MSU LIBRARIES m w RETURNING MATERIALS: P1ace in book drop to remove this checkout from your record. FINES wi11 be charged if book is returned after the date stamped be10w. Copyright by THOMAS LEE POWERS 1985 ABSTRACT AN EXAMINATION OF THE EFFECTS OF TRADE INCENTIVES ON LOGISTICAL PERFORMANCE IN THE GROCERY PRODUCTS INDUSTRY BY Thomas L. Powers Trade incentives are commonly utilized within distribu- tion channels, however their impact on logistical perfor- mance has not previously been identified. Without this information a determination cannot be made as to the desirability of trade incentive usage. The purpose of this research was to determine those logistical impacts and cost ramifications over a wide range of environmental conditions. The Simulated Product Sales Forecasting Model was utilized to generate information on logistical performance impacts. This was accomplished utilizing a wide range of input variables associated with incentive activity. The major conclusion of the research were: 1. Incentive activity had a substantial positive impact on inventory levels. These inventory levels once increased due to the presence of the incentive were rela- tively insensitive to other input parameters. 2. Incentive activity caused customer service levels to increase overall. These levels were sensitive to other Thomas L. Powers -2 input variables within the incentive scenarios tested. 3. Incentive activity caused the pattern of shipments by size to vary from the non-incentive pattern. A greater number of larger shipments occurred, at the same time that the number of smaller shipments was reduced. Within the incentive scenarios, the pattern was on occasion signifi— cantly different as levels of the other input parameters were varied. 4. The presence of an incentive did not significantly change the total number of shipments. Despite the rela- tively large change in the pattern of shipments, their total remained relatively constant. 5. When cost assignments were made to the measures of logistical performance, a wide range of results occurred. In the instance of an incentive price discount not being passed on to the final customer, the use of an incentive resulted in improved profitability for the distribution center/retail network. When this price discount was passed along, reduced profitability occured in the majority on incentive scenarios tested. The results of this research project would indicate that incentives have substantial impacts on logistical performance, and that in many cases these impacts result in reduced financial performance when a price discount is passed along through the channel and less than optimal xnarket response occurs. ACKNOWLEDGMENTS I wish to express my appreciation to the following members of my dissertation committee who made this project possible: Dr. David J. Closs, who chaired the dissertation committee. Dr. Closs provided a tremendous amount of input both on the overview of the entire project and on the specific aspects of the research design. Dr. M. Bixby Cooper, whose expertise in the area of retailing was instrumental in the conception and execution of this research. Dr. Donald J. Bowersox, whose vast experience in simulation modeling gave a sense of balance and background to the project. I I would also like to express thanks to the following individuals for their intellectual and financial support during my doctoral program: Dr. Charles Hoitash, Eastern Michigan University. Dr. Floyd Patrick, Eastern Michigan University. Dr. Donald Taylor, Michigan State University. ii TABLE OF CONTENTS Page Chapter I IntrOdUCtion O O O O O O O O 0 O O O O O O O O O O O O O O O O O O O O O O O O O O 0 O O 0 O O O 1 BaCkground O I O I O I O O O O O I O O O O O O O O O I 0 I O O O O O O O O ....... O 3 Reasons for the Existance of Trade Incentives ..... Calls for Research ..............L................. \O\l1.\ Market Response to Trade Incentives ............... Research Questions and General Hypotheses ......... 10 Methodology ....................................... 14 Model and Structure Utilized ...................... 15 Justifications and Limitations .................... 19 Footnotes ........................................ 22 Chapter II Review of the Literature .............................. 25 Price Theory Applied to Channels ................. 26 Division of Labor-Channel Structure .............. 31 Importance of Distribution Efficiency ............ 35 Single Echelon Models ............................ 37 Multiple Echelon Models .......................... 39 Market Response to Promotions and Incentives ..... '45 Contributions of the Present Research ............ 51 Footnotes ........................................ 53 Chapter III ResearCh MethOdOI-WY OOOOOOOOOOOOOO0.00000000000000000. 56 IntrOdUCtion .0...OOOOOOOOOOOOOOOOOO0.0.0.0.000... 56 Development of Hypotheses ........................ 56 iii MOdel Development 0.0.0.0....OOOOOOOOOOOOOOOOOOOO. ResearCh mSign 0.0.0.0....OOOOOOOOOOOOOOOOOOOOOO. Model Validation summary OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Footnotes 0 O O O O 0 Chapter IV Results and Conclusions Hypothesis One . Hypothesis Two . Hypothesis Three Hypothesis Four Hypothesis Five Footnotes ...... Chapter V Implications of the Research .......................... RGView Of HYPOtheSis OOOOOOOOOOOOOOOOOOOOOOOOOOOOO Implications of the Research ..................... Limitations of the Research ...................... Future Research Footnotes 0 O O O O O Appendices AppendiXA0.0000000000000000000000.0000000000000. Appendj-XBO......OOOOOOOOOOOOOOOOOOOOO0.0.00.0... Appendixc0.00...00....OOOOOOOOOOOOOOOOOOOOOOOOOO Appendix D ..... BibliograEh! 0.0.00.0...OOOOOOOOOOOOOOOOOOOOOOO0.0.0... iv 61 68 76 77 79 81 81 108 132 149 165 210 211 211 222 226 227 229 230 235 239 244 245 LIST OF TABLES Page Table 2.1 Competitive Market Structure ................ 29 Table 3.1 Testing Environment Variables ............... 69 Table 3.2 Testing Environment Simulation Run combinations 0.0......OOOOOOOOOOOOOOOOO0.0. 73 Table 4.1 Analysis of Variance of Inventory Level by Incentive Level at Low Level of uncertainty OOOOOOOOOOOOOOOOOOOOO... .......... 84 Table 4.2 Analysis of Variance of Inventory Level by Incentive Level at High Level of uncertainty COO...OOOOOOOOOOOOOOOOOOOO0.....0.85 Table 4.3 Analysis of Variance of Inventory Level by 20% Incentive with Different Sales Increases at Low Level of Uncertainty ........33 Table 4.4 Analysis of Variance of Inventory Level by 20% incentive with Different Sales Increases at High Level of Uncertainty ....... 89 Table 4.5 Analysis of Variance of Inventory Level by 40% Incentive with Different Sales Increases at Low Level of Uncertainty ........ 