u ftp. (K. ABSTRACT AN INTERACTIONAL APPROACH TO INVESTIGATING FOOD BUYING BEHAVIOR by Gordon W. Paul The need for obtaining more generalized knowledge about persons, when acting in the role of food purchasers, has been largely ignored by marketing research. The majority of marketing research studies have concentrated primarily upon obtaining quantitative meas- urements at the expense of building a body of causal relationships pertaining to food buying behavior. A combination of survey and ob- servational research techniques are needed to be used if interaction between buyers and sellers is to be investigated. This study reports on the initial phase of a three phase research program which attempts to identify and measure the market processes of the food distribution system serving an entire community. A body of store, consumer, and promotional knowledge was obtained in this research. The procedure involved operating a consumer food purchasing panel, conducting retail store audits, and monitoring all food market promotion and advertising simultaneously over a ten week period. Gordon W. Paul Audits of nine product categories, in food stores account- ing for ninety-five percent of total grocery sales made in this com- munity, provided a virtual census of retail food operations. The consumer panel furnished a weekly record of the buying behavior of a sample of households from the urban and rural community. Later phases of the research program propose to utilize data generated by this study to cross refer buying behavior with marketing efforts and sales results. At the conclusion of the ten week audit and panel re- porting period, information for constructing personality configurations for panel members, persons who had refused panel membership, and a random sample of community members was obtained. An investigation and evaluation into current types of food marketing research is made. The techniques and methodology em- ployed in the organization and operations necessary for conducting retail store audits are discussed along with the selection, recruitment, and maintenance of the consumer food purchase panel. Limitations, problems, and recommendations encountered in each stage of the re- search are discussed. Many of the techniques employed in this study appear worthy of consideration by other researchers in view of the excellent cooperation of respondents that was obtained. For example, of 325 panel recruiting contacts, there was an agreement to participate by 267 of these households, or eighty-one percent. Of those agreeing to panel membership, seventy-two percent became active members. Gordon W. Paul Store audits provided sales data against which panel pur- chases were compared. Comparisons between panel and audit figures were made for both private branded and national branded items. Panel data reflected an aggregate over-statement of purchases for both pri- vate and national brands when compared to actual sales data. However, further tests indicated that the difference was not significant for national brands, but was for private brands. Explanations are offered as to this discrepancy between audit and panel figures. AN INTERACTIONAL APPROACH TO INVESTIGATING FOOD BUYING BEHAVIOR BY Gordon Wilbur Paul A THESIS Submitted to Michigan State Univeristy in partial fulfillment of the requirements for the degree of DOCT OR OF PHILOSOPHY Department of Marketing and Transportation Administration 1966 ACKNOWLEDGMENTS The contribution of advice, financial support, efforts, and encouragement from persons too numerous to acknowledge in their entirty were made available to the writer. Mr. Martin Stoller, while serving as economist for MAR- PLAN Division of McCann-Ericksen and Dr. Charles Slater of Michigan State University provided the original idea and focus for the research. Technical advice provided the writer by Dr. I. J. Abrams, Vice- President Research, Marketing Research Corporation of America and Mr. William Epmier, Vice-President, A. C. Nielsen Company contributed substantially to the implementation of the research. Financial support necessary for this project was made available by the Food Marketing Program and Department of Marketing and Transportation Administration at Michigan State University. The research grant for preparation and processing of data generated by this study, and to be used in later analyses, was made available by International Business Machines. The efforts and energies of numerous field investigators were required. Without the assistance of graduate students enrolled in the Food Marketing Program at Michigan State University, this ii project would not have been possible. Retail store audits and panel recruiting were conducted by these students, and the time and effort each contributed is deeply appreciated. The writer's committee of Dr. Charles Slater, Dr. Robert Headen, and Dr. Leo Erickson offerred considerable advice and en- couragement throughout the research. In addition, Dr. James Shaffer, Department of Agricultural Economics, Michigan State University, willingly shared his vast knowledge of consumer panel operation with the writer. It is impossible to acknowledge the contribution made by my wife Gloria. Without her steady encouragement, patience, and sacrifices this project would neither have been instigated nor com- pleted. iii TABLE OF CONTENTS Page ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . ii LIST OF TABLES . . . . . . . . . . . . . . . . . . . . ix LIST OF EXHIBITS . . . . . . . . . . . . . . . . . . . x LISTOFAPPENDICES.................. xi Chapter I. INTRODUCTION................ 1 BackgroundofStudy.............. 1 Food Marketing Practice . . . 1 Complexity of Promotional Mix Decisions.. 2 Complexity of Product Mix Decisions . . . . . 4 Food Marketing Theory . . . . . . . . . . . . 9 Research Problem . . . . . . . . . . . . . . 11 Research Objectives . . . . . . . . . . . . . 13 Research Methodology. . . . . . . . . . . . . 16 Limitations.................. 18 Organization of Study . . . . . . . . . . . . . 19 II. SOURCES AND LIMITATIONS OF PRESENT FOOD MARKETING INFORMATION . . . . . . . . 22. Introduction.................. 22 Research in Food Marketing . . . . . . . . . . 25 iv TABLE OF CONTENTS (Continued) Chapter Page Need for More Generalized Studies . . . . . . . . 28 Survey and Observational Research . . . . . . . 36 Survey Research . . . . . . . . . . . . . . 37 Observational Research . . . . . . . . . . . 39 Audits and Panels as Extensions of Survey and Observational Research . . . . . . . . . . . 40 Retail Store Audits . . . . . . . . . . . . . 42 Uses of Store Audits . . . . . . . . . . . 46 Limitations of Audits . . . . . . . . . . . 50 Continuous Consumer Purchase Panels . . . . 54 Uses of Consumer Panel Data . . . . . . . 57 Limitations of Consumer Panels . . . . . . 51 Audits and Panels in Total Market Investigations. . 65 NeedforAudits............... 66 Need for Consumer Panels . . . . . . . . . . 67 Summary...................68 III. PREPARATIONS PRIOR TO FIELD INVESTIGATION . 71 Introduction . . . . . . . . . . . . . . . . . . 71 Community Selection Criteria . . . . . . . . . . 72 Isolation of Community . . . . . . . . . . 72 Population Characteristics of Foodtown . . . . 74 Cross Section of Food Outlets . . . . . . . . 76 Employment Base . . . . . . . . . . . . . . 78 Media Availability . . . . . . . . . . . . . 85 Obtaining Cooperation . . . . . . . . . . . . . 87 Warehouse Groups . . . . . . . . . . 87 Community Influences and Information Centers . 89 Independent Retailers . . . . . . . . . . . . 91 TABLE OF CONTENTS (Continued) Chapter Problems Encountered and Recommendations . Community Selection . . Cross Section of Food Outlets Obtaining Cooperation . Product Category Selection for Store Audits Auditability Broad Distribution Vitality of Movement Promotional Activity Categories Selected Recommendations for Selecting Product Categories . Diary Selection and Development Diary Limitations and Recommendations . Panel Incentive Program . Incentive Limitations and Recommendations Summar y . IV. EXECUTION OF STORE AUDITING AND PROMOTIONAL MONITORING Introduction . Period of Audits . Training of Auditors Execution of Store Audits . vi Page 95 95 96 98 99 99 101 102 103 103 104 105 107 109 111 113 115 115 116 117 120 TABLE OF CONTENTS (Continued) Chapter Page Audit Problems and Recommendations . . . . . 124 Obtaining Invoices . . .. . . . . . . . . . 124 New Products. . . . . . . . . 125 Weekly Dollar Sales of Cooperators . . . . . 126 Audit Forms . . . . . . . . . . . . . . . 127 Promotional Monitoring . . . . . . . . . . . 128 Summary . . . . . . . . . . . . . . . . . . 131 V. RECRUITING AND OPERATING THE FOODTOWN CONSUMERPANEL.............. 133 Introduction . . . . . . . . . . . . . . . . . 133 The Sampling Problem-Literature Review . . . 134 Foodtown Sampling Procedures . . . . . . . . 137 Sample Census . . . . . . . . . . 141 Selecting Areas and Households . . . . . . 142 Panel Recruitment . . . . . . . . . . . . . . 144 Panel and Area Demographic Comparisons . . . 147 Panel Cooperation . . . . . . . . . . . . . . 155 Panel Maintenance . . . . . . . . . . . . . . 158 Recruiting Replacements . . . . . . . . . 160 VI. PANEL RELIABILITY FOR BRAND REPORTING . . 162 Introduction . . . . . . . . . . . . . . . . . 162 Expenditure for Foodtown Panel vs. Other Sources 165 The Reporting Accuracy for Brands by Consumer Panels................... 169 Cross Referral of Foodtown Panel with Retail StoreAudits................ 173 TABLE OF CONTENTS (Continued) Chapter Method of Comparing Panel and Audits for Movement . . . . Test of Paired Comparisons . . Explanation of Results . Summary VII. SUMMARY AND CONCLUSIONS Introduction . Are More Complete Methods of Investigating Food Marketing Possible? Need for More Complete Studies of Food Marketing Strengths, Weaknesses, and Limitations of Foodtown Research Recommendations for Further Research Concluding Remarks BIBLIOGRAPHY APPENDIX viii Page 176 180 183 190 193 193 194 197 201 208 211 213 225 Table 10. 11. LIST OF TAB LES Family income for Foodtown, State of Michigan, and United States . Median years of education for Foodtown, State of Michigan, and United States Yearly sales of cooperating and non-cooperating food outlets servicing Foodtown Number of items audited in each product category for all cooperating retail food stores . Number of households agreeing to panel membership and numbers of those returning diaries by area for the ten week period . Income for sampled areas and panel members Income for selected ranges expressed as percents . Husband and wife's education for sampled areas and panel . Mean education for sampled areas and panel Size of family for sampled areas and panel . Panel members returning diaries ix Page 75 75 94 123 146 150 150 152 153 155 156 LIST OF EXHIBIT S Exhibits 1. Profile of retail food outlets in Foodtown . 2. Test of paired comparisons between store audit figures and panel data for twelve national brands 3. Test of paired comparisons between store audit figures and panel data for six private brands . Page 79 181 184 LIST OF APPENDIC ES Appendix Page A. Communications for Panel Recruitment, Maintenance, andOperation.................. 225 B. Retail Audit Training and Audit Forms . . . . . . . 262 C. Questionnaire for Panel Members, Non-Cooperators, and Random Sample of Community Households . . . 265 xi CHAPTER I INTRODUCTION Background of Study Food Marketing Practice The complexity of decision making for food executives when choosing among promotional strategy alternatives and developing prod- uct policies can be attributed to primarily two factors: (.1) the abundance of available alternatives from which to choose; and (2) the lack of ade- . . . . 1 quate market 1nformat1on relating to these alternatives. A number of other factors affecting promotion and product decisions are also present. Among these are included personnel turnover, management's influence on the decision, and communications within the organization. 1The Fabulous Von's Story (New York: Super-Market Publishing Co. , Inc. , 1960) , Part I. Mr. Harold Ward points out the need for a better understanding of promotional, as well as other mar- keting activities in the food industry. It is Mr. Ward's opinion that the role played by promotional activity is just beginning to be under- stood and present methods and data available could, if used correctly, provide for'better promotional decisions. 2'Patrick J. Robinson and David J. Luck, Promotional Decision Making: Practice and Theorl, Marketing Science Institute (New York: McGraw-Hill Book Company, 1964), p. 5. However, it was the consensus of executives interviewed by the Mar- keting Science Institute, in a study of promotional decision making, that lack of information was one of the most important deterrents to effective decisions. The 1963 Mid-Year Conference of the Super-Market Insti- tute focused upon the increasing complexity of retail food competition. This complexity was found present in the form of more stores, more promotional changes, more trading stamps, longer hours, more services, closer pricing, and tighter management for food retailers. There has been a decline in sales per square foot and a subsequent shrink in profits for the supermarket operator, as the cost elements of operation and the number of outlets have increased. 5 Although this is only reference to one portion of the retail food marketing system, the maturing character of the supermarket serves to emphasize the necessity for food management decisions based upon alternatives re- quiring data not available from present sources. The merchandise strategy alternatives considered and employed by the food retailer can generate additional profit dollars if 31bid.. pp. 24-25. 4Perspective for Decision Makers: A Study of the Emerg- ing Retail Environment (Chicago: Super-Market Institute, 1963) , p. 19. 51bid.. pp. 25-27. chosen well, but the myriad of alternatives makes such strategy choices exceedingly complex. A recent study illustrates the problem of making such decisions when only the variable of promotional pricing is con- . 6 . . . . Sidered. If promot1onal dec1s1ons are to be made Wisely, the food executive must have extensive information available about competition and consumers. The necessity of predicting competitive behavior and consumer response requires a continuous market information system both complex to administer and expensive to maintain. Complexity of Promotional Mix Decisions Food executives when choosing their promotional strategies“ must select from among a large number of alternatives and an equally large number of ways in which these may be combined. The decision to offer or not to offer trading stamps may illustrate the scope of such alternatives. The following questions illustrate the complexity of pro- motional decisions when the only promotional variable considered is trading stamps: 6 See: Wroe Alderson, ”Administered Prices and Retail Grocery Advertising, " Journal of Advertisingiesearch, Vol. III, no. 1 (March, 1963), pp. 2-6; George Fisk, Lawrence Nein, and Stanley J. Shapiro, ”Price Rivalry Among Philadelphia Food Chains, " Journal of Advertising Research, Vol. IV, no. 2 (June, 1964), pp. 12- 21; and Wroe Alderson and Stanley J. Shapiro, "Toward a Theory of Retail Competition, " Theory in Marketifl, ed. Reavis Cox, Wroe Alderson, Stanley Shapiro (Homewood: Richard D. Irwin, Inc., 1964), pp. 190-202. 1. If stamps are not to be offered, should the firm stress a policy of comparative price advertising with those competi— tors offering stamps as part of their promotional mix? 2. If stamps are to be adopted, should bonus stamps be given only for high margin items, and if so, how often should such items be repeated? 3. Should there be several days during the week in which extra stamps are given, and if so, which days, and is a minimum purchase required? The above are but a few of the more obvious questions needing answers if promotional decisions are to be made most effec- tively. In addition to trading stamps, also to be considered are the equally perplexing and wide choices of other promotional variables. Included among these would be discounts, gift premiums, and various in-store merchandising activities. In addition, underlying all such decisions is the question of who responds to the various alternatives once they are chosen and implemented. With such a wide range of promotional variations available and the amount of information at hand, it is difficult, if not impossible, to approach an optimal promotional mix. Complexity of Product Mix Decisions Difficulties are presented when attempting to arrive at an optimal mix of products. Scarce shelf space must be allocated efficiently between new items which have been introduced and those already carried in inventory. 7 A recent survey of leading grocery manufac- turers disclosed a total of 280 new items introduced by these companies per year, or an average of nine per manufacturer. 8 This was the equivalent of one new item for every nineteen established products these manufacturers produced. When one considers that the six thousand to eight thousand items stocked by the average supermarket comprise but a small portion of the total number available, it becomes apparent that the food executive must carefully select those items best serving his customers at a profit for his firm. Two of the many questions the addition or deletion of products presents to the retailer are: (1) will items considered for deletion because of movement and profit contribution be purchased elsewhere by a store's loyal customers if they are discontinued; and 7William Brown and W. T. Tucker, "The Marketing Center: Vanishing Shelf Space, " Atlanta Economic Review, XI (October, 1961), pp. 9-13. 8Grocery Business Annual Report-~1964 (New York: Progressive Grocer, 1964), p. F-2. 9Thirtieth Annual Nielsen Review of Retail Grocerj Store Trends (Chicago: A. C. Nielsen Company, 1964), p. 23. This report illustrates the difficulty that the supermarket operator has when faced with choosing among 35, 035 packaged items available and, in most cases, having only sufficient shelf space for 7, 000 such items. The problem is further intensified in certain departments such as frozen foods. Here there are three thousand items from which to choose and only three to four hundred that can be carried. (2) if so, will other purchases also be made somewhere else? Little information is available as to the product perception of the housewife and her sensitivity to a store's product changes. Product selection presents a difficult problem for the retailer, due in part to the lack of information upon which to base product mix decisions. The method to determine the optimal promotional and product mix appears relatively easy to outline but difficult to implement in practice. The retail food executive would need only determine the effectiveness of different promotions and product offerings by analysis of the increased sales attributed to these changes. The additional costs incurred by each promotion or product would then be weighed against additional sales and profits generated from the change. How- ever, much of the information necessary to follow such a procedure is not available. The inadequacy of information is a result of several factors. Among one of the more important of these factors is that of joint products. Additional or reduced promotional expenditures for certain products not only directly affects sales of the product receiving the promotional treatment, but also affects substitute and complementary product sales as well. Because cross elasticities between product categories are often unknown, the complexity of arriving at promotional and product mix decisions is further magnified. In addition, little is known about the decisional processes of food executives when adjusting promotional and product mixes. However, one of the principal problems in optimizing promotional and product mixes remains that of obtaining necessary marketing information and developing adequate methods for its analysis. Marketing management is aware that if a more complete knowledge of marketing interaction is to be obtained, it is necessary to have some continuous measurement of three factors. Included are a body of store knowledge, promotional and competitive measurement, and a body of respondent or consumer knowledge. To acquire such an amount of information on an aggregate basis appears an insurmountable task. However, a microscopic look at one particular market appears to offer a basis for obtaining additional insight into better understanding the independent marketing variables and their effects upon food shopping behavior. Such an exploratory investigation would, hopefully, determine the feasibility and possibilities such an approach has for further research. Several studies attempting total market investigations in one community have been made in the past. One study, attempting to explore milk consumption relationships, investigated an isolated market by saturation interviewing, census of dealers' sales records, and store sales. 10 Approximately $50, 000 was spent in this market of ten thou- sand persons and few meaningful relationships were identified. Another study attempted to determine the decision variables open to the entre- 10Dr. Max E. Brunk, personal correspondence, Dec. 16, 1964. preneur, rather than the variables impinging upon a particular com- modity. 11 Both of these studies employed different methodology from that prOposed in this research. Recently there has been diSplayed an increased recognition of the need for a continuous marketing intelligence program. New methods must be developed to monitor consumer goods markets if adequate knowledge of market segmentation is to be obtained. Demogra- phy alone is not an adequate measure for classifying market segments for establishing meanginful relationships, and the marketing intelligence system necessary would need to be more sensitive. This requires that the intelligence gathering techniques must be directed at segments of the consumer market at the same time that institutional brand and store loyalties are obtained. Differences such as buyer values, attitudes, and preferences must be considered as important as the traditional demographic characteristics now used in delineating market segments.13 New ways must be found to discover market segments based upon such considerations. These segments could then be used to develop the apprOpriate marketing mixes and strategies. 11 Bob R. Holdren, The Structure of a Retail Market and the Market Behavior of Retail Units (Englewood Cliffs: Prentice-Hall, Inc., 1960). laRobert Mainer and Charles C. Slater, "Markets in Motion, " Harvard Business Review (March-April, 1964) , pp. 75-83. 1 3Daniel Yankelovich, "New Criteria for Market Segmen- tation, " Harvard Business Review (March-April, 1964) , pp. 91-101. Because changes in the marketing mix do not affect a market uniformly, such methodology needs to provide a means of segmenting markets by both demography and personality configurations. If possible, this could provide the information necessary for analyzing the response to changes in marketing variables by different groups of consumers. Such a marketing intelligence program would require that data useful in establishing the pertinent dimensions best describing the market be generated. In addition, such a system would need to provide the neces- sary information required for the analysis of the cause of trends. This would require that an additional knowledge of consumers be obtained along with the monitoring of competitive activity. Food Marketing Theory The possibility of studying in detail food marketing in one particular community was discussed by Dr. Charles Slater with mem- bers of the food industry on numerous occasions. Several discussions held with Mr. Martin Stoller of McCann-Ericksen were directed toward the possibility of conducting an exploratory study designed to obtain store, consumer, and promotional information by utilizing present marketing research methods. Retail food store audits, consumer panels, and promotional monitoring all appeared to offer some contri- bution. Because each of these techniques has certain limitations, as presently used, an integration of these methods would be necessary. 10 If both audits and panels were used at the same time, discrepancies existing between these techniques could be investigated. The imple- mentation of a methodology whereby the buying behavior of households could be compared with the universe of food stores supplying their needs was viewed as a vehicle for providing information necessary for later investigations. These would be directed toward answering questions like the following: 1. To what extent do different market segments shift patronage as a result of promotional activities ? 2. Does sensitivity to promotion vary inversely with store . loyalty? 3. How do stores differ in their radius of customer draw? How, and to what extent, does this vary over time as merchandise or promotional variables are changed? 4. How do changes in promotional efforts affect the buying behavior of "bargain hunters"? 5. Do food chains compete more intensively with some rivals than with others ? Can this be verified by store ”switching'I patterns ? 6. Are changes in purchasing behavior and shopping patterns related to different personality configurations, and if so, can this effect be determined? Are panel members atypical of the population in their personality configurations ? Research techniques, as presently used, have been inade- quate for providing the necessary information for arriving at answers to the above questions. This can be attributed to the manner and com- bination in which they are used. In most cases, these techniques are employed to obtain quantitative measurements of brand movement and 11 market share rather than information useful for building a body of causal relationships relating to food buying behavior. It was because of the limited uses made of present methods that this research was conducted, and an attempt to combine several methods for investigating marketing interaction present in one community was made. Such an exploratory attempt at a more meaningful marketing information system appeared justified in light of the rising costs of promotion and the proliferation of product offering. Research Problem The problem visualized was that of developing and imple- menting a methodology whereby buying behavior could be compared with marketing efforts and sales. Information relating to the buying behavior of a sample of households in one market was to be cross referred with the sales results of the universe of food stores supplying their needs. From this cross referral the pertinent dimensions best describ- ing food marketing in this community were to be delineated. The magnitude of such an exploratory investigation into food marketing interaction called for a team effort by several persons. The complete research program consisted of the following three separate phases: l4Wendell Earle and John Sheehan, Operatirg Results of Food Chains, 1963-64 (Ithaca: Cornell University, 1964). 