91 Table 4.6 Analysis of Variance of Inventory Level by 40% Incentive with Different Sales Increases at High Level of Uncertainty .......92 Table 4.7 Analysis of Variance of Inventory Level by 20% Incentive with 20% Sales Response Increase with Different Payback Levels at Low Level of Uncertainty ............. ..... 95 Table 4.8 Analysis of Variance of Inventory Level by 20% Incentive with 20% Sales Response Increase with Different Payback Levels at High Levels of Uncertainty ........... ..... 96 Table Table Table Table Table Table Table Table Table Table Table 4.9 Analysis of Variance of Inventory Level by 20% Incentive with Different Payback Levels at Low Level of Uncertainty ......... 4.10 Analysis of Variance of Inventory Level by 20% Incentive with Differnt Payback Levels at High Levels of Uncertainty ....... 4.11 Analysis of Variance of Inventory Level by 40% Incentive with 20% Sales Increase with Different Payback Levels at Low 98 99 Level Of uncertainty 0000000000000 0000000000 101 4.12 Analysis of Variance of Inventory Level by 40% Incentive with 20% Sales Increase with Different Payback Levels at High Level of Uncertainty ....................... 4.13 Analysis of Variance of Inventory Level by 40% Incentive with 40% Incentive with Different Payback Levels at Low Level of Uncertainty ........................... ..... 4.14 Analysis of Variance of Inventory Level by 40% Incentive with 40% Sales Increase with Different Payback Levels at High Level of Uncertainty ............................. 4.15 Testing Environment Output, Hypothesis One, All Testing Conditions . ..... ............... 4.16 Analysis of Variance of Customer Service Level by Incentive Level at Low Level of uncertainty 00......OOOOOOOOOOOOOOOOOOOOOOOOt 4.17 Analysis of Variance of Customer Service Level by Incentive Level at High Level of uncertainty I.OOOOOOOOOOOOOOOOOOOOOOO0.0.0... 4.18 Analysis of Variance of Customer Service Level for 20% Incentive with Different Sales Increases at Low Level of Uncertainty ....... 4.19 Analysis of Variance of Customer Service Level for 20% Incentive with Different Sales Increases at High Level of Uncertainty ...... vi 102 105 106 107 110 111 114 115 Table Table Table Table Table Table Table Table Table Table 4.20 4.21 4.22 4.23 4.24 4.25 4.26 4.27 4.28 4.29 Analysis of Variance of Customer Service Level for 40% Incentive with Increases at Low Level of Uncertainty ....... Analysis of Variance of Different Sales Customer Service Level for 40% Incentive with Different Sales Increases at High Level of Uncertainty .......................... Analysis of Variance of Level by 20% Incentive Response Increase with Levels at Low Level of Analysis of Variance of Level by 20% Incentive Response Increase with Levels at High Level of Uncertainty ........ Analysis of Variance of Level by 20% Incentive Response Increase with Levels at Low Level of Analysis of Variance of Level by 20% Incentive Response Increase with Levels at High Level of Uncertainty ........ Analysis of Variance of Level by 40% Incentive Response Increase with Levels at High Level of Uncertainty Analysis of Variance of Level by 40% Incentive Response Increase with Levels at High Level of Uncertainty ......... Analysis of Variance of Level by 40% Incentive Customer Service with 20% Sales Different Payback Uncertainty ....... Customer Service with 20% Sales Different Payback Customer Service with 40% Sales Different Payback Uncertainty ......... Customer Service with 40% Sales Different Payback Customer Service with 20% Sales Different Payback Customer Service with 20% Sales Different Payback Customer Service with 40% SAles Response Increase at Different Payback Levels at Low Level of Uncertainty ......... Analysis of Variance of Customer Service Level by 40% Incentive with 40% Sales Response Increase at Different Payback Levels at High Level of Uncertainty ....... vii 116 no ooooo 117 121 .122 .123 .124 0.0.0.000127 128 .129 ..130 Table Table Table Table Table Table Table Table Table Table Table Table Table 4.30 4.31 4.32 4.33 4.34 4.35 4.36 4.37 4.38 4.39 4.40 4.41 4.42 Testing Environment Output, Hypothesis Two, All Testing Conditions .......... ....... 133 Shipment Summary for Different Incentive— Levels at Low Level of Uncertainty ..........134 Shipment Summary for Different Incentive Levels at High Level of Uncertainty ......... 135 Shipment Summary for 20% Incentive with Different Sales Increases at Low Level of uncertainty 000......OOOOOOOOOOOOOOOIOOOOO0..137 Shipment Summary for 20% Incentive with Different Sales Increases at High Level of Uncertainty ..............................138 Shipment Summary for 40% Incentive with Different Sales Increases at Low Level of Uncertainty ..............................139 Shipment Summary for 40% Incentive with Different Sales Increases at High Level of Uncertainty ............... . ..... 140 Shipment Summary for 20% Incentive with 20% Sales Response Increase with Different Payback Levels at Low Level of Uncertainty .. 141 Shipment Summary for 20% Incentive with 20% Sales Response Increase with Different Payback Levels at High Level of Uncertainty . 143 Shipment Summary for 20% Incentive with 40% Sales Response Increase with Different Payback Levels at Low Level of Uncertainty .. 144 Shipment Summary for 20% Incentive with 40% Sales Response Increase with Different Payback Levels at High Level of Uncertainty . 145 Shipment Summary for 40% Incentive with 20° Sales Response Increase with Different Payback Levels at Low Level of Uncertainty .. 146 Shipment Summary for 40% Incentive with 20% Sales Increase with Different Payback Levels at High Level of Uncertainty ......... 147 viii Table Table Table Table Table Table Table Table Table Table Table Table Table 4.43 4.44 4.45 4.46 4.47 4.48 4.49 .50 .4.51 4.52 4. 53 4.54 4. 55 Shipment Summary for 40% Incentive with 40% Sales Increase with Different Payback Levels at Low Level of Uncertainty ..... ..... 148 Shipment Summary for 40% Incentive with 40% Sales Increase with Different Payback Levels at High Level of Uncertainty ........ Total Shipments for Different Incentive Levels at Low Level of Uncertainty .. ........ Total Shipments for Different Incentive Levels at High Level of Uncertainty .. ...... . 152 Total Shipments for 20% Incentive with Differnet Sales Increases at Low Level of Uncertainty ....... Total Shipments for 20% Incentive with Different Sales Increases at High Level of uncertainty OOOOOOOOOOOOOOOOOOOO00...... ..... 154 Total Shipments for 40% Incentive with Different Sales Increases at Low Level Of uncertainty OOOOOOOOOOOOOOOOOIOOOOO ....... 