12 1. Phase 1: Methodology Investigation This phase includes a review of the need for such a research approach and the actual implementation of the major portion of the methodology. This phase of the project includes an evaluation of present sources of marketing information used by food marketers along with problems and limitations of these methods. An exploratory analysis illustrative of the possible research investigations such a study presents is included. The limitations, problem areas, and recommendations discovered in methodology imple- mentation are included to provide a reference source for researchers using similar techniques. 2. Phase II: Psychological Investigation of Consumer Panel Members An exploration of possible differences existing in the personality configurations of panel members vs. others is the major topic of this phase. The panel recruited and operated in Phase I of the research is utilized for part of this exploration. In addition, measures of store characteristics, personal influence on shopping behavior, and food shopping patterns are obtained and analyzed . 13 3. Phase III: Shopping Behavior Study Simulation models of food shopping in the market being investigated are included in this phase of the project. Phase III of the study uses data generated in the previous two phases of the research. The relationships and effects of certain marketing variables upon shopping behavior are explored in this portion of the research program. Methods of cross referring buying behavior with sales results are developed in this part of the research. This particular study reports 9111 on the Phase I portion of the research program. Because a team approach was necessary due to the comprehensiveness and magnitude of the project, there will occasionally be referrals made in this study to later phases of the research program conducted by other team members. However, an attempt has been made to confine such references to a minimum. Research Objectives The major research objectives of this study (Phase I of the research program) are threefold. Included in each of the three broad major objectives are several sub -objectives. These objectives are as follows: II. 14 To develop and implement a methodology whereby buying behavior can be compared with marketing efforts and sales result ology taken. A. s. The implementation and evaluation of the method- as the first phase of the research program is under- This includes: an examination of the literature to determine the need and applicability of the research methodology for studying food marketing interaction. the development, implementation, and evaluation of the methodology determining strengths, weaknesses, and limitations discovered in each portion. Sugges- tions for overcoming weaknesses in future generations of similar studies and the problems, limitations, and possibilities such an approach presents are analyzed. the generation of data necessary for later phases of the research program and suggestions as to possible approaches for analyzing this information. Future areas of exploration and tentative hypotheses to be tested using data generated in this phase are to be suggested. To provide other researchers with a source of information reflecting upon certain operational techniques oflthe study. Because of the secrecy shrouding operational aspects and 15 techniques of commercial research organizations using similar tools, little published literature is available to the student wishing to use these techniques . III. A. Included are: techniques and methodology employed in the selection, recruitment, and maintenance of a continuous con- sumer purchase panel. techniques and methodology employed in the organiza- tion and operations necessary for conducting retail store audits . To explore possible discrepancies which may exist between research techniques employed in the study. Retail store audit figures and consumer panel data are compared for: A. measures of aggregate over or under reporting of purchases by consumer panel members for national branded food items. measures of aggregate over or under reporting of purchases by panel members for private branded food products. purposes of evaluating consumer purchase panels for brand reporting by comparison to store audit informa- tion. 16 Research Methodology The research methodology obtains information necessary for investigating interaction between buyers and sellers of food items in one community. Information for later investigations into the causal relationships between people and goods was obtained by use of retail store audits and a continuous consumer purchase panel. Store audits obtained store and product information while the consumer panel pro- vided buyer information. The research methodology was a natural design in Phase I as only the events resulting from natural occurrences were observed and recorded. An experimental design for investigating personality configurations of panel members in Phase II was used. The research methodology was as follows: 1. A community relatively isolated for food shopping was selected for the research. The number of variables impinging on food shopping behavior in the community selected were reduced by selecting such a community. Community selection necessitated the presence of several retail food outlets serviced by various warehouse groups to insure a variation in merchandising between outlets. 2. Retail store audits for nine food product categories were taken weekly for ten weeks in the majority of food stores servicing the community. Product categories selected for 17 auditing were expected to receive promotional treatments during the ten week period. Price, space allocation, weekly product movement, and in-store promotional information for the product categories selected were audited. Total weekly dollar sales for the majority of stores were obtained. A consumer panel, large enough to discern major market segments, was recruited from the urban and nearby rural community. The consumer panel was operated concurrently with store audits for a ten week period. Weekly purchasing information relating to quantity, brands, prices, and stores shopped was obtained from panel members. A record of promotional variables for the panel operation and store audit period was maintained. Newspapers, radio, and television were monitored for food promotional activity. Consumer panel members were interviewed at the end of the ten week reporting period. Interviews obtained informa- tion for constructing personality configurations of panel members. Needs, values, measures of personal influence, and food shopping habits were obtained from panel members at this time. Interviews were also conducted at this time with persons who had been originally requested to partici- pate as panel members but had refused cooperation. In addition, interviews were also obtained from a random 18 sample of the community's population. All interviews attempted to obtain the same personality and demographic information . Retail store audit figures and consumer panel data were compared for product movement for both national and private branded items. A test of paired comparisons was made to determine the significance of any discrepancies present between consumer panel and store audit figures. Limitations Limitations appear inevitable in most research and this project was no exception. The problems presented by limited finances and time were ever present. In addition, the following limitations must also be mentioned: 1. The research was conducted in only one community. This may result in the inclusion of variables atypical to other localities. By purposely choosing a community where food purchasing alternatives are reduced, certain limitations are present. Panel members may well react differently in food shopping behavior if purchase alternatives are enlarged. However, this study does not purport to repre- sent other communities, and is only a statement of an exploratory investigation into one particular community. 19 2. All variables affecting food shopping behavior have not been obtained in this study. Only a limited amount of information has been secured. Certain omissions are present which would have effect upon precise results. 3. Because of the confidential nature of some of the data, information on community and cooperating companies is not as complete as may be desired. The identity of both the community and cooperators has been disguised, and in some cases, minor alterations have been made to preserve the anonymity of those extending cooperation. No changes have been made in the content of the data. 4. An erstWhile attempt has been made to be completely objective in both the collection of data and in an impartial analysis of the project. It is recognized that certain constraints may have induced the presence of some bias in the study, but it is the researchers' belief that a scientific method of inquiry has been maintained. Organization of the Study Partial explanation of the background and importance of the study has been presented in this chapter. This chapter establishes the. proper perspective of this study in relation to the research program and its importance to current food marketing practice. This chapter 581:8 out the research objectives for Phase I of the research program. 20 The objective of Chapter II is fourfold: (,1) to review present methods and sources of information available to food marketing man- agement, (2) to review the advantages, disadvantages, and limitations of survey and observational research methods, (3) to review the advan- tages, disadvantages, and limitations of consumer purchase panels and retail store audits as extensions of survey and observational research, and (4)to show the need for integrating both consumer panels and store audits if food marketing interaction is to be more fully investigated. In Chapter III, the preparation necessary for implementation of the methodology is presented. The primary objective of this chapter is to describe the operational basis for the investigation excluding consumer panel and retail audit execution. Throughout this chapter, the reader and later researchers are furnished with a step-by-step account of the limitations and problems encountered in the initial stages of this study. Recommendations made as a result of experience obtained are made throughout the chapter. The performance of the retail audits and promotional monitoring portion of the research is described in Chapter IV. The training necessary and field execution of store audits is discussed. Problems encountered in auditing are discussed with recommendations developed from the research. This research presents other researchers a Source of information relating to the technical aspects of store auditing. 21 Chapter V presents the selection, operation, and mainte- nance of the consumer purchase panel. The primary purpose of this chapter is to describe the operational techniques and problems encoun- tered in panel operation. Certain demographic comparisons are made between consumer panel households and other households in the areas from which panel members were chosen. An exploratory investigation using data obtained from the research is presented in Chapter VI. This analysis uses audit and panel data to investigate the consumer panel's performance in brand reporting. The analysis compares movements of selected national and private branded products as reflected by audit and panel data. Possible explanations are offered for discrepancies present between store audit and consumer panel figures. Conclusions are presented for the research in Chapter VII. The major strengths, weaknesses, and limitations of the research are reviewed at this time. Implications and recommendations for further research are made in this chapter. C HAPT ER II SOURCES AND LIMITATIONS OF PRESENT FOOD MARKETING INFORMATION Introduction An initial search of the literature was undertaken to pro- vide necessary background relating to present sources of food marketing information. This search focused upon present sources of information available and the manner these were utilized by food marketing man- agement. Periodicals, trade journals, and general marketing books were examined. This examination was directed toward obtaining a general overview of any similar marketing studies having been pre- viously conducted. At the same time, the limitations of marketing research methods now utilized were explored. The two most suitable research techniques for a study of food marketing interaction in one community appeared to be the retail store audit and the continuous consumer panel. The literature on panels and audits was then examined for limitations, current uses, and possibilities for application to this investigation. 22 23 Available published material concerning audits and consumer panels was insufficient in two respects. Most published material lacked information about the current uses of both consumer panels and audits. In addition, much of the available material reflected only upon the more obvious types of data generated. Continuous consumer panel and retail store audit references usually concentrated upon the use of these tech- niques for providing aggregate measurements of products sold at the retail or consumer level. 1 Brand loyalty and brand switching studies were usually cited as an extension of this information. The second deficiency was the lack of reference made to the limitations associated with these techniques. Much of this can be attributed to the same reasons that limitations are often under-empha- sized by many researchers. Consumer panels and retail store audits are conducted and operated, in most cases, by commercial research organizations. Of necessity, such organizations do not emphasize the shortcomings of their methodological approaches when this provides the source of revenue. It may be for this reason that firms continually stress the advantages associated with their techniques, as many of these same advantages are also limiting factors of competitors' techniques. 1Parker Holmes, Marketing Research: ' Principles and Readings (Cincinnati: South-Western Publishing Co., 1960) , Appen- dix A and C, pp. 597-629. These appendices are the basic information sOurce for published references to continuous consumer panels and retail store audits. 24 Because of the limited usefulness of published documents relating to these aspects, correspondence was directed to commercial marketing research organizations providing such services. This cor- respondence attempted to find answers to questions not answered in the literature. In addition to contacting the leading firms conducting retail store audits or operating continuous consumer panels, smaller organizations providing similar services were written. It was antici- pated there would be a reluctance on the part of commercial organiza- tions to divulge uses and limitations of their techniques. For this reason, agencies conducting or having conducted store audits or panels on a non-commercial basis were also contacted. Correspondence was received from university research groups and newspapers that had engaged in such research. Information about panel advantages and limitations was obtained by correspondence with fourteen individuals and firms. Because of the dominance of one organization conducting retail store audits, fewer individuals were contacted. Seven different organizations and individuals provided information relating to the uses and limitations of retail store audits. Correspondence provided useful information pertaining to limiting considerations, but little on the 'manner in which audits and Panels are utilized, beyond that obtained from published sources.‘ Visitations were made to the A. C. Nielsen Company, Market Research COrporation of America, and The Chicago Tribune to solicit informa- 25 tion on operational techniques, while at the same time hoping to dis- cover additional uses of audits and panels. Interviews conducted were extremely valuable in providing limitations and uses of audits and panels. Problem areas one might expect when employing these techniques in a total marketing study were also emphasized by those interviewed. Research into Food Marketing Numerous studies pertaining to the detailed technical or operational aspects of food marketing have been conducted in recent years, but few have directed themselves to general retailing strategies. Many of these studies attempt to isolate and determine the effect of a single independent variable upon product movement. Such studies frequently focus upon the relationship between shelf space, price, or placement and sales performance at the retail level. Representative of such studies is that by Coxe. Coxe, attempting to measure the relationship between product shelf space and sales, selected only those products having the same price during the test period with no sales promotion or advertising For an example of the specific nature of many studies see: flex of Super Market Articles (Chicago: Super-Market Institute), Vol. I-VIII. .3Keith Coxe, "The Responsiveness of Food Sales .to Shelf Space Changes in Supermarkets, " Journal of Marketing Research (May, 1964), pp. 63-68. 26 treatment present. Without a measure of promotional activity for both the cooperating stores and competitive activity, the meaningfulness of the results is questionable. Products that indicated a responsiveness to changes in shelf space may have only reflected changes in the total marketing effort of the retail outlet. Because the housewife purchases a "market basket" rather than isolated products, some initial measure of store choice in relationship to competitive activity would be neces- sary. Pessemier has stressed the need for studies attempting to measure relationships existing between shelf space and produce move- 4 ment as follows: Ideally, . if 'management knew just how sensitive were the sales of each item to the particular space allocated to it and just how large a contribution each item produced, it would be possible to make allocations which would return the largest profit. Studies made, with food industry sponsorship, have attempted to determine space and product movement relationships and product . . . . . 5 categories profit contribution such as Pessemier stresses. One source of difficulty with these studies is observed. The results too 4Edgar Pessemier, "Applying Supermarket Techniques to Non- Food Retailing, " Journal of Retailifl (Summer, 1960) , Vol. XXXVI, pp. 108-113. 5See: The P £C Food Markets Study: Projecting Product Prof- itability (Ithaca: Eastern Market Research Service, Inc. , 1964) and A Top Management Approach to Meat Merchandising (Washington: National Association of Food Chains, 1964) . 27 frequently show large profit contributions by certain product groups. Studies of shelf space allocation and product profitability contribution leave the impression that shelf space should be expanded indefinitely on high profit items. Coxe implies that hypotheses formulated about the responsiveness of food product sales to changes in shelf space allocation are too generalized and in need of additional refinement. However, at this time, product profit contribution approaches have remained unquestioned. Most product profitability and shelf space research has been confined to the analysis of a few stores of a cooperating chain or all of that chain's widely distributed member stores. 7 Each store has been observed in relation to the independent and dependent variable under investigation. As a methodology of discovering market inter- action, this type of approach may be likened to the "individual-unit" 8 procedure of sales forecasting. Many food marketing studies appear 6Coxe, op. cit., p. 66. 7Examples of this approach are five major studies conducted for Proggessive Grocer. Individual reports are available for Providence Public Markets, Foodtown Stores, Super-Value Stores, Dillon Stores, and Colonial Stores. Each of these studies emphasizes operating efficiencies and refinement of food marketing techniques. 8D. Maynard Phelps and J. Howard Westing, Marketing Ma_r1_agement (Homewood: Richard D. Irwin, Inc., 1960), pp. 241-244 and Harry Deane Wolfe, Business Forecasting Methods (New York: Holt, Rinehart and Winston, 1966), pp. 155-156. 28 to utilize such an approach. First, each separate variable is observed and then an attempt to aggregate these into a market model is made. If a more complete knowledge or understanding of food marketing inter- action or consumer dynamics is to be obtained, a different approach is necessary. Need for More Generalized Studies Food retailing may be described as being a "transaction" business. The retailer earns a profit on the customer's total "trans- action" not on each of the individual products making up the total "trans- action. " If food retailing operates in this manner, then it may appear useful to look first at aggregate food marketing behavior. The size and makeup of the total "transaction" may be influenced by merchan- dising skills. However, before investigating individual products and the merchandising skills affecting them, it would appear more bene- ficial to observe marketing interaction on a more general basis. This may lead to establishing those areas of food marketing behavior ap- pearing most profitable for further exploration. One recent and widelyfquoted food marketing study cites one of its limiting factors to be that data generated from the research is only part of the total information needed for marketing decision making. 9 Even though Direct Product Profit data is useful, as 9McKinsey General Foods Study: The Economics of Food Distributors, (White Plains: General Foods Corporation, 1963) p. 36. 29 illustrated by the McKinsey study, additional information about con- sumer demand, competitor's practices, price/volume relationships, consumer price awareness, space/movement relationships, and numerous other merchandising considerations are necessary. 10 The McKinsey study, by inference, points to the shortcomings of such ap- proaches and the need for a more complete knowledge of consumer dynamics. In the past food retailers have operated under the assump- tion that marketing efforts should be directed toward the average con- sumer. This has been done, in many cases, to obtain those economies available from a centralized operation. This has resulted in the marketing mix of many, if not the majority, of food retailers being one of likeness. As mechanical and technical features of retail food operations become more widely understood, this sameness will likely be increased. While the marketing effort has become more similar on the part of many retailers, there has, at the same time, developed a recognition that there is no such thing as an "average" customer. 11 Instead, the market is viewed as consisting of many sub-markets com- prising certain characteristics. It appears that one of the more im- portant needs for food retailers today is a basic knowledge about 10Ibid. , p. 37. 1 Walter Weir, On the Writing of Advertising, (New York: McGraw-Hill Book Co., Inc., 1960) p. 95. 30 consumers as demographic and psychological groups and their relation- ships to the marketer. In discussing retail research, Prof. Hollander emphasized that suitable techniques have not yet been developed for determining how stores are selected by such groups, the effect of retail advertise- ments on such groups, or the effect of price changes on sales. 2 The need for generalized research covering a wide range of items and situations was further accented. Prof. Hollander stressed that initial retail research attempts should be directed toward determining who customers are, with later research directed toward determiningwhy they make the selections they do. If the interaction that takes place between the customer and those providing her needs is to be better understood, it may first be necessary to study markets in the aggre- gate. Those areas appearing to warrant detailed research may then be investigated further. Such an approach, like the "over-all" sales forecasting procedure, would start with the "general" and proceed to the "particular. " Generalized approaches for investigating market interaction have seldom been used in the past for two reasons: ( 1) the costs in- curred by such a project; and (2) the lack of adequate methods of 12Stanley C. Hollander, "Retail Research, " Business Topics, Vol. VII, No. 2, (Spring, 1959) p. 52. l 3Phelps and Westing, op. cit., pp. 244-249. 31 analyzing the large amount of data such research generates. One of the researcher's purposes in conducting this total marketing study was to determine the costs of such a study. Correspondence with Dr. Brunk, of Cornell University, indicated the costs of a previous marketing study under his direction for one commodity in a single community to be fifty thousand dollars. When one considers the yearly cost of a bi-monthly national audit report for one product's movement to approximate ten thousand dollars, an analysis of the consumer dynamics influencing this product movement would seem to warrant at least the same expenditure. Such an expenditure, in even a limited market, could lead to results useful for planning mar- keting strategy for that product. The availability of adequate financial resources was one of the limiting factors of this research. This same limitation also was advantageous because it established parameters determining the extent to which the investigation could be conducted. The total direct expenses for this research were four -thousand dollars. Because much of the data processing remains to be done, this figure only includes those costs incurred to the completion of field investigation. This figure can be misleading because of the contribution of time and effort by graduate marketing students. If these efforts had been reimbursed, the total field costs for this food marketing study would have approximated ten-thousand dollars. Even if this amount 32 were to be doubled by future analysis of data, such research appears relatively inexpensive to food industry members when compared to the costs of commercial research now being purchased. 14 However, this cost must also be considered in respect to the extent of geographi- cal coverage. Although the information obtained from present com- mercial sources is based upon wide geographical dispersion, more complete investigations into a limited number of selected communities may provide information useful in making some marketing decisions, such as allocating promotional expenditures, that is not obtainable from present commercial store audit and consumer panel data. There are indications of a growing awareness on the part of food industry members for studies investigating local or district markets. One such study now being conducted and a proposed study, by one of the leading commercial marketing research organizations, illustrates this increased interest. A report on a market study sponsored by Kroger Company, Progressive Grocer magazine, and Reuben H. Donnelley Corporation was presented at the 1965 annual meeting of the National Association "Researchers Snap up Supermarket 'Secrets, "' Business Week, (March 5, 1966) p. 83. J. Clarke Mattimore refers to a charge of $50, 000 a year a product classification for national reports indicating movement as obtained from chain warehouse records. 33 1 of Food Chains. The first report of this research states that the: 5 first stage of study is an examination of differences among different consumer groups in terms of store selection, cate- gory and brand purchases, food budgets, attitudes, habits, problems, wants, satisfaction, complaints, reactions to pro- motion, new items, prices, stamps, perishables and hundreds of other factors that should and do concern modern supermarket operators. The methodology for this research begins to approximate a total food marketing study. Five distinct neighborhoods in Cleveland, Ohio, were selected on the basis of family size, age, income, and education. In each of these five areas, two Kroger stores were selected for audits of all items in units, dollars, and dollar margins over a period of eight weeks. A consumer panel was established con- sisting of five hundred families. One hundred panel members were selected around the location of each of five Kroger stores in the five areas. Each family was to keep a diary of all food purchases for a period of eight weeks. In addition, four hundred telephone interviews were conducted in those five trading areas. Because of the need to obtain information not available through the use of a consumer panel diary, depth interviews were conducted with 125 panel members. These depth interviews were designed to elicit information about store personality, promotional mix, and information most useful to the con- sumer in food shopping. 5 . . . . "Consumer Dynamics in the Super Markets, " Progesswe Grocer (Oct., 1965) p. K-2. 34 Newlyweds, recent movers, and ethnic groups are to be the subjects of additional investigations. The researchers anticipate that the diary and store audits will provide a portrait of sales in each of the five demographic areas under study. This should then lead to a projectable sampling of food purchases from all food stores by 100 representative families. Market Research Corporation of America is considering a similar approach to investigating consumer dynamics in Boston, Massachusetts. 16 Dr. William Applebaum of Harvard University has developed map overlays consisting of equal lines of income, age group- ings, and other demographic measures. From these maps will be drawn, randomly, sample communities which represent the total variance in the metropolitan market. A consumer panel is then to be established surrounding each community. Instead of then following the Cleveland practice of auditing selected stores, warehouse with- drawal slips are to be utilized. This project, like the Cleveland proj- ect, would attempt to discover the consumer dynamics present in food marketing in this particular community. Such a study would hopefully result in a short term predictive model of consumer sales. 1 6Personal interview with Dr. I. J. Abrams, Director of Research, MRCA, June, 1965. 17Samuel G. Barton, "Model for Short-term Prediction of Consumer Sales," Journal of Marketing, (July, 1965) pp. 19-29. 35 A step toward establishing a complete information service on supermarket product movement was recently made by the Market . . 18 . . Research Corporation of America. MRCA Signed an excluswe agreement to purchase warehouse inventory data from Kroger with which it plans to: track the pattern of usage and purchase beyond the warehouse, through the store, and right into the home. The company has obtained so-called "observation privileges" in Kroger and Grand Union stores, in order to measure such things as shelf space and display arrangements. Next step will be to set up market-by-market test panels to measure in-home use. Studies such as the Progressive Grocer Cleveland research and the proposed MRCA Boston study utilize two types of measurement; a consumer panel for purchases by household and retail store audits or warehouse withdrawals to secure measurements of total product movement. Because an important part of this research utilized both audits and panels, it is necessary to investigate further their uses, advantages, and limitations, and show the need for integrating the two. Continuous consumer purchase panels and retail store audits have both been referred to as extensions of survey and observational re- search. Because of the apparent confusion existing, it is first neces- sary to briefly review survey and observational research methods. Retail audits and consumer panels then will be further discussed as techniques of each type. \ l8"MRCA Gets Rich Supermarket Data, " Business Week (April 30, 1966), p. 134. 36 Survey and Observational Research Data collection methods have been classified in different . . 19 . . ways by various writers. In marketing research, the collection of data usually utilizes a methodology of questioning or observation. These two methods are distinguished by the actions undertaken by the researcher in securing information. In observation, the researcher asks no questions, but instead observes objects or actions of interest. The questioning method involves securing responses to questions either in person or in writing. Regardless of the type of research design used, "the necessary data are collected by one or both of these two methods. " Questionnaire research is often referred to as survey re- . 22 . . . search by many writers. Because of the value which descriptive survey research can provide, many different techniques have been developed. As a result of this proliferation of techniques, one writer has stressed that survey research has merged with observational 19Fred T. Schreier, Modern Marketipg Research: A Behavioral Science Approach, (Belmont: Wadsworth Pub. Co., 1963) p. 34. ‘ 2 0Harper W. Boyd, Jr. and Ralph Westfall, Marketing Research '( Homewood: Richard D. Irwin, Inc., 1964) p. 148. 21Ibid. ZChester Wasson, The Strategy of Marketing Research, (New York: Meridith Publishing Co., 1964) Chapter 5, pp. 130-164, and Holmes, op. cit., p. 222. 37 research on one side and experimental research on the other side, as . . . 23 . . . a means of gathering information. The distinction between survey and observational research may be considered minor, but this dis- tinction becomes important when the uses, advantages, and limitations of each are considered. This distinction is important because each is best suited for obtaining certain types of marketing information. Survey Resgarch Survey research studies both large and small populations by usually selecting and studying samples chosen to discover incidence, distribution, and interrelations of variables. Studies of samples are utilized because of the difficulty of studying entire populations. The survey approach and the techniques used focus upon people, facts of people, their beliefs, opinions, attitudes, motivations, and behavior in a marketing situation. 24 The most common survey techniques utilize mail, telephone, and personal interviewing. The techniques employed in survey research depend on some form of verbal or wirtten response to a question or other stimulus. Survey research results can only be considered a logical implication as to the actual response which would result in a real situation. This 23 Wasson, op. cit., p. 138. 24tFred N. Kerlinger, Foundations of Behavioral Research, (New York: Holt, Rinehart and Winston, Inc., 1964) pp. 392-408. 38 occurs from the reliance that must be placed on verbal or written re- sponses. Because of the versatility, ease of administration, speed, and associated costs, survey research has found wide application as a method of securing marketing information. Wasson has discussed several of the underlying assump- tions of the survey approach. 26 These focus upon the significance, validity, and relevance of the information obtained, because of the differences in meaning attributed to words and the individual reactions of those responding. It is because of these assumptions that some ad- ditional method of determining and recording actual behavior may be needed. If an accurate picture of the market situation is to be obtained, observational methods may be necessary to alleviate some of the prob- lems or deficiencies present when the survey approach is used alone. For an overview of the procedures and development of survey research see: A. Campbell and G. Katona, "The Sample Survey: A Technique for Social Research, " in L. Festinger and D. Katz, Research Methods in the Behavioral Sciences, (New York: Holt, Rinehart and Winston, Inc. , 1953) Chap. 1 and Richard D. Crisp, Company Practices in MarketingResearch, American Manage- ment Association Research Report 22, 1953. For the most compre- hensive source on the entire survey method see: Mildred Parten, Surveys, Polls, and Samples: Practical Procedures, (New York: Harper, 1950) . 26Wasson, op. cit., p. 138. 39 Observational Research In one manner, it is possible to think of observation re- search as a part of all data collection and not a separate method. Everyday observation allows the formulation of the necessary infor- mation required for survey research. The type of observation re- ferred to as a distinct method of research differs because it is systematic in approach rather than consisting primarily of random observations. Necessary characteristics for observational research to be considered a separate type of marketing research are: (1) it must be systematically planned to focus on specific propositions of the study; (2) it must include a systematic recording of events; and (3) it must utilize planned checks and controls to insure the completeness of the observation. The use of the observation technique permits actual be- havior to be recorded by relying primarily on physical observation of One or more phenomena under study. There is disagreement among marketing writers as to whether self-recorded observation by a re- sPondent should be considered or only that reported by an outside Observer. Wasson would consider self-observation recorded and 2 7Richard D. Crisp, Marketing Research, (New York: MCGraw-Hill Book Co., Inc., 1957) pp. 179-227. 40 reported to be a part of the category encompassing observational re- search whereas Crisp includes self—reporting as an extension of survey 28 . . . . research. This disagreement has led to con51derable conquion for classifying continuous consumer purchase panels and retail store audits as being either survey or observational research techniques. Classification of Audits and Panels as Survey or Observational Research Obtaining some evaluation of changes occurring over a period of time is an objective of both audits and consumer panels. It is for this reason that both have been occasionally referred to as the panel technique. Both utilize a panel of respondnets remaining some- what constant. This is done to eliminate the possibility of differences occurring between two separate samples when taken over a period of time. By utilizing a panel of respondents and observing this same panel over a period of time, it has been possible to obtain measure- ments on such items as real shifts in consumer usage rather than that 28Wasson, op. cit., p. 132 and Crisp, op. cit., pp. 209- 22.7. 2C"One such example is provided in Crisp, op. cit., pp. 209-227. In discussing panels as a technique of survey research, the A.C. Nielsen store auditing service is used as an example. Later dis- cussion of observational research refers to this firms methodology. Any existing confusion results from classifying the stores used in retail store audits as a "panel" of stores. There appears to be some agree- ment that panels are a form of survey research when using consumers' as reporters and observational research when used for retail store auditing. 41 which may have occurred by a chance variation in sampling. Although both utilize a panel of respondents, (households, retail stores, manu- facturers, professional persons, etc.) the basic method of recording the data varys. Many problems peculiar to consumer panels also are present with retail store audits. There are important differences serving to distinguish each technique even though similarities as to uses, advantages, and limitations of each are present. The primary distinction is in the manner in which the data is collected. In the case of the continuous consumer panel the necessary information is provided by the respondent, in this case an individual consumer or some other member of the household. Although the retail store audit may utilize a panel of stores, the information is systematically gathered and re- corded by the researcher or his staff by direct observation. Observational research is many times only used when survey methods are found to be inadequate in providing desired infor- mation. When there is question about the accuracy of information supplied by survey methods, and when a high degree of accuracy is required, observational techniques may be desired. The observational approach is often found necessary when data concerning the actions of competitors is desired and it is unavailable by other means. If one is to approach consumer dynamics by a total mar- keting investigation, it appears that observation can provide some of 42 the quantitative findings such as Alfred Politz stresses are needed. 30 Nevertheless, opinions, attitudes, and preferences are the basis of purchase behavior and it may be necessary to investigate these through survey research. This research utilized both survey and ob- servation methods by including retail store audits and a continuous consumer panel among the employed techniques. Continuous consumer panels are considered a technique of survey research and retail store audits as observational research throughout this study. One can infer that the limitations and advantages of consumer panels and store audits discussed are also limitations and advantages of the broader classifi- cations of research methodology which they illustrate. Retail Store Audits Retail audit research accounts for a large percentage of the commercial research purchased by the consumer packaged goods industry each year. Audits supply the subscriber of the service with useful information about sales, inventories, distribution of his and 30Alfred Politz, "Motivational Research from a Research Viewpoint, " The Public Opinion Quarterly, Vol. XX, No. 4, (Winter, 1956-57) pp. 663-673. Mr. Politz states: "If consumer surveys are to predict consumer behavior in a way which will help management to make marketing decisions, they should come up with quantitative find- ings. As well as being more useful, this is a requisite of the scientific approach. It is also more reason why definite reactions to stimuli must be studied rather than opinions, attitudes, and preferences which as such cannot be reduced to weights or measures. " 43 competitors' product, and some limited measure of promotional activity. This information is used by subscribers to make estimates of national or regional market trends and market penetration. Retail store audits are defined as: The measurement of sales of a product, or brand shares of a product category, determined by inventory and invoice checks in a sample of cooperating retail outlets, usually periodically. Retail store audits were first developed commercially dur- ing the 1920's, when manufacturers of consumer packaged food and drug items developed a growing awareness that the lag in distribution prevented sensitive measurements about quantities and rates in which final consumers were purchasing products. In 1933 the A. C. Nielsen Company inaugurated a marketing research service later known as the . 32 . . . . Nielsen Food-Drug Index. Although many firms prov1de Similar ser- vices, this organization is the largest commercial firm conducting re- tail store audits. Conversations with Nielsen personnel revealed that most major food manufacturers subscribed to the Nielsen service. Retail store audits, a method of observational marketing research, will be analyzed from the scope and method in which they are conducted 31Robert Ferber, Donald F. Blankertz, and Sidney Hollan- der, Jr., Marketig Research, (New York: The Ronald Press Com- pany, 1964) p. 662. 32 . . . . - A Brief DeSCI‘HDtIOII of Nielsen Food Index and Nielsen DigiIndex, (Chicago: A. C. Nielsen Company, 1957) p. 1. 44 by this organization. Before any assessment of the advantages and limitations of the retail store audit as a method of observational re- search can be made, it is first necessary to review the need for audits and methodology employed in such research. The lag in distribution resulting from the use of intermedi- aries causes a difference between manufacturer's factory sales and consumer sales to exist. Because the manufacturer's product does not move directly to the consumer but moves first through the various chain and wholesale warehouses and retail food stores, inventories are present throughout various levels of distribution. These inven- tories result in an appreciable difference between factory sales and consumer sales. Of course, factory sales information is already available to the firm, but the additional measurement of consumer sales is needed if there is to be adequate information about distri- bution, stock cover, display, and sales to consumers. It was in an effort to provide such information that retail audits were first de- veloped for commerical purposes. Most retail store audits conducted on a commercial basis are usually utilized for arriving at aggregate sales for a product and competitive products at the consumer level. The basic principle of the retail store audit is that re- tailer's sales over a certain period of time equal an opening inventory plus any purchases during this period minus present stocks at the time of the audit. Stocks at the beginning of the inventory period are physically 45 counted and invoices and similar records are observed to determine additions during the period under observation. The residual resulting from opening inventory, additions, and ending inventory is the amount purchased by consumers during the period. Most audits are conducted on a national basis although special audits are available for certain test markets. Stores selected for auditing are selected in such a manner that each store type, neigh- borhood served, and store size are represented in the proper proportion. However, certain chain stores are not included and the proportion of small stores audits remains questionable. In the case of the A. C. Nielsen audit service, stores‘audited have been selected using prob- ability methods. The Nielsen organization uses approximately 1600 retail food stores as the base for its service. Retail units contract with the Nielsen company to allow auditors to take inventories, check invoices, and record promotional activity every 60 days in exchange for cash compensation. This compensation varies with the type and size of retail store and amounts to approximately $15-$35 for each audit. Consumer sales figures obtained from this sample of stores is then expanded to a total for the entire country. Store audits are usually a technique employed in test marketing research for new products. The A. C. Nielsen Company provides several different facilities for measuring results of market 46 tests. 33 The stores audited for the national store audit may be broken- out by areas, districts, or limited cities. The same type of informa- tion obtained in the regular bimonthly report of the Nielsen Food-Drug Index is available for these smaller markets. Special analyses appli- cable to the test operation may also be obtained. Such analyses include shelf space measurements, special display evaluation, stores redeem- ing coupons, and age of stocks. Special test stores may also be audited in markets which are not Nielsen Test Cities or which are too small to permit a breakout of the existing sample used for the Nielsen Food or Drug Index. The stores chosen for auditing in these smaller markets are chosen on the basis of a stratified disproportionate sampling with ratio estimation. The universe is stratified with a higher percentage of large stores in the sample and combined with the ratio of each pro- duct's sales to total store sales in projecting total consumer sales for the product under investigation. 34 In both the test market store audit and those used for aggregate consumer sales data, only a sample of stores are audited. Uses of Store Audits Information obtained from retail store audits is analyzed and used in various ways. Although the majority of subscribers of commercial audit services utilize audit data for analyzing aggregate 3'3Markets in Miniature, (Chicago: A. C. Nielsen Company, 1964) p. 3. 34Ibid., p. 6. 47 sales and distribution, test market analyses are also made. The A. C. Nielsen Company provides the following list of data every sixty days for food stores by brand, territories, population range, stores, package size, and product type: sales to consumers _ purchases by retailers retail inventories days' supply store count distribution all-commodity distribution out-of-stock prices (wholesale and retail) special factory packs dealer support (displays, local advertising, coupon re- demption) special observations (order size, reorders, direct vs. wholesale) 1 total food store sales (all commodities) 13, major media advertising (secured from other sources) OxomxlO‘U'In-bUONr—n l—l I p—a I—I O p—u N The above information is usually used by the client in an effort to determine his consumer sales and those of competitors. Using retail audits, it is possible to obtain a measure of total con- sumer sales for all brands, and then break these down into percen- tages of the total market for each particular brand. These breakdowns can be further stratified into particular market segments by territory, population groups, store types, and store sizes to provide a basis for allocating promotional efforts. By measuring sales at the point of consumption a limited knowledge is gained about each particular market. 35A. C. Nielsen Company, op. cit., p. 5. 48 The sales accounted for by each particular store type (chains vs. in- dependents) is available and if corrective efforts are needed, audits can provide some necessary information useful in gaining distribution in weak areas. Cents-off promotions have been investigated as to their effectiveness by the use of retail store audits. 36 As is the case with most advertising or promotion, many variables enter the marketing "mix" and the influence of only one of these is difficult, if not impos- sible, to determine with any precision. Nevertheless, audit informa- tion, within broad limits, can be useful in determining the effect of the price promotion on the share of the market held by the particular product receiving the promotional treatment. Most such studies, which employ store audit techniques, investigate by case histories certain brands and promotion on an aggregate basis. Product category groups are chosen which have utilized the type of promotion being in- vestigated. Within these categories, brands are evaluated as to market share, based upon the size of the cents-off "deal" expressed as a per- centage of the brand's average selling price. Some measure of the effectiveness of the "deal" is obtained by comparisons of the sales of competing brands in that product. This is possible because store audits provide a measure of sales for all brands within the product category. 36"Cents-off Promotions, " The Nielsen Researcher, No. 1, (Chicago: A. C. Nielsen Company, 1964). 49 Conclusions based upon only store audit information are rather limited because variables such as advertising, other promotions, product quality, and those responding to the promotion are unavailable. Regard- less, some broad conclusions are usually arrived at on an aggregate basis. 37 Because conclusions often are so general, they often only serve to stress the need for more sensitive research into the problem area being investigated. In addition to studies designed to measure such variables as promotional deals and prices, store audit information has been used in determining sales trends by sizes, importance of store types, sales trends by areas, turnover, and exposure of product categories and in- dividual brands. 38 Such studies are normally made on an aggregate basis and the information used in the analysis is that available from a national retail store audit. This does not void the use of the retail store audit and similar analyses being utilized in smaller well defined markets. Information obtained from retail store audits conducted in small well defined test market cities is often analyzed in the same 7111c}. , p. 10. The single conclusion reached from this investigation into cents-off promotions was that cents-off deals, when large enough, can be effective on a short term basis. 38J. O. Peckham, "Guideposts for Pricing, “ A Presen- tation to the Grocery Manufacturers of America, (White Sulpher Springs, June 17, 1963) and Franklin H.V Graf, "Marketing Develop- ments in the United States, " A Presentation to the Foundation for Branded Consumer Goods Marketing Seminar, (Amsterdam, November 14, 1964) . 50 3 manner as that obtained from national store audits. 9 The retail store audit has found considerable use as a research technique because of the amount and type of information it provides which is unavailable by 40 . . . . other methods. Exposure of product categories, indiVidual brands, and monitoring of in-store merchandising activity are but a few examples of market information available exclusively by use of retail store audits. Limitations of Audits Many of the limitations concerning the use of retail store audit research center about errors resulting from the work of field in- vestigators. Stonborough has pointed out the following types of errors occurring in audits: 1. arithmetical errors which result in counting stock on shelves. 2. errors in omission of inventory stocked in out-of—the -way places. 3. errors due to missing invoices. 4. clerical and arithmetical errors on invoice proper. 5. errors in overlooking credits and computing sales. 6. errors due to lack of records and convenience pickups of merchandise. 7. sampling errors in selection of "panel" of stores. 392532133, (Chicago: A. C. Nielsen Company, 1957), 40Markets in Miniature, (Chicago: A. C. Nielsen Company, 1964) . 41 Thomas H. W. Stonborough, "Fixed Panels in Consumer Research, " ed. Hugh G. Wales and Robert Ferber, Marketing Research: Selected Literature, (Dubuque: Wm. C. Brown Co., 1952) p. 111. 51 Many of these errors can be reduced to a minimum by the use of well trained field auditors. The main limitation is that retail store audits provide no direct analysis of consumers. The retail store audit tech- nique does not furnish data which relates purchases directly to the people making the purchases. While providing certain knowledge as to sales, prices, and promotional effort, there is no knowledge of w_h9 accounted for such sales or £153 responded to m promotional effort. It is for this reason that retail store audits, by themselves, are not an adequate research tool to describe food marketing from a total perspective. By selecting only a sample of stores to be audited, certain pro— ducts' total sales are understated.42 This understatement results from added strength of some branded products among chains when compared to independent sotres and because of non-cooperation by some chains. When audits include only certain chains' member stores, an insufficient knowledge of promotional activity as well as product dis- tribution is present. The Metropolitan Supermarket Audit Division of Marketing Re- search Corporation of America is designed in such a way to eliminate part of the non-cooperator problem. By making no attempt to secure 42J. A. P. Treasure, "Retail Audit Research, " The Incorpor- ated Statistician, (October, 1953), Vol IV, No. 3. 