155 Total Shipments for 40% Incentive with Different Sales Increases at High Level of Uncertainty .. Total Shipments for 20% Incentive with 20% Sales Increase with Different Payback Levels at Low Level of Uncertainty .......... 157 Total Shipments for 20% Incentive with 20% Sales Response Increase with Different Payback Levels at High Level of Uncertainty . 158 Total Shipments for 20% Incentive with 40% Sales Response Increase with Different Payback Levels at Low Level of Uncertainty .. 159 for 20% Incentive with 40% Increase with Different at High Level of Uncertainty.. Total Shipments Sales Response Payback Levels 160 for 40% Incentive with 20% Increase with Different at Low Level of Uncertainty... Total Shipments Sales Response Payback Levels 161 ix Table Table Table Table Table Table Table Table Table Table 4.56 4.57 4.58 4.59 4.60 4.62 4.63 4.64 4.65 for 40% Incentive with 20% Increase with Different at High Level of Uncertainty. 162 Total Shipments Sales Response Payback Levels for 40% Incentive with 40% Increase with Different at Low Level of Uncertainty...163 Total shipments Sales Response Payback Levels for 40% Incentive with 40% Increase with Different at High Level of Uncertainty..164 Total Shipments Sales Response Payback Levels Distribution Cost Assignments ................l66 Analysis of Variance of Financial Performance by Incentive Level at Low Level of Uncertainty (Option 1) ....... 170 Analysis of Variance of Financial Performance by Incentive Level at Low Level of Uncertainty (Option 2)... 171 Analysis of Variance of Financial Performance by Incentive Level at High Level of Uncertainty (Option 1) ....l72 Analysis of Variance of Financial Performance by Incentive Level at High Level of Uncertainty (Option 2) .......l73 Analysis of Variance of Financial Performance for 20% Incentive with Different Sales Response Increases at Low Level of Uncertainty (Option 1) .....l76 Analysis of Variance of Financial Performance for 20% Incentive with Different Sales Response Increases at Low Level of Uncertainty (Option 2) ........l77 Table 4.66 Analysis of Variance of Financial Performance for 20% Incentive with Different Sales Response Increases at High Level of Uncertainty (Option 1) 178 Table Table Table Table Table Table Table Table Table 4.71 4.72 4.74 Analysis of Variance of Financial Performance for 20% Incentive with Different Sales Response Increases at High Level of Uncertainty (Option 2) ...... Analysis of Variance of Financial Performance for 40% Incentive with Different Sales Responses at Low Level of Uncertainty (Option 1) ................. Analysis of Variance of Financial Performance for 40% Incentive with Different Sales Responses at Low Level of Uncertainty (Option 2) ................. Analysis of Variance of Financial Performance for 40% Incentive with Different Sales Responses at High Level of Uncertainty (Option 1) ................. Analysis of Variance of Financial Performance for 40% Incentive with Different Sales Responses at High Level of Uncertainty (Option 2) ................. Analysis of Variance of Financial Performance for 20% Incentive with 20% Sales Response Increase with Different Payback Levels at Low Level of Uncertainty (Option 1) Analysis of Variance of Financial Performance for 20% Incentive with 20% Sales Response Increase with Different Payback Levels at Low Level of Uncertainty (Option 2) Analysis of Variance of Financial Performance for 20% Incentive with 20% Sales Response Increase with Different Payback Levels at High Level of Uncertainty (Option 1) Analysis of Variance of Financial Performance for 20% Incentive with 20% Sales Response Increase with Different Payback Levels at High Level of Uncertainty (Option 2) xi 179 182 183 184 185 189 190 191 192 Table Table Table Table Table Table Table Table Table 4.76 4.77 4.78 4.79 4.80 4.81 4.83 4.84 Analysis of Variance of Financial Performance for 20% Incentive with 40% Sales Response Increased with Different Payback Levels at Low Level of uncertainty (Option 1) Analysis of Variance of Financial Performance for 20% Incentive with 40% Sales Response Increased with Different Payback Levels at Low Level of Uncertainty (Option 2) Analysis of Variance of Financial Performance for 20% Incentive with 40% Sales Response Increased with Different Payback Levels at High Level of Uncertainty (Option 1) Analysis of Variance of Financial Performance for 20% Incentive with 40% Sales Response Increased with Different Payback Levels at High Level of Uncertainty (Option 2) Analysis of Variance of Financial Performance for 40% Incentive with 20% Sales Increase with Different Payback Levels at Low Level of Uncertainty (Option 1) Analysis of Variance of Financial Performance for 40% Incentive with 20% Sales Increase with Different Payback Levels at Low Level of Uncertainty (Option 2) Analysis of Variance of Financial Performance for 40% Incentive with 20% Sales Increase with Different Payback Levels at High Level of Uncertainty (Option 1) Analysis of Variance of Financial Performance for 40% Incentive with 20% Sales Increase with Different Payback Levels at High Level of Uncertainty (Option 2) Analysis of Variance of Financial Performance for 40% Incentive with 40% Sales Increase with Different Payback Levels at Low Level of Uncertainty (Option 1) xii 193 194 195 196 198 199 200 201 204 Table Table Table Table Table Table 4.85 Analysis of Variance of Financial Performance for 40% Incentive with 40% Sales Increase with Different Payback Levels at Low Level of Uncertainty (Option 2) ................................ 4.86 Analysis of Variance of Financial Performance for 40% Incentive with 40% Sales Increase with Different Payback Levels at High Level of Uncertainty (Option 1) ................................ 4.87 Analysis of Variance of Financial Performance for 40% Incentive with 40% Sales Increase with Different Payback Levels at High Level of Uncertainty (Option 2) ................................ 4.88 Testing Environment Output .................. 5.1 Financial Performance Summary Percent Change in Performance from Non-incentive scenario 0.0...OOOOOOOOOOOOOOOOOOOO0.0...O 5.2 Managerial Guidelines Summary ......... ..... xiii 205 206 207 209 220 223 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 1.1 2.2 2.3 4.2 LIST OF FIGURES Relationship Between Manufacturer and Retailer/Distrubtion Center Cost and Revenue Curves Trade-Off Limits: Contribution from % Change in Sales Volume vs. Carrying Cost of l-Week's Inventory .............. Response of a Production-Distribution System to a 10% Unexpected Rise and Fall in Retail Sales Over a One-Year Period Comparison of Inventory Levels for Incentive and Non—Incentive Testing Conditions Comparison of Inventory Levels for 20% SPSF Testing Environment .................. Causal Loop Structure ..................... Order Point and Dependent Demand .......... Channel System Configuration .............. Incentive Timing and Response ............. Incentive Sales Payback ................... Incentive with Different Sales Response Levels xiv Page 16 34 41 43 58 59 63 65 66 82 86 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 4.12 4.13 Comparison of Inventory Level for 40% Incentive with Different Sales Response LEVQIS oooooooooooooooooooooooooooooooo... 90 Comparison of Inventory Level for 20% Incentive with 20% Sales Response Increase with Different Payback Levels ............93 Comparison of Inventory Level for 20% Incentive with 40% Sales Response Increase with Different Payback Levels ........... 97 Comparison of Inventory Level for 40% Incentive with 20% Sales Increase with Different Payback Levels ................1oo Comparison of Inventory Level for 40% Incentive with 40% Sales Increase with Different Payback Levels ................104 Comparison of Customer Service Levels for Incentive and Non-Incentive Testing conditions ......OOOOOOOOOIO ..... 109 Comparison of Customer Service Level for 20% Incentive with Different Sales Response Levels ......OIOOOOOOOOOOOOOOII.113 Comparison of Customer Service Level for 40% Incentive with Different Sales Response Levels ......................... 118 Comparison of Customer Service Level for 20% Incentive with 20% Sales Response Increase with Different Payback Levels .. 119 Comparison of Customer Service Level for 20% Incentive with 40% Sales Response Increase with Differnt Payback Levels ... 125 Comparison of Customer Service Level for 40% Incentive with 20% Sales Increase with Different Payback Levels ........... 126 XV Figure Figure Figure Figure Figure Figure Figure Figure 4.14 4.15 4.16 4.17 4.18 4.19 4.20 4.21 Comparison of Customer Service Level for 40% Incentive with 40% Sales Increase with Different Payback Levels .......... Comparison of Financial Performance for Incentive and Non—Incentive Testing. conditions 00...........OOOOOOOOOOOOOOOO Comparison of Financial Performance for 20% Incentive with Different Sales Response Levels ......OOOOOOOOOOOIOOOOOOO Comparison of Financial Performance for 40% Incentive with Different Sales 131 169 175 Response Levels ......IOOOOOOOOOOOOOOOOOO181 Comparison of Financial Performance for 20% Incentive with 20% Sales Response Increase with Different Payback Levels Comparison of Financial Performance for 20% Incentive with 40% Sales Response Increase with Different Payback Levels Comparison of Financial Performance for 40% Incentive with 20% Sales Response Increase with Different Payback Levels Comparison of Financial Performance for 40% Incentive with 40% Sales Response Increase with Different Payback Levels xvi .186 .188 .197 .203 CHAPTER I ’ INTRODUCTION The objective of this dissertation is to examine the distribution channel performance impacts of industry trade incentives. Trade incentives affect many facets of a dis- tribution channel resulting in measurable impacts on cost and customer service. Past research has not investigated the channel performance impacts of incentives resulting in a fertile area of study to develop guidelines for the effec- tive use of trade incentives. The dollar magnitude of trade incentives has increased 1 and at the same time their 2 significantly in the recent past percentages of total marketing budgets has increased. There has not been a similar increase in the understanding 3 of the overall management of incentives although such a development has been called for in the literature.4 Trade incentives exist for a number of reasons, includ- ing the ability to shift the demand curve in the short 5 transfer inventory to lower points in the channel,6 term, and to gain economies of scale in production and distribu- tion processes. These and other reasons are believed by various channel members to increase their firm's financial performance. 2 This perceived financial gain may be for the channel in total, or suboptimally for an individual channel member. These relationships and impacts have not been empirically proven for a total channel, although they have been viewed in the context of a single echelon, the retail level. At the same time that the need exists for clarification of this issue, methodologies have been developed that make this type of research possible. Dynamic simulation 8 with appropriate experimental techniques such as 9 modeling, factorial designs utilizing analysis of variance, can be used as a testing environment to evaluate overall channel performance. The present research is designed to utilize the above methodology to examine the effects of trade incentives on distribution channel performance. In addition it is intended to develop guidelines that reflect tested relation- ships between various parameters and conditions that result in improved channel performance. The guidelines are intend- ed to provide the discipline with an increased understanding of channel operations under trade incentives. The guide- lines may also provide practitioners with a means of deter- mining how and under what conditions to implement trade incentives in consideration of channel performance impacts. This chapter outlines the background of the use and impact of trade incentives within a channel of distribution, the research questions and hypothesis to be examined in the 3 research, the research methodology to be utilized, and the justification and limitations of the research. Background Trade incentives are a commonplace activity within distribution channels and have grown at a rapid rate in recent years. Trade incentives are defined as activities including but not limited to seasonal, quantity, and case discounts. These incentives may appear at any level within a channel of distribution, and may or may not result in an. incentive in the form of a price discount and aimed directly 10 The growth in the use of, and at the ultimate consumer. the expenditures on incentives has increased a need to further understand the total ramification of incentive usage. Trade incentives impact not only sales responses at various levels within the channel, but also directly impact the performance of that channel. Performance is defined as the relationship between customer service capability and 11 The status of research on the associated total cost. subject reflects the former issue, however, not the latter. In as much as incentives