52 necessary information for projecting product sales, it is possible for this organization to take audits of a limited nature. The Metropolitan Supermarket Audit only reports distribution, space assigned to brands by type and package size, price of the products on the shelf, and other visual measurements such as special displays and location. This organization's audit utilizes a standing sample of 2268 supermarkets located in fifty-four key metropolitan areas each having yearly sales exceeding $1, 000, 000. With no working contact with the sample organizations audited, it is possible for the Metropolitan Supermarket Audit to provide data relating to brand distribution by chain organi- zations. The subscriber to this service obtains distribution and the amount of promotion given his brands by chain classifications. By limiting the sample to million dollar supermarkets, this audit over- looks many of the more important outlets for certain categories of food products. The lack of adequate promotional monitoring is one of the major limitations of store audits when used as the sole measure of a product's performance in a test market. Some manufacturers attempt to distort any national projections made from test market information, by loading the competitor's test market with extra advertising and pro- motional features during the test period.43 This competitive counter- strategy requires that manufacturers using audits for test market 43"Colgate vs. P. #G” " Forbes, (Feb. 1: 1966) P- 28° 53 information rely on more than audit information alone in evaluating product performance. Some measure of the amount of competitive promotional effort must be obtained if product performance is to be judged accurately. Recently, several new companies have been organized to purchase and process merchandise movement information from chains and wholesalers and resell it to manufacturers on a regular basis.44 Limitations of present retail store audits have been directly responsible for the development of these new market research organi- zations. Food manufacturers find audit information, such as A. C. Nielsen Co. provides, adequate on a national basis but lacking in reliability for individual market areas. This occurs because of the sampling techniques employed and the sample size of present audits. The small number of stores included in certain market areas make product movement projections for those markets unreliable. In ad- dition, store audit organizations sell bi-monthly reports which do not reach subscribers until a month after the reporting period. These new organizations utilizing warehouse-based product move- ment data have been able to reduce this time, and furnish product movement reports one week after the close of the four -week reporting —r— ‘ 4 4See: "Data Collectors are Moving in on Food Chain Warehouses, " Business Week, (April 19, 1966) p. 57; and "Researchers Snap up Supermarket 'Secrets, '" Business Week, (March 5, 1966) p. 83, also "MRCA Gets Rich Supermarket Data, " Business Week, (April 30, 1966) p. 134. 54 period. Manufacturers introducing new products have found these services advantageous because remedial actions, if necessary, can be taken almost immediately upon discovery rather than a possible one or two months later. The continuous panel of consumers provides much infor- mation not available from retail store audits. It is for this reason that many firms in the consumer packaged goods industry also pur- chase consumer panel data to use in conjunction with audit information. Continuous Consumer Purchase Panels The J. Walter Thompson agency began a pilot consumer panel research project in England in the early 1930's, which attempted to develop a method for obtaining data on consumer buying habits by brands and quantities. This initial project led to later experimental work in the United States. There are several commercial organizations which use continuous consumer panels for gathering data similar to that obtained by retail store audits. In many cases, however, the panel technique is employed for obtaining additional information not available by retail store audits. Because of limitations of retail store audits in providing adequate information about products and their re- lation to the consumer, many firms have found it necessary to obtain the type of data desired by purchase of consumer panel data. 55 The panel technique has been the subject of much publi- cation because of the uniqueness of panel information. A few pub- lished murces accent the methodological and operational aspects of 45 panels. However, many sources have concentrated upon the panel as a source of marketing information and its use as a research instru- . . 46 . . ment in soc1al research. While there appears to be a pauc1ty of published literature on store audits, there are several sources re- lated to the uses of consumer panels as a research method. Before discussing the types of information and limitations of panels, it is first necessary to define the consumer panel which was used in this research. The consumer panel referred to is defined as:47 c—Y m 45See: G. G. Quackenbush and J. D. Shaffer, Collecting Food Purchase Data by Consumer Panel, Technical Bulletin 279 (East Lansing: Michigan State University, 1960), and Harry E. Allison, Charles J. Zwick, and Ayres Brinser, "Recruiting and Maintaining a Consumer Panel," Journal of Marketing, (April, 1958) Vol. XXII, no. 4, pp. 377-390; also Establishing} National Consumer Panel from a Probabiligy Sample, Marketing Research Report no. 40, (Washington: U. S. Government Printing Office, June, 1953) and Problems of Es- tablishiia Consumer Panel in the New York Metropolitan Area, Mar- keting Research Report No. 8, (Washington: U. S. Government Printing Office, May, 1952). 46Harper W. Boyd and Ralph L. Westfall, An Evaluation of Continuous Consumer Panels as a Source of Marketinglnformation, (Chicago: American Marketing Association, 1960) and P. F. Lazers- feld, "The Uses of Panels in Social Research, " Proceedi’ngs of the American Philosgphical Society, (Nov., 1948) pp. 405-410. 7Raymond H. Ganley and Richard D. Crisp, "Consumer Purchase Panels Serve Advertisers, Agencies, and Media, " Printer's RE (Aug. 8, 1947) p.29. 56 a group of consumer units, either families or individuals, care- fully selected and controlled to constitute a proportional and representative cross section of all consumer units in the popu- lation group being sampled, who report weekly, all food pur- chases. The consumer panel is designed to obtain a continuous record of the behavior of consumers over a period of time. Both the store audit "panel" and the continuous consumer panel obtain infor- mation about the same cooperators. Because consumer panel obser- vations are continuous, inferences do not have to be made for the periods in which no observation was made as they do when observations are made at separate points in time. Through analysis of the compo;- nent parts of the consumer panel it is possible to account for aggre- gate changes which have taken place by tracing the movement of merchandise into the hands of the final consumer. Consumer panels are comprised of participating families who maintain records where the details of all purchases made of the commodity under investigation are recorded. Although panels vary, depending upon the purpose, the typical consumer panel obtains the following information from reporting panel members: brand of grocery or beauty aid purchased type, flavor, or variety of product package size quantity purchased price paid store purchased including all types and organizations . whether special inducement (price off, in-pack premium, or coupon) was received. NO‘U‘Io-hLNNH 4ISA Guide to the Organization and Services of Market Research Corporation of America, (Chicago: Market Research Corporation of America) p. 5. 57 The above data provides a portrayal of the purchase trans- action, and is combined with classifying data of the cooperating panel member. Classifying data of panel members consists of demographic information about size of family, income, economic group, occupation of head of household, age of members of household, education, and geographic location. Most panel members report purchases weekly and are compensated by merchandise awards chosen from a catalog on the basis of points awarded for various degrees of cooperation. Uses of Consumer Panel Data The major use of most consumer panels is for obtaining a measurement of both the net and gross change in consumer pur- chases. The net change measurement is the difference in level of purchases reported by panel members between two or more periods of reporting. The gross change reflects gains or losses in units pur- chased between the periods considered by combining both new and old purchasers for the specified time. Gross changes are often used to obtain a measure of "new" versus "old" purchasers of the product among panel members. The net change is merely descriptive whereas the gross change is utilized in various ways. Repeat buyers versus non-repeat buyers may be analyzed relative to their classification characteristics to determine the loss or gain of certain demographic market segments. By analysis of those panel members having not 58 repeated purchases of a product, it is possible to determine the ex- tent of substitutions of competitive products by non-repeat panel mem- bers. By projecting such panel members purchase data, the manu- facturer can obtain an aggregate view of total sales of his brand, all other brands, non-buyers of competitive brands, and the switching that takes place between brands.‘49 The complete written record provided by the consumer panel provides information as to what was purchased, where it was purchased, when it was purchased, and by what type of household. This information is then combined with the number of households purchasing for the first time, repeat purchasers, number of times each household buys, and the amount purchased each time. Careful analysis of this data can provide the marketer with information use- ful in determining a market profile for his product. The consumer panel is unique in providing the type of information used in promo- tional decisions. By correlating customer demographic profiles and their purchases of the product (heavy, light, non-use) and media audiences, it is possible to use panel information for improving media selection. Because it is not economically feasible to cover all the types of stores which may sell particular products by use of retail store audits, consumer panel data is often used in determining the 49Boyd and Westfall, op. cit., PP. 7-8. 59 relative importance of a variety of store types. Many types of outlets selling food products may not be included in the retail store audit but consumer panel members report all purchases regardless of where they are made. This information assists the panel user in recognizing distribution trends taking place and discovering the importance of the various types of outlets relative to his product. In addition to aggregate information about market size, brand shares, package types, and seasonal, cyclical and long term trends, the consumer panel would be useful in obtaining data in studies where market dynamics are investigated. Because of the panel's sen- sitivity to changes in consumer buying patterns, it can be useful in evaluating the impact of new competition by obtaining measurements of the effect of competitive promotion, price, or change in packaging. A partial listing of the various ways in which consumer panels have been used would include the following: 1. for discovering problems associated with product testing 2. for judging radio programs 50The Chicago Tribune Consumer Panel, (Chicago: The Tribune Co., 1962) p. 34. 1Marie Sellers, "Pretesting of Products by Consumer Juries, " Journal of Marketing, (April, 1943) No. 4, part 2. 5Z'Hazel Gaudet and Daniel Cuthbert, Radio Listener Panels, (Washington: The Federal Radio Education Commission, 1941) . 60 . . . 5 3. for determining magazme content and voting behaVior . . . 5 4. for measuring the sales power of advertising 5. for measuring the degree of exposure of magazines and television and the extent to which these reach markets for various products55 6. for measurements of department store buying 7. for opinion and attitude research 8. for predicting early the market success of a product. 53Paul Lazersfeld and Marjorie Fisk, "The Panel as a New Tool for Measuring Opinions, " Public Opinion Quarterly, (Oct. , 1938) . - 54Raymond Ganley and Richard D. Crisp, "How Would You Like to Measure the Sales Power of Your Advertising," Printer's 1315, (Sept. 26, 1947) pp. 44-52. 55A Study of the Magazine Market: Its Size, Quality, and Buying, (New York: The Magazine Advertising Bureau of Magazine Publisher's Association, 1960) . 56T. W. Black, "Using the Consumer Panel to Measure Department Store Buying, " Journal of Retailiim, (December, 1948) Vol. XXIV, pp. 151-157. 57R. A. Robinson, "Uses of the Panel in Opinion and Attitude Research, " International Journal of Opinion and Attitude Re- search, (March, 1947) pp. 83-86. 58Louis A. Faust and Joseph W. Woodlock, "Early Pre- diction of Market Success for New Grocery Products, " Journal of Marketing, (Oct., 1960) and W. Barcley, "A Probability Model for Early Prediction of New Product Market Success, " Journal of Mar- keting, (January, 1963) pp. 63-68. 61 9. for a source of data for price, income, and cross- elasticity computation 9 10. for a source of determining the extent to which a brand has become recognized in a community60 Limitations of Consumer Panels Regardless of the numerous advantages and uses of con- sumer panels, there are also certain limitations present. Certain of these center about the technical nature of establishing and main- taining a panel operation, but several are inherent limitations of this type of information source. In an address to The American Marketing Association, Mr. Samuel Barton listed the following restrictions on the uses of purchase panels because of inherent limitations: 1. they cannot be used to directly analyze the motivation underlying purchase or non-purchase behavior; 2. they cannot provide a measure as to the reactions of family members to various products; 3. they cannot provide the reasons underlying selection of the particular store where purchases were made; 4. they provide no measure of the readership of advertise- ments or identification of radio and television sponsors; 59G. G. Quackenbush, "Demand Analysis fromthe MSC Con- sumer Panel, " Journal of Farm Economics, (August, 1954) Vol. XXXVI, No. 5, pp. 691-698. 0George H. Brown, "Measuring Consumers' Attitudes Toward Products, " Journal of Marketing, Vol. XIV, No. 5, (April, 1950) pp. 691-698. ' l 6 Samuel G. Barton, "What You Can and Can't Do with a Consumer Panel, " an address to the American Marketing Association, Boston, (March 15, 1949). 62 5. they are unable to provide a measurement as to the avail- ability of products for purchase in stores, no measure of the posted prices of all products is available or the amount of display space for particular products by the consumer panel; 6. they cannot provide a measurement of retail inventory levels. If the consumer panel technique was employed to secure information about the previous items, it would result in a biasing and conditioning effect on the future activities of panel members. In addition to such inherent limitations, sampling and reporting problems are present. The continuous consumer panel, like the retail store audit, is based on sampling techniques. Recruiting and maintaining a panel from a sample may affect the representatives of the information secured from the panel. The dependence upon cooperation, and the problems in obtaining this cooperation prevents the use of a proba- bility sample design. Because of difficulties in securing a high rate of cooperation by those selected for panel membership, few panels are selected entirely by probability methods. 62 The problem of non- cooperation increases the sampling problem because of the difference in cooperation rates among various types of households. Because of the necessity to keep records, there is usually an upward bias of intelligence and education among panels. This may result in unreliable Establishinga National Consumer Panel from a Proba- bility Sample, Marketing Research Report 40, (Washington: U. S. Government Printing Office, June, 1953) p. 19. 63 information for products having different rates of usage by groups with different educational levels. Over a period of time, one can expect some mortality or attrition of panel members. For various reasons panel members will drop from the panel. Those who drop from the panel may cause the remaining panel to be composed of "atypical" members. 63 Phase II of this research program investigated the differences existing between cooperators, non-cooperators, and those discontinuing panel member- ship. Because no control group is usually maintained, there is no measure of the changes that take place among another group. The problems associated with panel reporting limits the use of- information provided by consumer panels. The accuracy of information obtained may be questionable, because other members of the household make purchases other than the housewife and these may be overlooked in recording. One must consider that of those husbands who help with shopping, three out of four shop for food or related items. 64 However, only forty percent of the male members of house- holds do practically no shopping, and the possibility of failing to report purchases by all household members is present when recording is done by other than the husband. 63Shaffer and Quackenbush, op. cit. , p. 89 and Allison, Zwick, and Brinser, op. cit., p. 377. 64The Changing Consumer, Research Report of the Public Opinion Index for Industry, (Princeton: Opinion Research Corporation, 1962) p. 16. 64 In an effort to maintain their panels, commercial research organizations exercise special cautions not to obtain too much infor- mation. Even though panel members receive compensation for their participation, this is often inadequate for obtaining information other- wise desired. Theincentive itself may produce further selectivity of the panel composition. If an attempt were made to secure data on the psychological make -up of panel members it could lead to a large drop- out of panel members. It is for this reason commercial panel members only classify their panels by demographic data. Such demographic data, although valuable, is not sufficient to adequately delineate market seg- ments. In an effort to minimize the danger of panel members be- coming "experts" as a result of re-interview effects, mo st commercial panels have a forced turnover of members. The MRCA panel has a turnover of twenty percent each year. This includes those who volun- tarily discontinue membership and those who are forced to withdraw so the total of twenty percent can be obtained. This turnover rate is to reduce conditioning which may occur. Nevertheless, there is still the possibility of remaining a panel member for a period of five or more years. Panels may be criticized on the grounds that participants are habit-ridden people as evidenced by their willingness to prepare neces- sary reports over such a long period of time. The question as to the 65 atypical nature of panel members has direct bearing upon the infor- mation they may provide in new brand loyalty studies. The continuous consumer purchase panel and the retail store audit both provide useful purchase information. The choice of using one or the other, or a combination, depends upon the nature of the problem to be solved, the kind of information needed for the solution, and the efficiency-cost relationship of the alternatives. Both consumer panel and retail store audit data are primarily utilized for projecting brand shares and trends. It is unfortunate that empha- sis has been given to routine projections rather than to the special analysis which can be made because of the unique nature of each method. Store Audits and Consumer Panels in Total Market Investigations If the interaction taking place in a particular market is to be observed, it is necessary to have some knowledge of the stores serving the market, the purchasers comprising the market, and the promotion activity that is present. If such data could be effectively and efficiently collected, it may lead to a better understanding of the nature in which one particular market is segmented. Such an 65Ross M. Cunningham, "Brand Loyalty, What, Where, How Much?" Harvard Business Review, (January-February, 1956) pp. 116-128. ‘ 66 understanding may offer possible explanations why certain promotional treatments are more effective than others. The possibility that certain segments are more sensitive than others and respond differently to pro- motional efforts could be evaluated. Such explanation could lead to the determination of why shifts in partonage took place both among products and stores. Because most markets for consumer goods are dynamic and changing, it is necessary to utilize research methodology which will obtain the needed information over time. The continuous nature of store audits and consumer panels could provide such information. However, because each has certain limitations and advantages, it is necessary to employ both for a better understanding of interaction oc- curring between buyer and seller. Each is best suited for the collec- tion of certain types of the needed information, and neither utilized alone provides the necessary information. Need for Audits In a total market exploration retail store audits are needed primarily as an information source about sellers, whereas the consumer panel would provide information pertaining to the pur- chaser side of the market. It is necessary to employ some method of retail store audits to obtain some indication of the overall com- petitive atmosphere. Consumer panels cannot obtain this information but the retail store audit can provide: 67 1. a measure of the distribution of products by brands. It would be necessary to have a measure of the offering of various outlets. This willprovide alternatives available to the purchaser. The retail audit also pro- vides information on out-of-stock conditions. 2. a measure of the in-store merchandising activity and extent of special retail support. Much in-store pro- motional activity goes unreported by consumer panels. 3. a reading on competitive prices. Only those prices of items purchased are available from the panel member's diary. The price of competitive products might have been the variable accounting for this product's purchase. 4. a measure of display and shelf space allocation in addition to other merchandising activity. 5. a reading as to total store makeup and composition in the way of services offered. 6. a measure of new products available. The A. C. Nielsen Co. feels this is the major advantage of retail store audits, as opposed to consumer purchase panel data, in making product or brand projections. They feel that products change more rapidly than do consumer prefer- ences and they can provide a better measurement of the distribution lag present. 7. a measure of the dollar sales by product, department, and ' stare. Need for Consumer Panels Although retail store audits provide needed information concerning product offering and outlet, the consumer panel is neces- sary to secure the following information pertaining to the buyer side of the exchange process: 68 l. a measurement of the demographic characteristics possibly accounting for variations in purchasing and brand posi- tions; 2. a measurement of the depth of the market in terms of the proportion of people purchasing; 3. a measurement of brand and store loyalties indicated by repeat purchasing; 4. the concentration of buying by relating frequency of pur- chase with number of families; 5. the number of families reached by special deals and pro- motional offers; 6. an accounting for items purchased in 3L1 stores where the product may be sold; 7. a measure of the effectiveness of certain promotions by accounting for coupon redemptions, extra stamps, and other specials; 8. a measure of the personality configurations of panel mem- bers if a panel was established and then systematically destroyed by obtaining too much information; 9. a measure of the importance of products making up the .market-basket rather than just those items easily identified, classified, and counted; 10. a measure of life style changes on purchasing. Summar y Much of the research pertaining to the food industry has attempted to isolate a single independent variable and analyze its effect upon sales of a particular product or group of products. This type of research has provided the food industry with information useful for 69 improving operating techniques but little in the way of discovering interaction that is present in the market. Recently, there has developed an interest by the food industry and some marketing research organizations in more com- plete food marketing studies. Such research is generalized to cover a wide range of items and situations and relies on both survey and observational research. Because of the advantages and limitations of each, it is necessary to utilize both in such studies. One method of observational research is the retail store audit. Retail store audits provide an accurate and practical way to evaluate aggregate consumer choice at point of sale but are inadequate in other respects. The continuous consumer panel, as a technique of survey research, can obtain much of the information needed for total marketing studies not available by use of retail store audits. Never- theless, the consumer panel is deficient in several respects. Because each of these research instruments has certain uses, advantages, and limitations, an integration of the two methods is desirable. By combining both retail store audits and a consumer pur- chase panel in one community, it may be possible to arrive at a more complete understanding of market segmentation and marketing effec- tiveness. This research attempted to integrate and implement both methods in one particular market. This was accomplished by obtaining measures of the buying behavior of a sample of households in a small 70 isolated market by consumer panel concurrently with audits of the sales results of the universe of food stores supplying their needs. The selection of the community, development of the panel reporting diary, and training of store auditors were necessary preliminary steps to implementing the proposed methodology. CHAPTER III PREPARATIONS PRIOR TO FIELD INVESTIGATION Introduction Several steps were necessary before retail store audits could be conducted or the consumer panel could be recruited and operated. The initial selection of a community for the research was as important consideration. Certain communities were expected to lend themselves more than others to the proposed research metho- dology. The community chosen had to be large enough to include several food stores but at the same time be small enough to accomo- date financial and personnel restrictions. The development of certain criteria for selecting a community was necessary before progressing further with the research. Upon selection of the community and securing indications of cooperation from needed persons, it was then necessary to prepare for the implementation of store audits and consumer panel. Diary forms, incentive programs, and products to be audited had to be chosen prior to field investigation. Some choices that were made in 71 72 this phase of the research led to recommendations applicable to similar studies of this type. It was necessary to approach the tasks, events, and activi- ties necessary in some sequential manner. This allowed for the re- searcher to anticipate future difficulties and plan the time and costs required more effectively. The sequential stages for this portion of the research were: (1) selection of a community; (2) obtaining the cooperation necessary; (3) selecting product categories to be audited; and (4) the preparatory work necessary for operating a consumer panel. Community Selection Criteria Isolation of Community A community relatively isolated for food purchasing patterns was necessary for this research. It was necessary to locate an isolated community because: (1) there was a need to reduce the influence of alternatives available for food purchasing; and (2) to isolate the more important variables effecting food shopping behavior. For purposes of this research the term "isolation" was defined as being separated from food shopping alternatives because of spatial and temperal considerations resulting in the impracticality of making palnned trips for food purchases. In addition to time and distance, other considerations would influence the selection of the city. Among these factors were product offering and store size, but because of the exploratory nature of this 73 study, only traveltime and distance to other communities was con- sidered. LaLonde, in studying supermarket drawing power, found small town supermarkets drew ninety percent of their customers from . 