impact sales response and logis- tical activities, the problem must be examined from both perspectives, as they are highly interrelated. Before this can be accomplished, it becomes necessary to understand the rationales for the existence of incentives, the causes of their increased use in recent years, and the logic of why the issue of performance measurement is seen to be increas- ingly important. In addition, it is important to understand 4 the impact that incentives have (M) the consumer market in terms of sales response. Reasons for the Existance of Trade Incentives Trade incentives exist for two primary reasons: one market related, one production related. The market rationale for incentives is usually associated with a short 12 In addition to term increase in sales and market share. this primary rationale, there are further objectives that are related. These include modifying the distribution pattern by increasing the number of members at a given level within the channel. Further objectives may be consumer related, such as acquiring new customers or accelerating the rate of product usage. Finally, within market related rationales, there may be objectives related to competition, such as increasing market share or responding to compe- titors' promotions.13 In addition to the market rationale, there exists a number of reasons that are based within the production activity of the firm. Of these, the most prevalent rationale is that of utilizing incentives to move merchan- dise down the channel thereby shifting inventory carrying costs down the channel. Implicit in this rationale is the assumption that there exists a mismatch between supply and demand at a given point in time. This may occur due to seasonal demand fluctuations that are being anticipated by a higher than normal production level, or batch type produc- tion with high set up costs that make large production runs necessary to minimize per unit costs. 5 A number of reasons exist for the dramatic increases seen in the use of, and expenditures on, promotion and trade incentives. These include both causes internal and external to the firm. Internal causes observed include:15 o Promotions have become more acceptable to senior management. 0 More executives are better qualified to handle promotion activities. 0 The use of the product manager system, with individuals desiring a quick return on invest- ment, has caused an increased emphasis on promotion due to its short term payoff. External causes cited include:16 o The number of brands has increased; as new brands typically use promotion at the time of introduction, the number of brands directly increases the amount of promotion. 0 The increased number of brands causes an increase in promotion as they compete for limited amounts of shelf space. 0 The increased number of brands also may result in shorter product life cycles. Promotions may be used to prevent or delay the onset of the decline stage. 0 Competitors are becoming more promotion minded, therefore, firms are forced to become part of a scenario of increased promotional activity for competitive reasons. 0 General economic conditions have declined, causing a need to engage in activities that will result in increased demand. An example of the last point is that of the tremendous increase in rebates to consumers for products such as auto- lnobiles during the period of 1973-1975. This rebate was a fonm of trade incentive that bypassed the dealer to ensure compliance. Prior to that period incentives within this .- 2‘. r-“- 6 industry had been given to retail dealers without specific performance requirements as to passing on price reductions. The period 1979-1982 also saw an increased use in this type of activity, however, for lower-priced, convenience-type merchandise. The above internal and external reasons for the increased use of promotions represent an overview of the factors causing this phenomena to occur. In addition, the reasons cited will undoubtedly continue to increase and continue to place further pressure on the development and use of promotions and incentives. This past set of circumstances that has caused increases in promotional activity has also resulted in substantial changes in the overall structure of marketing expenditures. Between 1969 and 1975 promotion expenditures 18 According increased at a rate twice that of advertising. to a survey of sixty-five packaged goods companies in 1980, their marketing budgets devoted 34.9% of total dollars spent 19 to trade promotion and 24.7% to consumer promotion. The total of these two categories exceeds the amount spent on consumer advertising. Another study has shown for grocery products and health and beauty aids that between sixty and seventy-five percent of goods in these categories were 20 The total shipped on some form of an incentive basis. expenditure for these categories on promotion and incentives was shown to have exceeded 9.8 billion dollars annually. 7 Although the above figures represent dramatic evidence of the widespread use of promotions and trade incentives, it must be kept in mind that these figures are, if anything, understated. Many times the expense of an incentive, in the form of a price discount or allowance, is not reflected in the financial statements of a company. The top line may reflect sales net of any discount or allowance. Thus the issue of incentive cost is ignored or taken for granted, much in the same way that freight costs have been tradi- tionally treated by many firms.21 Trade incentives and promotional activity have been discussed simultaneously although a difference does exist. Trade incentives have as their focus selling activities to other channel members whereas promotional activity has as its focus selling activities to the final consumer. In many cases, the activity aimed at the consumer has as its origin a trade incentive with a specific performance requirement. A description of various types of trade incentives appears in Appendix A. Calls for Research The tremendous increase in promotion and trade incentive activity has created a need for further under- standing of the cost effectiveness of these activities. As stated by one author, trade promotions are where advertising was twenty years ago, “large, rapidly growing budgets with no real measurement, planning, or management tools to apply .22 these budgets. Another perspective on the problem has h23 been brought forth by Baug who states: 8 Unless it is documented that trade promotions are unprofitable, expenditures will continue to increase faster than sales because of continuous competitive pressure because trade promotions increase sales in the short run and because the sales department will feel the pressure from the trade for higher discounts. Other authors have stated that improved measurement, planning and management can result in improved promotion effectiveness, and that these techniques exist or can be 24 It developed. is also believed that improvements in effectiveness will take years, however, even small efficiency improvements will impact profits dramatically.25 ’In addition, it has been observed that there are wide differences between promotional patterns, problems, and opportunities for various companies and industries, although common measurement and management tools can be utilized.26 It has been suggested that management information systems be modified to identify promotion costs. These include changes to identify costs associated with special packaging, consumer education programs, sales and dealer incentives, and loss of revenue from temporary price 27 From a total strategic perspective, reduction programs. it has been observed that these activities may result in improved promotion management by analyzing past spending, establishing objectives, selecting appropriate techniques, and pretesting.28 The rapid growth in the use of trade incentives and the costs associated with incentives has created a very real need to understand the impacts that these activities have on 9 the performance of a distribution channel. The purpose of the present research is to determine that relationship. The retail sales impact of promotions, within the context of trade incentives, has been examined in previous research and provides a background that can be useful in understanding the nature of market response and for establishing para- meters for testing the model. Market Response to Trade Incentives Market response as it relates to incentives can be separated into two areas. The first area is the response of channel members to incentive activities conducted by other channel members. The second area is that of final consumer response. The literature on intermediary response is extremely limited, however, there is a wide body of litera- ture on end-level, consumer response to incentive activi— ties. Consumer response to promotion and incentive type activities has been examined in the literature from a number of different perspectives. These include the reasons that consumers have for accepting a price deal, which has been seen to be based on a relatively lower cost of holding inventory or the portion of cost that is associated with 29 Associated with this finding is that the source storage. of the sales response in: a promotional deal may be a stockpiling activity on the part of the consumer and not brand switching activity. In comparing the responses to promotions and advertising, it has been observed that promotions yield faster results than advertising, however, 10 30 Advertising, they do not produce new, long-term buyers. on the other hand, has been seen as being able to increase long-term sales. Additional studies have examined market response to promotion in the context of market structure and product 31 32 33 life cycle, brand loyalty, pricing, and package size.34 In summary, promotion and trade incentive activities are widely used for consumer goods. Their relative use compared to other marketing activities is growing at a rapid rate. Their cost effectiveness has been questioned, and remains unanswered. in) answer this question, two primary areas must be examined. These are the response to this type of activity in the marketplace and the impact of this activity upon the channel. The present investigation is designed to utilize assumptions of the former to develop empirical findings of the latter. The operational focus of the research is (”1 trade incentives, however, promotional activities aimed at the consumer are implicit with these trade incentives. Researchyguestions and General Hypotheses The overall nature of the problem examined is the impact of trade incentives on logistical channel perfor- mance. Impact may be considered to include all associated changes on the channel of distribution that are caused by the existence of the incentive activity. These associated changes include but are not limited to, changes in inventory 11 levels at various points in the channel, changes in the flow of merchandise through the channel, changes in the cost and type of transportation utilized, changes in material handl- ing procedures, changes in the flow of information and changes in margin structure through the channel. These associated changes may be thought of as the variations that occur on the activities related to the product that is involved in the incentive under examination. In addition, incentive activities associated with one product may have impacts of the above nature on other products that may ultimately affect product and customer mix in the long term. Channel performance may be considered to be all elements of cost and customer service capability. Costs can be measured on many dimensions including inventory, trans- portation, material handling, communications, and overall administration costs. Customer service capability can be measured on dimensions including order cycle time, back- orders, fill rate, and incorrect orders shipped. As with impacts, these measures of performance may be considered not only for the product that is under an incentive, but for all products and customers of the firm. The essential problem relates to the overall efficiency of the system, and the effect that incentive activities have on that efficiency. This problem of efficiency has been 35 in their discussion of seen by House and Karrenbauer logistical models as follows: The basic problem under investigation is deceiv- ingly simple. If we have m productive facilities 12 P ...P and n consumption points, C ...C r then we can deufine the role of a logistic sysrtem to be that of providing the linkage between the pro- ductive facilities and the consuming units. A more precise statement of the problem emphasizes the determination of the needs of the n consumers in terms of product and service, the determination of the product/service delivery capabilities of the m producers, and the matching of production and consumption units with the most economical system which satisfies customer service con- straints. This issue of efficiency or ”economical system" as described by House and Karrenbauer is also reflected in the following statement by Bowersox:36 Physical distribution is an integrative field which combines transportation, inventory, ware- house operations, order communication, and material handling as components of a single performance system. The structure of the system is determined by the number, size, and geographi- cal arrangement of warehouse facilities that serve as nodal points in the integration of the above components. The planning objective in physical distribution is to isolate the operating system structure and procedures which most satisfactorily contribute to the firm's profit goals. This statement expands the idea of ”economical system" to the notion of ”operating system structure and procedures." In addition, the concept of tradeoff between production/ service capabilities and customer requirements is expanded to be ”the firm's profit goals.” From this it can be seen that the nature of the problem can be expanded to include not only the impact of incentives upon logistical channel performance in terms of costs and customer service capabilities, but also their combined impact on the overall profitability of the firm and channel. 