2 . . . . 1. 38 miles. If a shopping district of the community selected were to approximate a regional shopping center, this distance increased to . 3 . . . . . 2. 53 miles. These yardsticks were initially employed for selecting the isolated community used in this study. One study found, using travel time rather than distance, that consumers will spend only about twelve to fifteen minutes traveling 4 to reach a shopping area. The regional shopping center distance, or —v 1 William Applebaum and Saul B. Cohen, "The Dynamics of Store Trading Areas and Market Equilibrium, " The Annals of the American Association of Geflraphers, Vol. 51, (March, 1961) p. 3. 2'Bernard J. LaLonde, Differentials in Supermarket Draw- i_ngPower and per Capita Sales by Store Complex and Store Size, Research Report in Mass Marketing Management No. 2, (East Lansing: Michigan State University, 1961), p. 90. This study investigates store size and products relation to distance traveled, and defined a small town store as a supermarket located ina town of less than 5, 000 popu- lation. 3_I_1_)_i_c_l., p. 21. A regional shopping center was defined as having in excess of fifty separate retail units dominated by a complete, full size department store. Although this investigation was to take place in one community rather than a shopping center, it was felt that these guidelines provided an adequate initial criteria to delineate com- munities for investigation. 4Victor Gruen and Larry Smith, Shopping Towns, U.S.A. , (New York: Reinhold Publishing, 1960), p. 33. 74 time for travel, whichever greater, was used as a minimum require- ment for selecting the community. Several communities within a radius of one -hundred miles of East Lansing, Michigan, were screened. Each community was additionally screened for population characteris- tics, cross-section of food outlets, employment base, and media availability. Foodtown was chosen as the community best meeting these considerations . Population Characteristics of Foodtown A community approximating national and state population characteristics was desired. Although the research was not intended to be representative of other communities, some similarity was de- sirable to enable other studies to be based on the data. The population factors analyzed were median family income, percentage above and below median income, average size of family, and education. United States Bureau of Census data for 1960 indicated an approximate population of 7, 000 for Foodtown proper and 12, 000 for both city and suburban area. This 12, 000 population figure represents Anonymity was assured for all persons and firms cooper- ating in the research. In order to preserve the confidential nature of some information, e. g. weekly dollar sales of cooperating retailers, product movement, and market share information, it was necessary to disguise the name of the community selected. If this had not been done, individual cooperating stores could easily be identified because of the nature of the retail food marketing structure present in this community. 75 slightly over twenty-five percent of the total population of the country in which Foodtown is located. Although not representative in some respects, Foodtown was chosen because it more nearly approximated these factors than other communities considered. Table 1. --Family income for Foodtown, State of Michigan, and United States (in dollars and percents). Income United States Michigan Foodtown Median Family Income $5, 625 $6, 256 $5, 681 Percent Under $3, 000 21. 7 15. 7 19. 8 Percent Over $10, 000 14. 3 17. 4 13.1 Source: United States Bureau of Census, 1960. Table 2. --Median Years of Education for Foodtown, State of Michigan and United States. Education United States Michigan Foodtown Median Education in Years 10. 3 10. 8 12. 0 Source: United States Bureau of Census, 1960. The increase in population between the 1950 and 1960 U. S. Census indicated that Foodtown did not increase in population at the same rate as did the county in which it is located or the State of Michi- gan. The population for the city of Foodtown increased 5. 3% during the 76 1950-1960 period. This smaller population growth may be accounted for by the addition of several new urban areas adjacent to the city but not included within the city limits. These new areas are included in county figures for census purposes. Cross Section of Food Outlets A project of this scope required that there be a variation in the merchandising practices of the stores included in the community. These differences allowed various changes in promotion to be moni- tored throughout the length of the study. Foodtown had the necessary differences, because the several stores were serviced by different warehouse groups. Ideally, one would select a community with several outlets available and adjust the pormotional variables of one store holding the promotional activity of the remainder constant. Because of the com- plexity of variables affecting sales beyond the marketing experimenters control, such scientific experimentation is impossible. In addition, the problem of extensive cooperation to the point of adjusting merchandising variables would most likely not be available from retailers serving the community. m For a description of different promotional strategies em- ployed by various food chains see: "The Great Supermarket Profit Squeeze: Relief in 1964"? Forbes, (Feb. 15, 1964), pp. 20-24. 77 By observing previous weekly sales plans provided by chain warehouse groups servicing outlets in this community, it was evident that varying merchandise approaches could be expected. 7 In addition, newspapers for exactly the same period, one year perviously, were examined to obtain some indication of the promotional policies of stores not following warehouse sales plans. The available newspaper advertising indicated different approaches had been utilized during the previous year. Fifteen retail outlets provided the grocery needs for Food- town at the beginning of the study. One outlet discontinued business during the course of the research. These fifteen outlets were serviced by two national chain warehouses, one voluntary--cooperative ware- house group, and one regional chain warehouse. Several of the smaller The weekly sales plan provides warehouse member stores with the merchandising and promotional plan to be followed for a speci- fied period. Typical weekly sales plans contain information regarding all promotional treatments for the time period. In addition, future pro- motional activities are outlined to allow member stores adequate prep- aration for these events. Additional information contained may include equipment for sale, availability of special bonus or priced merchandise, and certain management information regarding store operation. Weekly sales plans obtained indicated certain chains could be expected to use different merchandising approaches than those of other chains. For example, one chain's weekly sales plan indicated a heavy emphasis would be placed upon the continuation of bonus trading stamps with ap- propriate newspaper copy emphasizing bonus stamps. However, another warehouse group's weekly sales plan indicated that newspaper and in- store displays would revolve about a policy of low prices with no trad- ing stamps. 78 outlets were serviced by a combination of three grocery wholesalers. It was observed that these wholesalers provided no merchandising plans or assistance to these customers. A profile of the retail grocery outlets and merchandising approach is seen in Exhibit A. Employment Base The employment base of the city chosen had be to large enough to preclude large portions of the population commuting to other communities for employment. A relatively self-contained community enabled the observer to monitor changes in promotion bearing upon those panel members making weekly purchases. If many of these purchases had been made in communities other than the one under observation, no measurement of what had occurred in promotion in these outside food outlets would have been available. Likewise, the size of the city chosen was not to be such that it would draw a large number of persons commuting for employment from other nearby areas. If a large amount of migration into the community were pres.- ent, there would be no representation by these individuals on the con- sumer panel. By selecting a community where the daily migration to and from the community for employment was minimized, grocery pur- chases made by those persons residing outside the community were expected to be minimized. Also, the possibility of food purchases being made by panel members in other communities where they were employed was reduced. 79 .wcnmfiuoxepm Hmoofi oz .mxfiup “mom was noon. mo £593. 133 on. poops: .oom .wHw mo modem 3.334 $980 new .mvdfiup flow down who named. 08.30.» gwwm .mumoe coosocsauomeDMm was Swamp .mowuooonw than mxooum Smog gmonm no oodpoum oz .ooud mcwdom mo .um 933mm? ououmug. .mmgfim mfipduu OZ .Um mac .uoummmon 28 so“? :mom was 802: m .ooo.¢oza mo modem 3.384 .233 can .863 603 one mofimm mo uncommon owned 4 .mpOOH Gonouw mo on: nouns: m moauumO .ommo soapofim .wcwxnmm woouummmo on mm: was magnum ooguomtfiom Edam was mono urea out/now mags: 63m ”—0: mooQ .wfiflnpcmg Ho 4 .mmhm maiden Ho .um .Um oomfi 38.254 .mqfimfiuocrpm H803 .GOSOEOHQ vacuole. OZ .uoumdwou moo 333 shown ewuououw gougoQ N .000 Jma mo organ Hrs-Em £02.? .um .Wm omN 33.254 .Goococdfi lemonade no .Gofioom demon...” .oudpofim don—059a ououmtfi no madman Smog Amman oz .mmgsm oGSOmom not? .0939 1.83;. H.603 on 33 magnum wagon» mon tgoo pom non—3mm.“ one new? :mom was 802: A no 85 naufiom wfimwpcmfiuuog Goaumwuomofl smegma/H .Evoonoh mat/How 30350 poem Smash mo ofimounmnu .< sofixm 0 8 483385 Hmcofloaoum was wearing ououmug. womb .wfimfiuoxrpm Hmoofi no magnum wagon» OZ .oco .oofiw oofimm Huang .ommo poom Genoa“ was .Gofiumm Goonogfitommmdmm .oodpofim Amman new? .305 smoum ma ogsfio> moHMm wooxoom $3.234 .mfiuommo poififi a??? on: 2.9m .um .wm OONH pad mnoumwmou 03» £33 uoxumEquSm Spam o .86 - 8x: 9:856 8.5 .25 .38 mo nozon Hosanna poor, loam Goa mofimm thooB mBofioh .Emao o>3m~omooo P332519, mo non-H82 .wfispdup >336? oodmtowm .mfi .mgm >385 paw aims: .mowuooouw 13.96 p08 so: HOW :oawmm: >me03 co promo. Pep .oosmoonm pad Home amonm 5:? on: wcfimfiuozfiw .coSoEoum ougmtfi .mfimfi 25h .uw .Um ooo.~ 38.2544 .muoucdoo upo>pm homommspoc Hmoofi .mmEmum mumps: mo>1U ”393vo 0&5 aux? uoxumEHoQSm Edam m .uoucdoo Hora $3.32ch ooo .omo 4m >Houm8wxoummm mo um. Gowucofim poNZNGOmuom ouoe guts moodum mofimm .mdoum ®>SMHoQOOUt>umEdHo> gunman twigwfim mo enofiom m. 3020M o» muoommafi m Ho “@3802 .363 fwusmon paw Simon .coEoEoum upon—mus: woman—mom was whommm .mofaououw ewpp .oodpofim .368 m0 03m [wk/on >326 anon. 5. ”nominee/pane .m>.mp madam 15.35. who? oaonEoo cwddm 44. .1..on @3506 mafia you mcomdoo can >306? >36 wESom mo .uw .qm come pom muoucsoo madam 3350p 6:0 #8.? magnum wasps: won/10 udoxooco p.98 £33 uoxumguomdm w. HQQEDZ twofionm mcwmdpdmgouoz Goaumfiuomofl upon—m 335.800 T- .< ”.3..an .l. 8 .33an8: no mammogram CZ .025.“ .303 :0 undo: onoum >18 oomwuuo>p4 .mtfwamflo 12.0on no 2030595 ongoing. OZ .aznaooo ooh .th no nozom .omoonu pad £25 £505 Gorgocdfi newsmsom £33 poxogm uofiooo pomofio .6 93.6300 Smog no ouspoum Amoum OZ .mEofi #088 one woman .mowmuo>on Eouw mmofimdo. mfi mo “mock moon page dowumuoao mo on?» concave/Goo 4 .moum wfiflom mo .uw .wm oom .xoummm was Houasoo uzoxoogo moo €53 :monm was 832: m .25 .25 .Nw mo mofimm 3.384 .mpoom amOm mo 93.830 @958: wfifiofipm no mom .muoucsoo udoxoogo Ema». .Gofioom pooch .GofioEoum oncoming. pom mdwmfiuozom cooaocsfitommmsmuw smug $3.3 Somehow uoammmapoo Hmoofi 5a.? mono mcwxnma woouum «dog powmxosmtonm .pm .Um ooo .2 .30 owned .moowum BofitummEoum on mo enoZonH .xoummm €33 «omeEHoQSm on: fish w .25 .ooo 4% mofimm 3384 .mp3 lagoon was £3603 paw .mommu noumnmou pom ommpcmaouoa mdfifim .mowuooouw F6 .oosmooum Jame: gmoum mo >USOQ m. 5:? nommmmsvma 5. mammfiuonrom mo on: muonEoO .fi .wm ooo .m new? 3004 .mtnmamfio u :ofiogoum 6.33-5 Sop muoucdoo usoxooao .Soh .Emno 3:038: madam 6350p 9.5 Hz? magnum mumps: mon mo nongogtuuoxumauomsm on: fish N. knoflom mcwmwpcmcouoz cofifiuomofl uoagz onoupm AposaaocoUT- .4 tnExm 82 .mmgdum wagon» no don—050.3 0.83th .mfimfinon/pm H.600." oZ .ooo .wmm moHMm 12:34 .0539, amen 05.3 was noom .693 omensamm mo urea soot pom—ME: paw .oodpona noon“ oz .poxoofiw mgofi >uoooum pofinfid .um .Um omm was heumwmou 6:0 #33 :mom was 832: N~ .mfidapgfi no 953.39, 1pm. .mmgmum wasps: .mtwmamwp 0.33.15 02 .oow ova .xoummd modem 3354 .368 smonw oz can .oodpoum gmoum mo unsound Seam m .mgofi twuoooum mo wcwuommo coupe: 4 .oommm madden mo .uw .wm coo .xoummm paw uoacdoo usoxoono 6:0 as? :monH was 502: fifi .magmum mazes: no madnpcm: .wcnmfiuozum Hmoofi o: .Gofiogonm 0.83:5 oZ .moHMm 05.3 was .303 932533 ooo .mww mo modem finance H33 fin? our: 053 was .803 smug 4 .mvwnomouwo podponm sandman/m 5 Memo smug M £53 mEofi >uoooum ooqoddoxaoo >Humoe mxoouam .oodponm can noucdou «morn not: o. has noon 93:3 no .3 .8 con £33m?" oco 2.3M? :aom was EOE: A: twofiom mnwmdpcmgououz con—gnomofl nonEdZ macaw Aooacaopoo T- .< 2.“:an 83 .aoSoEoum whom—mtg. no .mmgmum magmas. .wfimfiuocrpm 133 oz .ooe.wvw moHMm 33.84 .wawnvwwo >uoooum twat pofiflhd .cofioom mpoom Gouofi oz .oodpoum soon“ on one among smoky US$54 .um .Um oom $3.234 .uouqsoo «doxooso 0:0 5:3 :aom can 802: mg .mewmdmmnp 3092.3 omumfi >6. mfimwuoam moudummh .mgmfiuozom Hoood .GoSoEoum paw mtfmamzo ongoing 680m meg .mmgmum mazes: mop/10 .25 62% modem H.954 .Gofioom £608 cononm Edam .oodpoum pad «dog good nuon mo mawuommo coupes: .mnoucdoo udoxomEo 03» was .um .vm oow~ a??? geogop 663004 .593 Hmcofimc 8 mo noQEoE a: .mmEmum wagon» Ho .GoSoEoum vacuole: .mcwmfiuosbm 303 oz .ooo.mma mofidm H.354 .Gowuomm coosocdfi tommmsmm pas mpog conouw £33 .uw .vm oo: .GOmmom E panama 03350 0056 tonm omumfi £33 mcwuomo oocowcofiuoo £33.on moo gan.? :mom was 802: Ma n» 030nm m5. mdpcogouog cofimtomofl non—852 ououam Aooaozooo T- .4 rain... 84 The possibility of selecting a community fulfilling all of these requirements in their entirity was next to impossible and had to be relaxed. Foodtown was chosen as a community which closely met the above employment restrictions. Some indication of this small employment migration was observed from panel members weekly diaries. Only six persons on the panel reported food pur- chases made outside of Foodtown during the ten-week period. These outside purchases were made in the city where the head of the family was employed. However, there was no measure of the extent of pur- chases made in Foodtown by persons outside the community at this time. Twenty industries, varying in size from a small tool and die shop employing eight persons to a large assembly maker for the automotive industry employing over five-hundred persons, were located in Foodtown. Furniture, athletic wear, and foodstuffs were but a few of the wide range of products produced. Also present were several state institutions adding considerably to the employment base of the community. This presence of a fairly large number of state employees may have accounted for the high median educational level of this town because of state civil service educational requirements. With an average unemployment of less than three percent and the local availability of employment for both male and female, it was felt that Foodtown provided an adequate employment base for 85 purposes of this study. Although no data was available for migration into and out of the community for employment and possible subsequent grocery purchasing, conversations with Chamber of Commerce officials and the largest employer's personnel manager indicated this would not be a significant number. Media Availability Because of the extensive use of newspapers by retail food stores for food promotion, it was necessary to have a minimum of one daily newspaper serving the community. 8 Ideally this was to be a local publication rather than one of a nearby metropolitan community. This would confine certain promotional variables to the particular com- munity used in the research. It would have been difficult to monitor the outside promotional material influencing purchasing in the city being studied. Several communities initially considered as having local media available were inadequate in other respects. Foodtown had two newspapers published locally. One of these was a daily having an ap- proximate circulation of 4300 during the period of the research, while For a description of the various uses of newspaper ad- vertising by retail grocers and the extent of such advertising see: Colin B. Church, "How is Newspaper Advertising Used as a Competi- tive Tool by Chain Grocers in the Philadelphia Area? " Unpublished Masters Thesis, Wharton School, University of Pennsylvania, 1962. 86 the other publication was a weekly having a circulation approximating 3700. The majority of grocery store advertising was placed in the daily newspaper. This paper featured a Wednesday food section but of a limited nature. In addition to local news, news of other com- munities in the county was featured. Such a format gave the newspaper the appearance of a county rather than city publication. There were six weekly newspapers published throughout the county with a combined circulation of 12, 200. The weekly Food- town newspaper circulation comprised about one -third of this total, but grocery advertising was limited. In the past, the publisher of this paper printed enclosures with the paper for several local grocery outlets, but this did not occur during this investigation. Other community media was limited to one 5000-watt local radio station but radio and television were available from the two nearby metropolitan areas. Television was received on six different channels, but two of these channels had marginal reception. 2 Because of the proximity of these metropolitan areas some measure of the newspapers which were read was desired. It was felt that one daily metropolitan paper featuring advertisements of the same chains with outlets in Foodtown would recieve rather wide circulation in this community. If this were the case, some monitoring of this paper was necessary. A questionnaire was sent to over 700 Foodtown 87 householders. Those responding indicated that this was not an im- portant factor. Only ten percent of those responding indicated they read this paper, and panel members indicated that only twenty-nine percent read a newspaper other than one of the two local newspapers. Although no precise measure was obtained, interviews and notations on returned questionnaires indicated that outside newspaper reader- ship was generally confined to Sunday editions . Obtaining Cooperation After choosing the community most closely meeting the requirements for isolation, the next problem was to obtain the cooper- ation necessary from all sources. A ranking order was established to minimize the possibility of not being able to secure cooperation from some strategic member of respondents, stores, or media. The fol- lowing ranking was used to insure that the necessary services would be contacted in the following order of importance: (1) warehouse groups; (2) community influences and information centers; (3) inde- pendent retailers; and (4) panel. Warehouse Groups Food outlets in the community were ranked by estimated 9 . sales. Sales estimates were based upon the number of checkout counters, product categories stocked, and services offered. The These estimates were made by two experienced students 88 purpose of this initial ranking was to determine a priority in seeking stores' cooperation for audits. If a lack of cooperation was not forth- coming from one or more units, comprising a significant portion of total sales, it would be required that the selected community be eli- minated from consideration. The three stores with the largest sales volume were mem- bers of various national chain and voluntary cooperative warehouse groups. Through the efforts of Dr. Edward Brand appointments were made to interview the grocery merchandisers for each ware- house group. 10 A presentation of the research project was made at the interview. Several important points were made in the interviews including outlining the background of the study, problem areas to be explored, the methodology to be employed, and possible results. Store managers were requested to allow the researchers to collect or provide to them the following data: store audits, weekly sales plans, and weekly dollar sales. The confidentiality of information in the Michigan State University food marketing program. Each of these students had several years experience in the capacity of retail grocery store managers, and on the basis of this experience made weekly sales estimates for each unit. These appointments were made through the office of Dr. Edward Brand, Assistant Dean of the College of Business Ad- ministration, Michigan State University. Dr. Brand's prior tenure as Director of the Food Marketing Program and reputation with mem- bers of the food industry assisted greatly in making these appointments. It was questionable if the researcher could have secured the cooper- ation needed without Dr. Brand's efforts. 89 was emphasized and it was pointed out that the results would be dis- guised. In exchange for such cooperation each group was to receive a profile of their Foodtown member store's customer. In addition to the profile, each store was to receive a summary of what customers and non-customers liked or disliked about the Food town store. Sug- gestions for product groups to be audited were solicited. The three warehouse groups approached were receptive to the proposal and each sent letters authorizing store audits to be taken to store managers. Community Influences and Information Centers Preliminary contacts were made with various community organizations before local independent store operators were contacted. The local Chamber of Commerce, Better Business Bureau, and the Retail Merchants Association were appraised of the research. This preceeded any local contacts with potential panel members or grocery store operators. Such a procedure was to allow the researcher to establish credibility with these organizations and enhance the possi- bility of obtaining further community contacts. Personal calls were made to the three previous mentioned community organizations. Proper identification was shown and the research project explained. Any assistance these organizations could offer was solicited at this time. Assurance was given that all field workers would carry proper identification and a letter of introduction 90 from the project director. These agencies' knowledge of the research aided the field investigators considerably, when on several occasions community members and store operators checked with these agencies on identity of investigators. This contact was also used to solicit demographic background information about the community. Local maps, industrial studies, and pamphlets describing the community were obtained from these sources. The tWO local newspapers serving the community were contacted at this time. An explanation of the research project was made to the news editor and cooperation was sought for later insertion of a news item concerning the project. Initial contact with newspapers was made early enough to insure that such an item would appear. If the paper would not allow a news item, then a paid advertisement was prepared for placement at the proper time. In all cases, local community organizations and persons contacted in Foodtown were receptive to the project. Throughout all conversations regarding the project, the dominant theme followed was one of asking for assistance in the university's efforts in studying food marketing. Because of familiarity with the university such an appeal appeared effective. Without this sponsorship, different re- sults may have occurred. 