13 To reflect the problem as it has been developed, the general question is stated as follows: What is the impact of levels of variables associ- ated with a trade incentive on measurable outputs of logistical performance for a channel of dis- tribution including overall financial performance? From 'this, the following research. questions reflect individual specific areas of the first part of the above question: What effect does a trade incentive have on: - inventory levels? - customer service levels? - shipment sizes? - number of shipments? These questions reflect the impact of incentives on the logistical performance of a channel. Given information relative to the customer service trade-offs and requirements for a specific firm, the changes in revenue due to the incentive, and the sales response due to the incentive, the impact on the overall profitability of the firm and channel can also be identified. The above research questions are restated in hypothesis form as follows: H The use of an incentive increases inventory levels for the product under the incentive at the distri- bution center. H2 The use of an incentive decreases the overall customer service level for products under the incentive. H3 The use of an incentive results in variation in shipment sizes between the manufacturer and the distribution center. H4 The use of an incentive results in an increase in the total number of shipments from the manufacturer to the distribution center. 1 Finally, to reflect the impact of trade incentives on the profitability of the distribution center/retail network, 14 given the above impacts plus cost and volume changes, the last hypothesis states: H5 The use of an incentive results in a decreased‘ level of profitability for the distribution center/retail network. The following section_ describes the research design utilized to test the hypotheses and to develop guidelines for the cost ramifications of trade incentives on channel performance. Methodology In order to test for the impacts of trade incentives on logistical activities of a firm and channel, the research design must be capable of evaluating logistical system performance under conditions of incentive and non-incentive activity. In addition, in order to develop guidelines for a broad range of conditions, the design must evaluate results from different combinations of alternative conditions. Specifically, the design must be capable of the following requirements: 0 Testing a logistical system under conditions of an incentive, and making comparisons with that same system under non-incentive conditions. 0 Evaluating specific operational impacts of changes in levels of parameters associated with an incentive activity, such as the level of sales response, and the payback in sales at the conclusion of the incentive period. 0 Replicate a multi-echelon distribution system, with multiple members in the channel at the distribution center and retail level. 0 Replicate a multi-product system in order to evaluate comparisons between products under incentive activities with products under normal conditions. 15 0 Account for dynamic elements of a logistical operating system, including the interaction of temporal and spacial considerations. 0 Replicate reality as closely as possible. Dynamic simulation models represent a methodology that is capable of achieving the above objectives. Models that have been previously been designed, tested and validated at Michigan State University include the LREPS (Long-Range )37 model and SPSF Environmental Planning Simulator (Simulated Product Sales Forecasting)38 model. The SPSF model, which was developed after the LREPS model, combines attributes of dynamic simulation, market area demand simulation, and statistical sales forecasting. Model and Structure Utilized The present research utilized a modified SPSF model that was developed in order to allow for the manipulation of variables associated with trade incentive activity. This model was operable on an IBM Personal Computer. Data that was generated in each simulation run was collected into a data file stored on diskettes. This data file was then transferred to the Michigan State University Cyber mainframe system for analysis. The general design of the SPSF testing environment appears in Figure 1.1 below: 16 Figure 1.1 OPIRATIONS MODULE PRODUCIS PRODUCTS ABC DE HANUFACTURERS DlSTRlBUTION ABCDE CENTERS ———————————— -I>—-————-. I l /?/J\ | I I ABCDE Incas ABCDE REIAILERS : : I i I —}—J—db-—-.--1l ! I I I _ 5 .l l - I l L_ ' ‘ I l . . .. IMHO; : _ mmm m3 .1. . ' ' MSE ‘ I I I I , I . I ; FE‘ ' g I swam mam ~«»----4 I : nnmu mus - -------- 1 F““““' : ' ' I | ' mm mm :': I : I g I I l i I I wmn : I nuns , | 1 _____ NH"_ flmwfll : I ‘1 mun I h—-—-—-—-——J : L--_--_--_---_-_..----_ TOW .----- .................. ..l mm A 55mm: Source: Simulated Product Sales Forecasting. D.J. Bowersox, D.J. Closs, J.T. Mentzer, J.R. Sims. MSU Business Studies, 1979. In the context of the SPSF testing environment, the structure of the model and the factors examined are as follows: Echelon Structure: The echelon structure repre- sents a grocery products distribution system. This includes a manufacturer, five distribution centers, each with five retail locations. 