91 Independent Retail er s The identity of the research team was established with recognized information centers and contacts were made with unaf- filiated retailers. Personal calls were made to the remaining out- lets by ex-grocery products salesmen. Because of previous experi- enceiwith grocery store operators, it was felt these interviewers could handle all possible objections and questions. Each owner-operator was contacted during the morning hours Tuesday through Thursday. Such a time was found to be ef- fective in securing the necessary time and attention required to fully explain the project. In explaining the project, emphasis upon the pos- sible advantages to the owner-operator were explained. Each owner or operator was told that the completion of the project required the cooperation from the majority of stores in the community, and those previously contacted had agreed to participate. It was explained that in return for such cooperation a customer profile would be provided and suggestions for utilizing the data would be given. It was empha- sized that the value of such profile would depend solely upon its utilization. Each retailer was asked for permission to conduct weekly audits, provide weekly sales data, and information as to any pro- motional activity utilized. Confidentiality of data was to be honored and a minimum amount of time required on the operator's part was stressed. 92 Several unanticipated objections were encountered. The primary objection appeared to be a suspicion on the part of the smaller retailer that such a study was being conducted at the insti- gation of large chains to achieve further penetration in the local com- munity. Failure to secure the participation of one retailer resulted from the failure to obviate this belief. It was found advisable in approaching owners or operators of the smaller units to present the research project at a much more basic level than that presented to the warehouse group grocery mer- chandisers. Even in following this practice, several objections cen- tered around the failure to fully understand the purpose and approach to the research. In several cases the project was outlined repeatedly but without adequate comprehension on the part of the operator. Several agreed to lend assistance even though failing to adequately understand what was required. All except fourof the fifteen outlets contacted agreed to allow audits to be taken weekly. Although weekly sales data was de- sired, several stores agreed to cooperate only on the condition that this data would not be provided. Those stores not agreeing to furnish sales data were estimated to account for a relatively minor proportion of total food sales in the community. No reasons were given as to why sales data was not to be provided. 93 Non-cooperators indicated various reasons for refusing permission to allow audits. One store initially agreeing to allow field workers to audit sales reversed this position when the field worker arrived to take the initial inventory. This particular store operator indicated that he was in the process of selling his store, and the owner did not desire his sales data to be made public. Attempts were made to assure him of the confidentiality of such data but to no avail. This store did change ownership during the period of the research. Store Number Twelve was operated by an elderly lady of foreign birth and because of a communidation barrier the research was never adequately explained to her. Her son was contacted in an attempt to secure audit information from this store with similar re- sults. The owner of store Number Fifteen indicated that he felt it was "too much bother to have people coming and going in his stock- room. " This particular owner did agree to furnish sales data for the period of the research even though not allowing audits to be taken. Store Fourteen was a member store of a national chain which has historically followed a policy of not participating in such projects. Attempts to secure permission for audits were made through national headquarters but without satisfactory results. How- ever, this attempt did result in the securing of average weekly sales for this store. 94 Even though failing to secure the cooperation from the universe of food stores serving the community, Table 3 indicates that the number of stores cooperating was sufficient for the study. Table 3. --Yearly sales of cooperating and non-cooperating food outlets serving Foodtown. Percent of Y 1 1 ear V 3a es Total Sales Cooperators $4, 760, 800 94. 3% Non-Cooperators 287, 700 5. 7 Total $5, 048, 500 100. 0% Source: Actual sales data from eleven cooperators and estimates for non-cooperators. Although the percentage of total sales of non-cooperators would at first appear to be understated, because of the number of these stores (4) in relation to total universe (15), this is not the case. The inclusion of a national chain, not cooperating for audits, has little bearing upon this total. This particular store was located in the cen- ter of the city with no off-street parking facilities and a limited mer- chandise assortment. National headquarters personnel expressed their displeasure with sales of this outlet and indicated studies were presently being made for a decision to either relocate or discontinue this outlet. The other three non-participating stores were of the small "Mom and Pop" variety with poor locations and extremely 95 limited merchandise. Fresh meat and produce was stocked by only one of these stores and later diary entries by panel members indicated the relationship of these stores to the total sales of all stores was as represented in the above table. Problems Encountered and Recommendations Community Selection Selection of the community in which the research was to be conducted was restricted by travel limitations for field workers. No hired field staff was available because of limited funds allocated for this project. This required that the community selected be located within a driving distance allowing the field worker to complete his audit and return in the same day. Such a restriction resulted in cer- tain communities not being considered. The arbitrary selection of the initial communities considered would necessarily induce a certain amount of bias. In most cases, these communities were looked at superficially by the researcher because of time and financial con- straints. Only those cities appearing on the surface to be isolated called for further investigation. In addition, because of the close proximity of the community and familiarity with the work of the uni- versity, cooperation was found available which may have been un- available in more distant locations. 96 The size of the community chosen would result in more personal influence affecting shopping behavior than in a larger com- munity. Some measure of the degree of personal influence was ob- tained in Phase II of the research program. Word-of—mouth behavior patterns for food shopping may be present to a larger extent in a com- munity of this size than in metropolitan areas. A more comprehensive method of delineating the Retail Trading Area or extent of isolation for the community selected would be necessary. This would allow for establishment of a panel that in- cludes members located throughout the trading area rather than just those living within the city limits and nearby urban and rural area. Several various methods may be utilized for establishing these boun- daries. These methods may follow either an empirical or a gravi- tational approach and the specific method used would depend upon the objectives of the study, financial support, and time constraints. Cross Section of Food Outlets One of the objectives of the study was to obtain data useful in the analysis of consumers' reactions to changes in promotional treatments. Those food outlets servicing Foodtown's needs could be divided into three groups by order of promotional importance as follows: 11 A description of various methods and their advantages and disadvantages is found in Lalonde, o . cit., pp. 150-170. 97 1. Large supermarkets which account for both the majority of sales and promotion. (Stores 4, 7, 8) 2. Small supermarkets which do limited promotion and next in amount of sales.(Stores 5, 6, l4) 3. "Mom and Pops" which account for a small amount of total sales and follow an extremely limited policy of promotion. (Stores 1, 2, 3, 9, 10, ll, 12, 13, 15) Diaries indicated that customers changed stores for food purchasing, but not to the extent anticipated. Limited store switching may be attributed to the relatively small amount of promotion by several outlets. It may be desirable in future studies of this type to select either a smaller or larger community for observation. If a smaller community were selected, it should be one with a limited number of stores, but one in which intense promotional rivalry takes place. A community containing three or more stores similar in size and offering and exhibiting extensive promotional rivalry may provide a more complete measure of store switching by customers. In a larger community the amount of rivalry present is exhibited, to a large ex- tent, in the weekly food section of the newspapers serving that com- munity. Observation would lead one to conclude that promotional rivalry among grocery outlets would be much more intense than that present in this community. 98 Obtaining Cooperation Cooperation was not obtained from the majority of persons from whom it was sought. Several sources necessary for auditing the universe of stores were not willing to lend cooperation to this research. Several recommendations applying to obtaining cooperation can be made from the experience gained in this research. These include: 1. Letters introducing the project should be sent to all sources involved in the research other than panel mem- bers before actual contact is made with those persons. Such a letter should explain the background and scope of the study and prepare the reader for later personal contact. This would provide the field worker with an ent're'. Such introductory letters should be sent to com- munity information sources, independent retailers, and warehouse groups with follow-up telephone calls made to establish appointments. Written research outlines should be given to all desired participants in the research except panel members. Out- lines should stress the cooperative aspects of such a venture. Such an outline would serve two functions: (1) it would provide answers to questions concerning the research that cooperators have; and (2) it may develop questions from cooperators concerning points that the researcher has overlooked. Such an outline would have to be limited to reduce the bias which it may introduce at the retail level. Retailers may not follow their normal merchandising pattern if they are aware of all variables investigated. An offer of a limited payment to those retailers otherwise not cooperating in allowing store audits to be taken may be necessary. Remuneration for audits has been found to be successful by A. C. Nielsen Company in securing co- operation for retail audits. Conversations with non- cooperators at the end of the research indicated that two more stores could have been secured if a payment had originally been extended. If a payment is made for audits 99 it has to be made to all cooperators not just those initially failing to participate. Such a plan of compensation would be based upon size and importance of the establishment audited. Product Category Selection for Store Audits The selection of product categories to be audited was ac- complished in two steps: (1) a initial selection of categories to be observed; and (2) a final selection after interviews with cooperating retailers. This two stage procedure enabled the researcher to ini- tially provide warehouse groups and independent stores with an indi- cation of the scope and extent of the retail audits. Final selection of audited categories was not made until after retailers were interviewed, and the initial selection was revised because of suggestions made by cooperators. Several store managers displayed an interest in specific product categories during interviews, and some of these were included in the final product category list. The initial selection of product categories was based upon the following factors: 1 . Auditability 2. Broad Distribution 3. Vitality of Movement 4. Promotional Activity Auditability Only those products which would provide a minimum of difficulty in auditing were selected, because field reports were to be 100 collected by part time personnel with a minimum of training. Cor- respondence and visitations with A. C. Nielsen Company personnel aided considerably in eliminating products presenting special prob- lems in auditing. Members of the Nielsen staff stressed that the primary problems facing inexperienced auditors could be expected in "classifying" products and in obtaining all invoices. Several pro- duct categories were suggested as groups which would be more easily audited. Perishables were not to be included because of the diffi- culty in obtaining actual consumer sales. Produce, bread and dairy products have either considerable spoilage or merchandise returned to suppliers for credit. These categories were not considered because of lack of adequate return and spoilage records. If audited, spoilage and returns may have been reflected in product movement because of the lack of these records. In addition, sources of supply for some product categoriesmay be from producers, from a number of wholesale drop shippers, or from warehouse groups. Securing the necessary weekly invoices for cash purchase becomes complicated and in some cases impossible. Similar invoice problems occur in bread where returns are credited as stale merchandise is replaced. For purposes of this study only dry groceries were to be audited. Several categories were chosen for their relative ease in auditing. These categories were initially selected using only ease of 101 auditing as the principle consideration. This initial selection was then evaluated using other factors and reduced to nine product cate- gories. Broad Distribution Since the majority of stores providing the communities food needs were to be audited, a broad distribution of products audited was desired. The lack of frozen food cases, fresh meat, and fresh produce precluded the selection of products in these categories. Frozen foods were also eliminated because of the difficulty in obtain- ing some measurement of facings or exposures to the shopper. Observation of several frozen food sections provided an indication of the difficulty in obtaining such data. In many cases a disarray of items was observed and difficulty in obtaining accurate field results was forecast. Ready-to-eat cereals were initially considered but a trial audit found this section to be too complex and time consuming within the restrictions present. The dry cereal section illustrates the need to restrict product categories to a workable number of brands and sizes. The largest supermarket contained approximately 120 lineal feet of facings for ready-to-eat cereals where the smallest indepen- dent store carried only three or four of the most popular items. If such a category was selected it would be necessary to conduct audits 102 on the largest section with only three or four items being used for com- parisons between all stores. Vitality of Movement It was not anticipated that during the research product categories would be replaced by substitutes. However, this would be a consideration in research over longer periods. Nevertheless, prod-F uct categories were evaluated on the basis of their vitality. Relatively stable categories were desired where a continuation of existing brands could be expected, and the addition of new brands would be limited. Ideally, one would desire the introduction of several new products so that questions concerning the effect of product introductions and dis- continuances upon shopping behavior could be investigated. However, inexperienced auditors dictated that categories be chosen where these would be minimal. In spite of cautioning field workers to include any new brand introduced in the product category during the period under observation, this was overlooked in two cases. The stability of the brands included in the product category also reduced the number of old products stocked for the first time by the store during the research. This aided the field worker in obtaining more accurate observations by minimizing first- time stocked items . 103 Promotional Activity This research attempted to obtain information useful for later evaluating promotional effectiveness. Measures of promotional effectiveness would require that items audited receive promotional treatments during the research period. Items receiving price specials, deals, cents-off labeling, advertised specials, and in-store merchan- dising treatments would be included. In prior studies, products re- ceiving price or promotional changes have often been discarded, due to the researcher's inability to account for variation in sales resulting from these uncontrolled variables. 12 Unlike attempts to control the experiment to measure the effect of one variable, this study attempted to secure some measure of as many observable variables affecting sales as possible. Therefore, product categories where a variation in promotion could be expected were desired. Categories Selected Nine product categories were chosen to be audited. These categories most closely met the criteria for inclusion. The following product categories were easily auditable, represented broadly, and could be expected to receive some promotion treatment over a ten week period: 1 2Coxe, op. cit., p. 64. 104 Canned Peaches Canned Peas Catsup Canned Tuna Fish Aero-Spray Starches Instant Coffee Wax and Plastic Wraps Cooking Oils Dry Dog Food \OmNO‘mtliWNH Two categories included deserve special mention. Canned tuna was chosen because a portion of the research was to be conducted during the Lenten Season. This category was expected to receive heavy promotion treatments during this time by various stories. The dog food category was included at the request of a large supermarket cooperator who suggested this item to be one heavily promoted in this community. Recommendations for Selecting Product Categories Information about the number of different brands and sizes for each product category was necessary before determining the num- ber of categories to be audited. Cooperating warehouse groups made order forms and price lists available for this purpose. Even after analyzing these, other sizes and brands were discovered during the preliminary audit. It would be advisable to visit larger retailers to secure the number of items the audited categories include. Such items as dry dog food would have been reconsidered upon discovering that one supermarket carries over thirty-five individual brands and sizes. 105 It may be feasible to select product categories based upon the largest distribution by one outlet. This would provide the maximum number of items to be audited in any one outlet, and give some indication of the time required per audited store. Diary Selection and Development Substantial literature is available relating to the type of diary kept by panel members. 13 Articles containing descriptive or illustrative material pertaining to diary types and formats were re- viewed, and samples of diaries used by panel operators were obtained. Diaries were received for investigation from MRCA panel, Chicago Tribune, National Family Opinion, NIARPLAN division of McCann- Ericksen, University of California, and Michigan State University. The content and format of these diaries was carefully re- viewed and the decision was made to use the diary developed for a previous Michigan State University Consumer Food Purchase Panel conducted by Shaffer in Lansing, Michigan. This diary was chosen 13Seymour Sudman, "On the Accuracy of Recording of Consumer Panels, " unpublished Ph. D. Thesis, University of Chicago, 1962 and Seymour Sudman, "On the Accuracy of Recording of Con- sumer Panels." I and II, Journal or Marketirgg Research, May 1964, pp. 14-20 and August, 1964, pp. 69-83. Also, James D. Shaffer, "Methodological Bases for the Operation of a Consumer Purchase Panel, " unpublished Ph.D. Thesis, Michigan State College, 1952, and Franklin Cawl, "The Continuing Panel Technique, " Journal of Marketing, (August, 1943) pp. 45-50. ' 106 for several reasons. The M.S.U. diary was a ledger type diary. The ledger format for consumer diaries consists of fairly complete break- downs of product categories and items and purchases are recorded in the proper pre-coded section of the diary by the panel cooperator. If no items are purchased for the category, the panel cooperator is re- quired to indicate this by placing a check in a no purchase made space. The journal diary allows the panel cooperator to record purchases in any sequence she chooses within very broad categories. The ledger diary may assist in recall for recording by panel members. The journal type diary leads to further expenses in extracting and proces- sing data as compared to a pre-coded ledger diary. The M.S.U. diary was developed after experimentation with five diary formats. The M. S. U. diary was chosen because the type of diary used influences survey results and there was previous data available concerning the reliability of this diary. 14 In addition, this diary could be utilized, with minor changes, for it had been developed for use in the same geographical area and was available to the researcher. The original Michigan State University diary contained no space for the recorder to list where purchases were made, whether promotional specials were available, or the specific brand or name of the item purchased. Another limitation of the M. S.U. diary resulted 14 For a description of the different diary forms evaluated and tested see: Shaffer, op. cit., Chapter 16. 107 from the heavy concentration on food items by the operators of this panel . Non-food items were also to be audited as well as food items in this research, and it was necessary to add certain product cate- gories to the M.S. U. diary. (See Appendix A-12.) The original M. S. U. weekly consumer food purchase diary was modified as follows: 1. Spaces were provided for recording brand purchases, where purchased, and special price, coupon, or other promotional activity. In order to enable the researcher to have sufficient space for these items, it was necessary to delete in some cases grades and other descriptive data contained in some product classifications. Product classifications for soaps, detergents, starch, paper products and dog foods were added. These classi- fications replaced nuts, nut products, vitamins, and minerals. Questions pertaining to income received by the household for the week reported were deleted. It was felt that the initial information received from the questionnaire would be adequate because the length of this research was to . be only ten weeks. In addition income data was to be sought in a follow up interview by later researchers. Directions for keeping the diary were modified to accomo- dat‘e the use of a different incentive program for panel members. Diary Limitations and Recommendations Correspondence and requests for diary suggestions from panel members during the research period indicated the following information for modifying the diary format used: 108 Fruits and vegetable product categories should be broken into a fresh and processed section. Several members indicated this would have facilitated their completion of the diaries and eliminated the possibility of only consider- ing the classification as being only fresh or packed. Several comments were received early in the study and a reminder that _a_ll fresh and processed fruits and vegetables were to be included under the major category was com- municated to consumers through a panel newsletter. A provision for number of shopping trips and the days of these trips should be made. This would provide useful information in future studies of promotional effectiveness. Such promotional treatments as double stamp days, bonus stamps, and special promotions could be obtained from promotional audits. These items could then be compared to determine if certain patterns emerged for particular classifications (demographic or psychological) of panel members. Certain items could be combined that were not of impor- tance to the research. This would reduce the task of com- pleting the weekly report by the panel member. General categories could be established allowing the panel member to fill in purchases made under these classifications. The ledger diary would still be utilized but it would feature some of the advantages inherent in the journal type diary. Combining various classifications of meat products into one category would be an example. If only limited product categories are to be audited, the diary could be reduced to include only these categories plus a limited number of other classifications. No provision was made for the recorder to report where bread and milk were purchased. Because a larger pro- portion of some store's sales are accounted for by these items, restrictions of later types of analyses that can be made are present. 5Sudman, o . cit., 75-76. 109 Panel Incentive Program Little evidence is available on which to base the type and amount of material incentive needed to obtain maximum cooperation from panel members. Cash payments, merchandise incentives, or a combination of both appear most frequently used. The amounts of incentive necessary for panel members varies with individual panels. Sudman, in studying the accuracy of panel reporting, found the per- centage of households willing to maintain a continuing purchase diary and the accuracy of their recording appeared to be independent of the level of compensation. Correspondence with panel operators indicated that the proposed panel should be compensated for fullest cooperation for the ten week reporting period with an equivalent wholesale gift priced at $1. 50. This figure was increased to approximately $3, 00 because extensive interviews of panel cooperators were to be conducted at the close of the recording period. Such an additional compensation was used in an effort to obtain increased cooperation for this phase of the research. Several mail order premium houses were contacted in an effort to secure merchandise incentives in the $1.00 to $3.00 price range. In most cases, those available were $5.00 to $25.00. Because 1 6Ibid. 110 of the unavailability of satisfactory merchandise premiums and pos- sible administrative difficulties in making cash payments, a merchan- dise point schedule was developed in cooperation with a mail order house. (See Appendix A-5.) Published guidelines for choosing a premium were followed in developing the merchandise listl.7 The following four considerations were made in developing the premium schedule: 1. Premiums should be something the prospective respondent would find desirable. 2. Premiums had to be something that did not introduce bias, --A premium should be the kind of thing everybody in a given universe could be interested in. 3. Premiums had to be small and light enough to be easily mailed. 4. Premiums were not to exceed $3.00. Panel members were remunerated by points with bonus points given to encourage a greater continual return of completed diaries. Points needed to secure the majority of premiums required almost complete cooperation for ten weeks. By weighing the number of items from which premiums could be selected toward the high point side, continuous cooperation was expected to be more easily obtained. 1 7Paul L. Erdos, "Successful Mail Surveys: High Returns and How to Get Them, " Printer's Ink, (March 1, 1957) pp. 56-60. 111 Bonus points were given for mailing the completed diary on Sunday or Monday and certifying that 311 members of the house- hold's purchases were included. Points were added after the tenth week and totals provided to panel members who then selected the item( 3) which they wished to receive. On the return card indicating preference, panel members were also asked to indicate whether they saved trading stamps or not, and if so, which kinds. Any questions overlooked previously should be obtained at this time because of the relative assurance of a high response. Only two percent of the panel members receiving this card failed to either answer questions about stamps or return the card. Incentive Limitations and Recommendations In spite of what may appear to be a relatively problem- free portion of the study, several major problems were presented by this facet of the research. The two major areas in which these could be classified were: ( 1) point misunderstanding on the part of panel members; and (2) prize selection and procedure. It would be advisable for future studies of this type to con- fine prize selection to limited categories. A choice of two or three items in each point range of 100 points would eliminate much adminis- trative work necessitated by panel members requesting premiums for which they were not eligible. Panel members lacking only a few 112 points, on several occasions, attempted to obtain premiums which required additional points. A limited descriptive catalogue could be compiled which would assist in the recruiting effort of prospective panel members for a relatively small amount. In attempting to delineate factors bearing upon new prod- uct acceptance, a combination of incentive plans may be utilized. Panel members at the beginning of the study could be given a choice of choosing from the prize list or taking a chance on a larger more expensive prize to be selected by drawing. Once this decision was made it could not be changed. Points could then be given as chances for this lottery. Hopefully, some broad general measure of overall risk taking on the part of the consumer could be determined. The usefulness of such a measure of risk could then be applied to new product purchases by high risk takers as opposed to low risk takers. Points awarded should be mailed weekly in the form of coupons which can later be redeemed. This will eliminate confusion as to points earned by the panel member. Several panel members failed to adequately understand the method in which bonus points were earned, and much later correspondence could have been eliminated by use of coupons and more adequate explanation of the point system. This confusion developed from certain panel members not under stand- ing that they could only qualify for the number of diaries completed for bonus points one time. 113 Summar y Several communities were investigated as possible re- search sites. Foodtown was chosen as the community most meeting established requirements. This city was relatively isolated for food shopping from other communities by both distance and travel time. In addition, Foodtown had two local newspapers available normally featuring food store advertising. The employment base of the com- munity was large enough to preclude large number of residents mi- grating to other communities for employment and subsequent grocery shopping. It was anticipated that the number of outside variables ef- fecting food shopping could be reduced by conducting the research in such a self-contained and isolated community. Expressions of cooperation were obtained from warehouse groups, newspapers, and grocery store managers. In few cases was cooperation not available. However, due to the lack of permis- sion to audit four grocery stores, only eleven of the fifteen grocery stores present were subject to examination. These eleven stores accounted for about ninety-five percent of all grocery sales made in Foodtown. Nine product categories were chosen to be audited. This selection was based upon ease of auditing, distribution among outlets, vitality of movement, and the amount of promotional activity expected during the research. The panel diary was developed from a diary 114 previously used in a M. S. U. study. It was necessary to modify the original M. S. U. diary to accomodate this research. Certain product categories were deleted and other categories included to provide for reporting of non-food purchases. Panel members were to receive points for reporting which could be exchanged for merchandise selected from a premium list. Some limited field investigation was necessary preparatory to choosing product categories to be audited and diary development. Contacts were made with all grocery store operators in Foodtown following previous visitations for estimating each store's sales im- portance and product distribution among stores. However, prelimi- nary field investigations did not include actual store auditing. Audits were taken only after the completion of some audit training by field inve stigator s . CHAPTER IV EXECUTION OF STORE AUDITING AND PROMOTIONAL MONITORING Introduction Retail store audits were taken weekly in the eleven co- operating stores for thirteen weeks. These audits obtained consumer sales information for nine product categories from January 23, 1965 until April 24, 1965. The Foodtown consumer panel reported pur- chases during the last ten weeks that store audits were taken. Prior to auditing, a limited amount of training was given to field investigators. This training was not as extensive as desirable, but was thought sufficient due to the types of products being audited and the previous experience of auditors. In spite of this training and cautions made to field investigators, several errors in auditing later occurred. Food promotion taking place in Foodtown for the thirteen week period was monitored. Special emphasis was placed upon ob- taining promotional information directly relating to the nine audited product categories, but more generalized promotional activity was also observed and recorded. Several grocery stores in Foodtown 115 116 used a very limited amount of merchandising activity and the majority of promotional effort was made by five stores. Four of these five stores were extending cooperation for auditing their establishments. Period of Aduit The thirteen week period chosen for store auditing was selected for three reasons: (1) to insure the presence of different promotional treatments for the audited categories; (2) to provide a three week trial period of auditing for field investigators in addition to the ten week period of audits taken simultaneously with the panel recording period; and (3) to facilitate the use of part-time field per- sonnel. Several product categories were expected to receive heavy promotional treatments during this time. Warehouse grocery mer- chandisers had perviously given an indication of some products to be featured in later weekly sales plans when first interviewed. Canned tuna fish and dry dog food categories were selected for auditing based primarily on conversations with grocery merchandisers. In addition, canned tuna was expected to receive extensive promotional treatments by most stores during the Lenten Season. It was therefore desirable that the audit period include the entire period of Lent. Because inexperienced personnel were used in auditing, it was desirable that the period of audits be long enough to allow for some on-the -job experience and store product familiarization, by 117 investigators. Investigators were graduate students enrolled in a graduate food marketing series, and store audits had to be taken during a time that facilitated the availability of these personnel. The presence of university vacations was minimized during this thirteen week period. Training of Auditors Audits were taken by students enrolled in the graduate food marketing seminar at Michigan State University. The background of these students dictated a’ decision to minimize the amount of audit training. 1 The limited number of product categories to be audited also contributed to this reduced training period. A visitation was made to the A. C. Nielsen Company's Training School for Auditors before field personnel were trained in auditing procedures. During this visit, Nielsen personnel stressed the extensive preparation that their auditor trainees received prior to qualifying for field positions. 2 The importance of accuracy, fa- miliarity with forms, and product classification problems were areas that Nielsen personnel suggested be covered in training. These students were experienced personnel in grocery store operations with experience as store managers, food products salesmen, or independent grocery store owner-operators. The ex- perience in food store operations of these men ranged from three to twenty-two years. 2'Trainees are trained in audit procedures and techniques for a period of 6 months by the A. C. Nielsen Company before being given field assignments. 118 Training of student auditors concentrated primarily on theory of retail store audits, audit forms to be used, product cate- gories to be audited, importance of retailer corporation, and controls to be excercised. These topics were covered in class sessions fol- lowed by a three week period of trial audits. Problem areas discovered during the weekly audits were covered in follow-up class sessions. Information supplied by A. C. Nielsen Company was used in the explanation of the theory and uses of retail audits. Nielsen per- sonnel emphasized that field auditors should fir st be exposed to the basic idea underlying retail audit research before recurring problem areas are covered. The basic idea of retail audits was presented using the framework of the equation: consumer sales = opening, inven- tory + purchases - present inventory. This equation was later modi- fied to accomodate various problems leading to possible inaccuracies. Among the points presented were: (1) inaccurate counts; (2) missing invoices; (3) inter-store invoices; and (4) returns for credit. These four problem areas were covered by sample problems illustrating their effects upon the basic equation and results. Nielsen methods used in recognizing and remedying problem areas were presented. The audit form used in this research was developed by incorporating several features of commercial audit forms. Auditors spent a total of six hours in practice sessions computing product movements using this audit form. Special examples and sample 119 problems were chosen to emphasize the cautions needing to be exer- cised in field work. (See Appendix B-l. ) Auditors not only became cognizant of the reasons for inaccuracies, but also gained a working knowledge with the audit form in these practice periods. A list of all brands stocked in cooperating stores was com- piled from order forms and price lists of warehouses. This master list indicated that several loca1,regional,and private brands were carried by cooperating grocery stores. Lists of brands expected for the individual stores were compiled by product group and given to each auditor. Auditors were expected to familiarize themselves with the brands carried by the store for which they were responsible. These brands were to be transferred to audit sheets, and were expected to be of some assistance in taking counts of back-room stocks during the initial inventory. The master list provided each auditor may not have contained all brands, and each auditor was cautioned to add brands carried to his audit sheets on his initial inventory. The pos- sibility of retailers stocking brands for the first time during the audit period was discussed, and auditors were cautioned to constantly be alert for such items. Store operators had little incentive to continue in their cooperation, and the importance of a friendly and courteous manner for establishing and maintaining rapport with owners and operators was stressed. Good relationships were necessary not only with store 120 personnel but customers as well. Shelves and store rooms were to be arranged as found at the beginning of each audit. Permission was to be requested from the retailer, at the initial audit, for putting all products in their proper space at the end of each visit without chang- ing facings or merchandise arrangement. Products needing to be re- marked for price or out -of-stock conditions were to be called to the attention of the manager. Retail audits entail an outside person to be present in the store for several hours, and the possibility of a cumu- lative effect of this "nusiance" may result in the withdrawal of cooper- ation. It was emphasized to field auditorsthat only by extending as- sistance and maintaining good relationships with store personnel could cooperation be expected to continue. Auditors were reimbursed only for travel expenses and milage forms were to be submitted weekly. Completed audit forms for each weekly audit were to be submitted at the same time. Auditors were to be monitored by both early observational trips and an indepen- dent audit taken at various times during the research. In addition, invoices were inspected periodically at certain warehouses by the writer. Execution of Store Audits Each auditor was assigned to audit the same store each week throughout the research. The three stores having the largest estimated weekly sales volume were each assigned an additional person. 121 Three larger stores. were audited by persons having prior experience as managers of similar type stores. Letters of introduction were provided each auditor for presentation at the time of the initial inter- View. Stores were audited each Tuesday morning. Tuesday was selected for several reasons. Because of heavy weekend shopping traffic, several stores did extensive restocking of shelves on Monday, and this increased the possibility of auditors failing to account for all in-transit items. Shelf stocks may have been replaced from carts or pallets moving from stocks to selling floor after shelf items had been inventoried by the auditor. Wednesday, Fridays, and Saturdays were undesirable days for executing audits because these were heavy shopper traffic days. Double or bonus stamps were given on Wednesday by two of the three largest grocery stores, The normal heavy traffic on weekends pre- cluded audits being taken on these days if auditors were to be least obtrusive. The first three weeks of store audits were not conducted in conjunction with the operation of the consumer panel. This three week period was needed to provide audit experience and store familiarity for auditors. The list of brands carried by cooperating stores had to be modified during the opening inventory and it was necessary for store auditors to eliminate certain brands from weekly audit forms that were 122 not carried and add others to the form. These forms were then re- produced by individual auditors in the quantity necessary for the re- mainder of the research. Questions regarding the classification of items were resolved during the three week trial period. Dietetic canned goods and specialized dog foods were two product categories requiring special attention. The decision not to audit these items or banquet sizes of canned goods was made. Special problems were encountered by auditors in physical counts of special merchandise displays. The importance of accurate counts was stressed even if they could be obtained only at the cost of a tearing down and recon- structing displays. At the end of the three week period when auditors had become acquainted with locations of product categories and stocks, a measure of the time involved in each stores audit was obtained. The time required for two persons to audit the nine product categories for the three largest grocery stores was one -and-one -half to two-and- one -half hours. The eight smaller stores required an average of one- hour for auditing the nine classifications. In one of the smallest stores, the entire audit was accomplished in twenty minutes. 123 Ououm germ .NoH N E. 3 «NH 42 NN No No. .1 N «N poses... neon: .NGQEDZ argon. NN N N N S a N N. 2 N r a .333 3. N a. N e N N m e a N N manna ornfla a .83 NN N r a 2 Na N N a a N N .35 Noarooo NNN N a a 0N N a Z 8 a N a oooa Noo a: N Z N N 8 N Na 2 N N N oouoo oceans ON 2 N a a N a N m a a z ransom aonam NN N N N Na a N N 2 a N N Nora and. as N N a Na 2 N N N N a N room oooooo No a e a 2 E a N. N N r N moroooa oooooo anoNoooo room 2 2 a. w N N. N N, a N. N z 5 won—5.94 manna mo nongz fimuoh * «@330 fimuom wdfldusmoou m 9. Rowen—MO auguounm .mmuoh poo.“ drawn mcwumuodooo Sm .NoH Eamon—mo pospoud come 5. won—mp5s was”: mo nonfisztt .w 3an. 12.4 Audit Problems and Recommendations Obtaining_lnvoice s It became apparent after the initial two weeks of audits that invoices for two stores would be difficult to obtain. One of these stores received deliveries on Tuesday and Thursday afternoons from the warehouse group. These invoices were sent the same night to another store owned and operated by the same management. Central records were kept at this out of town location a considerable distance away. Auditors, because of the lack of invoices, could only take inventories each week, and not calculate weekly sales. To overcome this problem, arrangements were made to secure invoices every fourth week by mak- ing a special visit to the central office to inspect all invoices and allowances for credit. Invoices were not available for the largest grocery store Foodtown. All records were maintained at the chain's central office and only shipping documents accompanied warehouse deliveries. ' Any shortages or overages were recorded on these documents for credits or deductions. These corrected shipping records were then returned to the central accounting office. Corrected invoices had to be copied on a bi-weekly basis at chain headquarters. However, due to an over- sight by the person recording this information, two weeks of invoice records were not obtained until after the completion‘of audits. This necessitated going back and computing weekly product movement for 125 the nine audited categories at a later date. Computations for product movement for this one store required approximately 125 more hours. This time could have been reduced considerably if movements had been computed by the auditors of that particular store. If necessary funds had been available to pay for duplicate copies of invoices, many later problems could have been alleviated. In addition to classification and invoice problems appearing early in the field auditing, several minor problem areas also developed. These problems were as follows: (1) new products or first time stocked items; (2) weekly dollar sales from stores; (3) scheduling and control of auditors; and (4) audit form used. New Products Two new products were introduced to one of the audited product categories during the research. Hunts Pizza Catsup and Hunts Hickory Catsup. Auditors in all stores but one recorded these new additions to product line. In this one store, these items were intro- duced by use of a special merchandise display not closely located to other products in this category. The auditor assigned to this store failed to record the addition on two successive weeks. Only when this oversight was called to his attention by the writer, acting in the role of an observer, did he become cognizant of this product. Total prod- uct movement for this item was determined from invoices, but it was 126 necessary to average product movement for the two week period. This was thought to be the best method for treating this error even though it was not completely accurate, because of the likelihood of differences in exposure and acceptance for the two weekly periods. A systematic procedure to be used for monitoring new products would be desirous. This is particularly important when using part-time personnel. A check off item may be necessary for each product category to serve as a reminder to the auditor. An appropriate space for (new products - yes or no) could be included for each clas- sification. Each auditor would be required each week to check this item against both inventory and invoices. Possibly, the best solution to insuring the auditing of new additions is to increase the'amount of emphasis given to this in training sessions. Weekly Dollar Sales of Cooperators Cooperators were originally requested to provide weekly sales data. Sales data was expected to be obtained for total weekly sales with additional break-downs into meat, produce, dairy, and dry grocery sales. Several cooperators initially agreeing to furnish such data later would not make sales totals available. Only seven of the eleven audited stores provided weekly total sales figures. Three stores furnished only an estimated yearly sales total, and two of these stores made no weekly totals and only kept monthly figures as required by 127 state sales tax regulations. One store had cooperated for auditing only on the condition sales information would not be made available. Two of the stores providing only yearly sales totals were among the least important stores in sales in the community. Smaller stores were found more reluctant in furnishing dollar sales information than the larger outlets. Because of their small size, it may be advisable to extend a small payment for maintaining and furnishing sales infor- mation. With the apparent small income from these establishments, such an offer would appear justified and in all probability acceptable. Audit Forms Extracting data from audit forms was complicated by the lack of pre-printed audit forms by brands and sizes. Pre-printed forms were unavailable because the product offer of each cooperating store was unknown until the opening inventory had been taken. Because each auditor added or deleted certain brands or sizes at the time of his initial visit, brands or sizes carried by several stores were not in the same order for all stores. Pre-printing audit forms by brand and sizes after the opening inventory would have minimized much of the time required to extract data. Even though several stores had a very limited product offering, pre-printing all brands and sizes avail- able in the community would have allowed for pre-coding and facilitated later tabulation of the data. 128 Promotional Monitoring Preliminary plans for Phase III of the research program called for developing some estimate of the impact various promotional treatments have upon shopping behavior. The desirability of using various promotional treatments depends not only upon the retailer who elects to implement them, but also upon the impact the promotion has upon different groups of customers. Three groups of customers would be affected by promotional variations: ( l) the present customers of the retailer introducing the variation; (2) those customers purchasing from direct competitors; and (3) customers of other establishments not considered to be direct competitors. If measures of effectiveness for various promotional alternatives are later to be made, it was necessary at this time to monitor the impace of various promotions upon customer segments. Some method of obtaining measure of competitive reactions also appeared necessary. It was therefore desirable to obtain measures of promotional variations using both a macro and micro approach. An audit of all promotional activity directly relating to the nine audited product categories was made. In addition, a limited measure of all food promotion in this community was obtained. Be- cause retailers are concerned with total transactions rather than individual items, it was necessary to monitor food promotions indirectly affecting the product movement of the nine audited categories. 129 The consumer panel diaries were designed to obtain limited information relating to impact of promotional variations on market segments. Panel diaries provided space for recording special prices, coupons, or other promotional activity relating to specific purchased products. Diaries are useful for providing measures of positive actions taken to promotion, but can furnish little information about the total promotional activity present. Retail store audits were used to obtain some measure of the promotional information directly associated with the selected categories chosen for auditing. Shelf space allocation, price, and merchandising activity were recorded. With adequate resources, it would be possible to obtain measures of all in-store promotion with no foreseeable difficulty. In addition to panel diaries and store audits, other sources were monitored for promotion. All newspaper grocery store adver- tising for a thirteen week period was obtained. Both Foodtown news- papers were observed and a file of grocery advertisements was main- tained on a daily and weekly basis. All cooperating and non-cooperating food store advertisements were recorded. Weekly sales plans were provided by each warehouse group that made these available to member stores. Three cooperating stores religiously followed the warehouse weekly plan in their merchandising activity. A fourth cooperating store was receiving a weekly warehouse 130 sales plan but did not adhere to it closely. Each weekly sales plan was obtained prior to the week it covered. Store auditors then famil- iarized themselves with the plan and observed the extent to which it was employed by cooperating stores. Store managers were requested at each audit to indicate the amount of local radio or television placed the previous week by their store. In only one case was radio advertising used and this was institutional. No monitoring of magazine advertising, radio, or tele- vision was employed. The importance of this omission is realized and was dictated by manpower and financial restrictions. The explanatory nature of this investigation did not seem to justify this part of the monitoring to the exclusion of other aspects. At the completion of the research panel members were asked to report whether or not they saved trading stamps and the kinds that they collected. A measurement of the characteristics of stores "most" liked and "least" liked was obtained by the researcher conducting Phase II of the project, from both panel members and other persons who had refused to become panel cooperators or had not been solicited. Measures of certain store characteristics were obtained by use of a scale using a most liked - least liked - spectrum. Measures of respon- siveness and non-responsiveness of food store advertisements and personal influence were obtained at this time. (See Appendix C.) 131 Summar y Weekly store audits were taken of eleven cooperating grocery stores in Foodtown. These audits were taken by inexperienced auditors who had received only a limited amount of previous auditing training. The first three weeks of the thirteen week audit period was used as a training period for the field auditors. In spite of cautions given in prior classroom training sessions about possible problem areas, several mistakes were made by auditors during trial audits. Several problems developed during the last ten weeks of store audits but most of minor importance. However, one serious problem devel- oped which places limitations on future analyses of the research data. Cooperating retailers had expressed a willingness to provide the researcher with weekly dollar sales figures, but only seven of the eleven stores later provided this information. Fortunately, those not providing sales figures were smaller stores. A request for yearly sales information was then made rather than pressing for weekly sales information and possibly losing cooperation for audits. Three of the four stores agreed to cooperate in providing yearly sales information. Promotional variations were monitored by use of panel diaries and store audits. In addition, local newspaper, radio, and TV advertising was recorded. No measures of magazines entering the households of panel members or outside radio and television were obtained. At the end of the consumer panel reporting period, 132 panel members were asked to furnish additional information about trading stamps, perception of Foodtown grocery stores, and their observation of grocery store advertising. CHAPTER V RECRUITING AND OPERATING THE FOODTOWN CONSUMER PANEL Introduction An integral part of the research plan for a total market investigation was the establishment and operation of a consumer pur- chase panel. This panel was to be operated concurrently with the retail store audits of food outlets servicing the community. The ob- jective of the panel was to obtain some measurement of food purchasing behavior by a selected group of consumers on a current basis. The data obtained was to then be analyzed in aggregate and compared with audit data for reliability and use in projecting product movement. A later phase of the research program called for the analysis by certain demographic and psychological characteristics of reporting panel members. The panel was to provide the repetitive-continual informa- tion necessary for delineating changes in buying habits for the period. Such information, compiled and analyzed on an individual household reporting basis, would facilitate the examination of internal shifts or changes that had taken place. In addition, the panel was to be utilized for a later phase exploratory study investigating the characteristics 133 134 of panel members, non-panel members, and a random sample of the population. This was to ascertain whether these groups differed in relation to measures of needs, values, and demographic characteris- tics. The establishment and operation of the consumer panel centered about three major areas. These were: (1) the sampling problem; (2) obtaining cooperation and information for panel participants and non-participants; and (3) the reporting burden and degree of coopera- tion. The Sampling Problem--Literature Review A probability sample design accurately drawn allows pre- cise inferences to be made about the population for the period of time covered. The panel operator must consider the problem of securing COOperation when choosing his sampling technique. This becomes more Critical when one considers that cooperation must be secured on a Continuing basis rather than for just one time. Even though the supe- riOrity of the probability design for panel construction as compared to a unnm H309 mfiGD argon—Gaza Sum pcmum uncum goo? >Udam geopooh you womb Euoh ”nodes upon—m. 3.»..qu NIm @5930: Iaofiaoflfimmafio podpoum 264 APPENDIX C Questionnaire for Panel Members, Non-Cooperators, and Random Sample of Community Households 266 INTERVIEW CONTROL SHEEI, l. Interviewer's Name , g_ Control Number II. Respondent's Name - _~_Date Address ___ Time _fi Control Number Ill. Interview Completed I i Interview Not Completed Reason: Not at home Refused Why? 267 V - 0. Show respondent card #I. Ask respondent to select the three occupations that she things are most important. I. Volunteer church worker (R) 2. ______Banker (E) 3. ______Doctor (5) h. __ Head of a company (E) 5. ______Social worker . (S) 6.______ Principal of a school (5) 7. _____ Real estate developer (E) 8. ______Sunday school teacher . (R) 9. _________ Clergyman _ (R) 268 INSTRUCTIONS - PART | Please indicate for each of the following statements that point on the scale which best describes you. Example: I enjoy outdoor sports. very much somewhat like neither like me somewhat unlike very much‘ like me me nor unlike me me like me If you enjoy outdoor sports very much, you would check Point I on the scale. very much like me If you sort of dislike outdoor sports, you would check Point h on the scale. somewhat unlike me If you do not care one way or another about outdoor sports, you would check Point 3 on the scale. neither like me nor unlike me l.. I have a definite day each week to do my washing and ironing. very much somewhat like neither like me somewhat unlike very much like me me nor unlike me me unlike me 2. I try to.flnd ways to improve my housekeeping. very much somewhat like neither like me somewhat unlike very much like me me nor unlike me ' me unlike me 3° When there's a choice, I would rather work with other people than work alone. very much somewhat like neither like me somewhat unlike very much like me me nor unlike me me unlike me A. I look to others to help me decide things about my home and family. very much somewhat like neither like me somewhat unlike very much like me me nor unlike me me unlike me 269 5. I settle arguments and disputes between others. neither like me nor unlike me smmMmtle me very much like me somewhat me very much somewhat likei neither like me EEEEwhat unlike like me me nor unlike me me ' 6.. I'm excited by new fads and fashions. very much somewhat like neither like me somewhat unlike like me me nor unlike me me 7. I can I'tell off” other people in public. very much somewhat like neither like me somewhat unlike like me me nor unlike me me 8. I do exactly what I want to do. very much somewhat like neither like me somewhat unlike like me me nor unlike me me 9. I come and go as I choose. very much somewhat like neither like me somewhat unlike like me me nor unlike me me ID. I get other people's opinions before I decide. very much somewhat like neither like me - somewhat unlike like me me nor unlike me me ll. I like being the center of attention. very much somewhat like neither like me somewhat unlike like me me nor unlike me me l2. I solve difficult puzzles and problems. very much somewhat like neither like me somewhat unlike like me me nor unlike me me l3. I keep bills, letters, and other things according to a system. unlike very much unlike me very much unlike me very much unlike me .very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me Z70 lh. i take trips to places I know well. very much somewhat like neither like me somewhat like me me nor unlike me- me IS. I often participate in friendly groups. very much somewhat like neither like me somewhat like me me nor unlike me me l6. i tell others off when I disagree with them. very much somewhat like neither like me somewhat like me me nor unlike me me I7. I try new things and experiment with new things. very much somewhat like neither like me somewhat like me me. nor unlike me me l8. I tell amusing jokes and stories. very much somewhat like neither like me somewhat like me me nor unlike me me IS. I follow instructions exactly. very much somewhat like neither like me somewhat like me me nor unlike me me 20. I'm independent of others in making decisions. very much somewhat like neither like me somewhat like me me nor unlike me me 2i. I prepare meals at the same time every day. very much somewhat like neither like me somewhat like me me nor unlike me me 22. I contact friends frequently by mail or telephone. very much somewhat like neither like me somewhat like me me nor unlike me me unlike unlike unlike unlike unlike unlike unlike unlike unlike very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me “IL 23. I'm a leader in groups to which I belong. very much ESEEhhat like (neither like me like me me nor unlike me 24. I share things with friends. very much somewhat like neither like me. like me me nor unlike me 25. ‘I'm reluctant to take on torgh jobs. neither like me very much somewhat like like me me nor unlike me 26. I take trips to new and different places. very much somewhat like neither like me like me me nor unlike me 27. I prefer to have someone else in charge of very much somewhat like neither like me nor unlike me like me me 28. I argue for my point of view. .very much somewhat like neither like me like me me nor unlike me 29. .I do new and different things. very much somewhat like neither like me like me me nor unlike me 30. I talk about personal accomplishments. very much somewhat like neither like me like me me nor unlike me 3]. Household chores always seem to pile up. very much somewhat like neither like me like me me nor unlike me 32. i plan a weekly meal menu. neither like me nor unlike me very much somewhat like like me me somewhat unlike me somewhat unlike me somewhat unlike me somewhat uni i ke' me group activity. somewhat unlike me somewhat unlike me somewhat unlike me somewhat unlike me I‘- somewhat unlike me somewhat unlike me very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me 272 33. I accomplish tasks that others recognize as requiring skill and effort. very much somewhat liken neither like me somewhat unlike like me me. nor unlike me me 3h. I rarely contact friends by mail or telephone. very much somewhat like neither like me somewhat unlike lIke me me nor unlike me me 35. I like respOnsibilities and obligations. I very much somewhat like neither like me somewhat unlike lIke me me nor unlike me me ' 36. i say witty and clever things. very much somewhat like neither like me' somewhat unlike like me me nor unlike me me 37. l supervise and direct the actions of others. very much somewhat like neither like me somewhat unlike like me me nor unlike me me 38. I get "back at" others for insults. very much somewhat like neither like me somewhat unlike like me me nor unlike me me 39. I get angry easily. very much somewhat like neither like me somewhat unlike like me me nor unlike me me 40. I tell others that they have done a good job. very much somewhat like neither like me somewhat unlike like me me nor unlike me me “I. I pay very little attention to instructions on food packages. very much somewhat like neither like me somewhat unlike like me me nor unlike me , me very much unlike me very much unlike me very much, unlike me very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me 273 #2. I'm reluctant to criticize other people. very much somewhat like ’neither like me like me me nor unlike me somewhat me #3. I'm uncomfortable when l'm the center of attention. very much somewhat like neither like me like me me nor unlike me #4. i accept and perform difficult jobs. very much somewhat like , neither like me like me me nor unlike me 45. I avoid responsibilities and obligations. very much somewhat like neither like me like me me nor unlike me somewhat me somewhat me I somewhat, me #6. I am one of the first to buy new brands of foods. very much somewhat like neither like me like me me nor unlike me somewhat me unlike unlike unlike unlike unlike ‘ very much unlike me very much unlike me very much unlike me very much unlike me very much unlike me 274 INSTRUCTIONS - PART ll Please indicate the extent to which you agree or disagree with each of the following statementS. Example: Giving to the United Fund. strongly agree agree undecided disagree strongly disagree If you strongly agree with giving to the United Fund, place a check ( ) above ”strongly agree." strongly agree agree . undecided disagree strongly disagree If you just disagree with giving to the United Fund, place a check ( ) above "disagree." - strongly agree agree undecided disagree strongly disagree If you just agree with giving to the United Fund, place a check ( ) above "agree.” strongly agree agree undecided disagree strongly disagree I. Volunteer social or public service work. strongly agree agree undecided disagree strongly disagree 2. _Watching movies and TV programs about business. strongly agree agree undecided disagree strongly disagree 3. Going to church every Sunday and every religious holiday. strongly agree agree undecided disagree strongly disagree A. Government control of business. strongly agfee agree undecided disagree strongly disagree 5. Watching movies and TV programs about nursing and the medical profession. strongly agree agree undecided disagree strongly disagree 6. Shopping only at stores that offer the best prices. strongly agree agree undecided disagree V strongly disagree 275 7. Defending religious and racial freedom. strongly agree agree undecided 8. Religious training in the home. strangly agree agree undecided 9. Contributions to religious causes. strongly agree agree undecided l0. Achuiring wealth. strongly agree agree undecided ll. House visits from the clergy. strongly agree agree undecided l2. Helping other people with their problems. strongly agree agree undecided l3. Being successful in business. strongly agree agree undecided lh. Religious discussions with family or friends. strongly agree agree undecided IS. Contributions to community activities such as Boy Scouts or strongly agree agree undecided disagree disagree disagree disagree. disagree disagree disagree disagree disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree the United Fund. strongly disagree PART III 276_ Show respondent card #2. Ask respondent to indicate in rank order those five characteristics from the following list of characteristics which she thinks are most important for a food store to have. If she thinks it is most important for a food store to be ”clean and neat," put a #l next to ”clean and neat.” If ”modern” is the next most important characteristic, put a #2 next to ”modern." The respondent's last selection will be indicated by the #5. Clean and neat _____.Large ______ Friendly ______ Modern Well-established . _____.A lot of price specials Reliable _____ Wide assortment of merchandise Low prices ,______Easy}to shop at Show respondent card #3. Ask respondent to select the three siOtB Which She likes to 5UP at most. indicate preferences by the initial M. I. _____.Mileski's Grocery .2. ______Gordon's Cash Market 3. _____.Fred's Super Market ii. _____A s P (NC) 5. _____DWest End Grocery (NC) 6. ______ Fate's 7. _____ Kroger's 8. ______0'Connor's 9. ______City Fruit Market l0. _____,Falsetta's Superette ll. ______Wilson's Grocery l2.._____ Meijer's l3. __ Al's C 8 M Grocery lh.._____ Piselli's Grocery (NC) l5. ______Farmer's Market (NC) l6. _____ Safeway Now ask respondent to indicate those three stores which she likes to shop at least. indicate respondent's preferences by the initial L. 277 Please indicate how you feel about each of the following characteristics of those stores which you like to shop at most. EXplanation: If you think a store is more "old-fashioned” than "modern," you would put a check (J) in a space (:__L_/_:) which is closer to the word "old-fashioned” than it is to the word “modern." Example I. Modern: : : : : l/: : :Oid-fash ioned If you think a store is very "old-fashioned," you would put a check in a space which is very close to the word ”old-fashioned." Example 2. Modern: : : : : : : /: Old-fashioned Please do this for all the following pairs of words. Modern :______:____:_____:____:____:____:_____: 0i d-fa sh ioned Clean and neat:___:____:____:____:____:___:____:Dus ty and unorganized. Unf r lend l y :_.____:____'__:_____:___:_____:____:______: Fr i end I y High prices :____':‘___:_____:_____:_____:_____:___:Low pr ices Smal 1 :_____:____:_____:____‘:___:____:___:Large Wei l-es tab l i shed :____:____:____:_____:___:____:____: Newcomer Few price Specials:__._:____:_____:____:____:_____:___:A lot of price specials Hard to shop at:______:____:___:_____:____:___:____:£asy to shop at Rel iable:_____:_____:_______:_‘____:______:_____:____:Unrel iable Wide assortment:_____:____:______:___:___:____:_____:Narrow assortment of of merchandise merchandise 278 3 Please indicate how you feel about each of the following characteristics of those stores which you like to shop at least. Modern: : : : : :Old-fashioned Clean and neat: : : : ': :Dusty and unorganized Unfriendly: : : : : ' ' :Friendly High prices :___:___:___:___:___:___:__;_:Low pr ices Smal l :______: : : : :_____:___:Large Wei I -es tabl i shed :______:_____:____:_____:____.:_____:__-____: Newcomer Few price Specials:______:____:___:____:______:_____:____:A'lot of price specials Hard to shop at:____;___ : : : : : :Easy to shop at Reliable:____: : : : : : :Unreliable Hide assortment: : : : : : : ~ :Narrow assortment of of merchandise ' merchandise Please indicate the kind of people you think shop at the store which you like to shop at most. ' Energetic: : : : : : : :Lazy Thrifty shoppers: .: : : : : : :Spenders Old-fashioned: : : : : : : :Hodern Close family ties: :Family members come and go :Cook because they have to ee ee e ee ee 09 Enjoy cooking: Live in expensive homes: : : : :Live in average homes :Older people Younger people: Smart food shoppers: :Buy on impulse Nice people: :Snobs O. I. O O. O. O. ShOp because they have to: : m :Enjoy food buying Take life easy:' : : ‘ :Always on the go 279 Please indicate the kind of people you think shop at the stores which you like to shop at least. Energetic:_____: : : :____:_____:_____:i.azy Thr i f ty shoppers :____:____:_____:___:____:_____:___:Spenders OI d-fash i oned :____:_____:_____:______:_'_____:____:_____: Modern Close family ties :____:__:____:_____:_____:____:___:Family members come and Enjoy cooking:_____:______:_____:____;:____:____: :ggok because they have to Live in expensive homes:_____:____:_____:___:____:____:___:Live in average homes Younger peep] e :_____:_____:____:_____:_____:____:_____: Older people Smart food shoppers:_____:____:____:___:____:____:____:Buy an impulse Nice peopl e:_____:____:_____:_____:_____:____:_____:Snobs ShOp because they have to:_____:____:_____:_____;_:____:_____:___:Enjoy food buying Take life easy: : : : : : : :Always on the go 280 8.1. Do you usually read food store advertisements? Yes No “ If NO, ask Question #8.2. If YES, ask: 8.1.1. What do you look for in these advertisements? 8.1.2. What do you like most about these advertisements? J S. 1.3. What do you dislike most about these advertise- ments? - 8.2. How many times a week do you sh0p? 1 2 3 4 or more 8.3. Do you make all your food purchases at one store? Yes No If YES, proceed using Schedule B. Ask Question #S.4.B. If NO, proceed using Schedule A. Ask Question #S.4. 8.4. 3.5. 8.6. 8.7. 281 SCHEDULE A When making food purchases, how many stores do you shOp at? 2 3 4 5 or more Do you buy your meat in the same store in which you buy other grocery items? Yes No 3.5.1. For either answer, ask "Why?" 8.5.2. If N0, ask: "Where do you purchase meat?" store name store name Do you tend to Shep at different food Stores for different types of products? For example, would you purchase canned goods in one store, produce items in another store and packaged groceries in another? Record total answer: Which stores do you shOp at most frequently? (Rank in decreasing frequency.) most frequent less frequent less frequent less frequent 3.8. 8.9. .8010. 8.11. 8.12. 90L: Who among your close friends Sh0p at (name of store most frequently shappcd at from Question #S.5.)? ‘ name name name Do you know anyone else who shops there? Yes No name name How often have you seen advertisements for this food store? every day twice a week _____pnce a week occasionally never mwaI-a What do you like most about these advertisements? Why do you shOp most frequently at (name of most frequent store in Question #S.S.)? Respondent may use more than one category. Use these items on 3 x 5 cards. Quality of food . Appearance - clean, neat Friendly Low prices Wide assortments of merchandise Well-established Modern Price Specials Easy to get to . Easy to shop at i . Other: (Ask for OTHER responses) HOOQNOU—L‘UNI—l i—‘i-I' 283 8.13. Why do you shOp less frequently at (name of least frequent store in Question #S.S.)? Respondent may use more than one category. Quality of food is sometimes questionable Not clean and neat appearing all the time Not too friendly Prices are above average ‘ Not a broad enough selection of merchandise It's a new store in town Not very modern Not enough price specials Not too easy to get to __w__Not too easy to shOp at Other: (Ask for OTHER reaponses) HD—l ‘ HOAOODNO‘Mwap-J 8.14. Who among your close friends shop there? name name name 8.15. Do you know of anyone else who sheps there? name name name 8.16. How often have you seen advertisements for this food store? every day _“__'_twice a week —. __once a week ”n- occasionally ____n__neve L’ UIJ-‘LJMH 8.17. what do you like or dislike about these advertisements? 8.18. 3.19.‘ 8.20. 8.21. 8.22. 284 During the last four months or so, have you shopped at any food stores other than the food store at which you shOp most frequently? Yes No 8.18.1 If YES, which ones? (Interviewer: _You will have to remember the name(s) of this store (these stores) for forthcoming questions). If NO, ask Question $.27. Was it the first time you had ever shOpped there? Yes No Other *0— What prompted you to change food stores? Who among your close friends shOp there? name name- name Do you know of anyone else who shOps there? name name name 285 8.23. Before you switched to this new food store, was there anything about the store that attracted you? $.24. Did anyone recommend the new store to you? Yes No 8.24;l If YES, who was it? name -.r- _..- ..- name 3.24.2 If YES, what did they say? 8.25. How often have you seen or heard advertisements for this store? Would you say... 1 everyday 2 twice a week 3 once a week 4 5 occasionally never 8.26. What do you like most about these advertisements? 8.27. Does anyone help you with the food shOpping? Yes No 3.27.1 If YES, who? relative (ask relationship) friend - name 8.28. 3.29. 8.30. S.3l.' 8.32. 286 Does anyone accompany you when you shep? Yes ' No 8.28.1 If YES, how often? ' 8.28.2. If was, who? relative (ask relationship) friend - name Do you usually prepare a shOpping list? Yes No If YES, ask Question #S.30. If NO, ask Question #S.32. Do food advertisements help you in preparing a sh0pping list? Yes No 8.30.1. For either answer, ask "Why?" On your shopping list do you indicate the items you want by brand name? (For example, do you indicate "Peaches" or "Libby Peaches?") By item By brand Do you find that food store advertisements which you have seen or. heard are helpful while you are in the store shOpping? Yes No 8.32.1. For either answer, ask "Why?" 287 8.33. If you wanted to find out any information about a particular brand of food or a food store, who would you be most likely to ask? name address 3.34. How long has she lived in this community? l‘- years 8.35. Do you know her personally? Yes No 8.35.1. If YES, how did you happen to meet her? 8.36. What is her husband's occupation? $.37. Does she work? Yes No 3.37.1 If YES, ask "What does she do?" 8.38. Do you know approximately how old she is? 5.4.8. 5.4.8 $06.8. 5.7.8. $08.8. $0908. 288 SCHEDULE 8 Which food store do you_shop at? Who among your close friends shop there? name name name Do you know of anyone else who shops there? Yes No name name t; How often have you seen advertisements for this food store? I. every day 2. twice a week 3. _____.pnce a week h. occasionally 5. never What do you like most about these advertisements? Why do you shop at (name of store)? ReSpondent may use more than one category. Quality of food Appearance - neat, clean Friendly Low prices Wide assortments of merchandise Well-established Modern Price specials Easy to get to Easy to shop at Other: (Ask for OTHER responses) mm -‘CDUDCO\JO‘UWJ?U9h3-i o dd 289 5.10.8. Does anyone help you with the food shopping? Yes No 5.10.8. If YES, who? relative (ask relationship) friend - name 5.11.8. Does anyone accompany you when you shOp? Yes No mm 5.11.8.2. If YES, who? relative (ask relationship) friend - name 5.12.8. Do you usually prepare a shopping list? Yes No m If YES, ask Question S.l3.8. If N0, ask Question 5.15.8. S.l3.B. 00 food advertisements help you in preparing a shopping list? Yes No 5.30.1 For either answer, ask ”WHY?” S.1h.8. On your shOpping list do you indicate the items you want by brand name? (For example, do you indicate “Peaches” or "Libby Peaches?”) By item 8y brand 5.15.8. Do you find that food store advertisements which you have seen or heard are helpful while you are in the store shopping? Yes No 5.15.8.1. For either answer, ask “WHY?” 290 8.16.8. If you wanted to find out any information about a particular brand of food or a food store, who would you be most likely to ask? name address 5.17.8. How long has she lived in this community? years ~ 5.18.8. Do you know her personally? Yes No 5.18.8.1. 'If YES, how did you happen to meet her? 5.19.8. What is her husband's occupation? 5.20.8. Does she work? Yes No 5.20.8.1. If YES, ask ”What does she do?” 5.21.8. Do you know approximately how old she is? 291 Questions 8.1. and 8.2. - to be completed by interviewer R.l. Sex: M F R.2. Type of dwelling: House ______Aparbment._____ Trailer R.3. For interviewees residing in a house ask: Do you own or rent this house? Own __ Rent __ Other R.h. Are you married? Yes No ——_c———h If YES, omit Question #R. 5. and ask Question #R. 6. If N0, ask Question #R. S. R.5. Have you ever been married? Yes No “ If YES, ask Question #R.6. If NO, ask Question #R.7. R.6. How many children do you have? None 1 2 3 4 over h 8.7. In what year did you complete your last year of school? R.8. What is the occupation of the head of the house? (Interviewer - lf household head is not the husband, please indicate who it is.) Occupation R.9. Does your husband (or head of house) work at more than one job? Yes No If YES, what is the other job? R.10. Do any members of your family or your husband's family live here with you? Yes No lf YES, what is the relationship? 292 ’R.ll. 'How many members of your family or your husband‘s family who live here work? what do they do? g9; . RELATIONSHIP R.l2. Would you please tell me the highest grade in school that you and your husband have completed? (Encircle grade) YOU_IWIFE) . Grade School I 2 3 h S 6 7 8 High School l 2 3 A College l 2 3 h 5 6 or more HUSBAND Grade School l 2 3 h S 6 '7 8 High School I 2 3 A College l 2 3 h S 6 or more er3. Which local organizations or clubs do you belong to? '_.—— R.lh. Do you belong to any national organizations or clubs? 'Yes No Which ones? R.lS. R.l6. R.l7. R.l8. R.l9. . R.20. R.2l. R.22. 293 What is your religious preference? Catholic Jewish Protestant If Protestant, ask: What is your church preference? Do you own a car? Yes No if YES, what Year , Make Model 00 you own more than one car? Yes No How many? How long have you lived in this community? years How long has your husband lived in this community? years How many other cities have you lived in during the last ten (l0) years? l 2 3________llor more For statistical purposes, would you please indicate from the following categories the total household income of all families who reside here? Under $l,000 $l,000 to 2,999 3,000 to h,999 5.000 to 5,999 6,000 to 6,999 7.000 to 9.999 l0,000 to lh,999 15,000 and over IHHI 03V O'QUT :WN—A o Is your age: under 20 2| to 30 3| to ho hi to 50 over 50 w I“ m m m