17 Product and Product Mix: The products tested in the model consist of a mix of ten identical products. Five of the products are subject to trade incentives: the remaining five are on a non-incentive basis. The products represent typical non-perishable grocery items. - Transymrtation Method: The transportation method utilized consists of plant inventory location to distribution centers being shipped by common carrier and rail. Frmm distribution centers to retail locations, shipments are made by private fleet on a regularly scheduled basis. Inventory Level: Initial inventory levels are established at a level of 12 days supply. Inventory Reorder Logic: The reorder logic utili- zed in the model is a reorder point and quantity method. Reorder points and quantities are based on a level that would result in a customer service level 90-95% with a moderate degree of demand uncertainty and 80-85% with a high level of uncertainty. In the case of a sales incentive, the entire incentive amount for the incentive period became the minimum order quantity on the first day of the incentive. Customer Service Levels: Initial customer service levels were established for the testing environ- ment in conjunction with the inventory reorder logic during a non-incentive simulation scenerio. These levels became the initial performance standards from which deviations were noted. Time Period: The time period utilized in the testing environment consists of a three year simulation run. Each year contains two incen- tives, each two months in length. Sales Incentive—Increase: Incentives were tested in three scenarios. The first is a non—incentive scenario, whereby no incentive is in effect. The second scenarho is that of an increase in ship- ments during each incentive period of 20% above the normal sales level. The third scenario is that of an increased level of shipments at 40% above the normal sales level. In each case, the level of sales increase is tested independently of the incentive increase. 18 Sales Response-Increase: The sales response at the customer level is tested at three levels. The first is a zero level response. The second is a sales response of 20% above the normal sales level. The third sales response increase is 40% above the normal sales level. _ Sales Response-Uncertainty: Sales response uncertainty is examined in two scenarios. The first is a relatively low level of uncertainty. The second scenario is that of a relatively high level of uncertainty. Sales Response-Payback: Sales payback consisting of a decrease in sales after the incentive period is tested in three scenarios. The first is a zero payback at the conclusion of each incentive, i.e. sales returned to the level prior to the incen- tive. In the second scenario, a sales payback occurs equal to one-half of the sales increase during the incentive. The third scenario con- sisted of a payback amount equal to the entire increase during the incentive period. In each payback scenario, the pattern remains constant and is equal in time to the incentive period. The model structure as outlined above allows com- parisons of incentive versus non-incentive scenarios in addition to examining the effects of critical variables associated with a trade incentive activity. Based on this structure, a summary of the requisite simulation runs is: Factor Runs Required Incentive Level 3 Sales Response-Increase 3 Sales Response-Uncertainty 2 Sales Response-Payback 3 Total runs required: 26 .A total of twenty-six runs is required to examine all combinations of relevant factors. Twenty four runs depict l9 incentive scenarios whereas two runs depict non-incentive scenarios. In addition to the 26 simulation runs required to test all possible combinations, additional cost and revenue measures are assigned to each measure of output for the testing of hypothesis five. The SPSF Testing Environment in combination with the research design provides a means to test the research hypotheses and to provide additional information relative to the impacts of different levels of parameters that are associated with logistical system performance. Inputs to the model that are independent variables associated with incentive activity generated in the Forecast and Demand Modules. The operating structure of the logistical system is defined in the Operations Module. The Analysis Module provides detailed output on logistical performance as impacted by the various combinations of parameter levels. This includes information about sales, stockouts, customer service levels, inventory, replenishment, and shipping volumes. Justification and Limitations This research provides a meaningful contribution from a number of perspectives. First, it addresses an area that has not been examined previously. This area is important due to the widespread acceptance and use of trade promotions and the cost ramifications of associated logistical acti- vities on the firm and channel financial performance. The effects of trade promotions on logistical activity have not 20 been known. Without this knowledge, it is impossible to determine if trade promotions produce a positive benefit to the institutions involved. This research is designed to make this determination. In addition, the research provides additional insight into logistical performance over a wide range of operating conditions that may be of immediate use and also may be the basis of research_beyond this disserta- tion. The present research has implications for the discip- line and practitioners. The research design, having in addition to the capability to answer the hypotheses developed for testing, provides a set of guidelines that illustrate the impacts of incentive activities for a wide range of operational considerations. These include various levels of market response to those incentives. These guidelines can be used to assess the feasibility of a given incentive activity prior to its inception. This knowledge provides substantial cost savings potential for a firm and channel. Although the approach utilized in this research is designed to produce results that are as broadly applicable as possible, certain limitations exist. The findings of the research are applicable only to channel structure and product environments that are similar to the one tested in this study. In addition, the findings from testing various levels of parameters within the model structure are depen- dent upon those parameters and levels. 21 The chapter which follows examines the relevant litera- ture relating to the theoretical base of the study, previous examinatitni of this and related problems, and literature dealing with sales response to trade incentive and promo- tional activities. Chapter II examines the literature related to the present research. Chapter III outline the research methodology and hypotheses tested. Chapter IV details the results and conclusions of the testing environment. Finally, Chapter V discusses the implications of the research. 22 CHAPTER I -- FOOTNOTES 1Strang, Roger A., ”Sales Promotion-Fast Growth, Faulty Management,” Harvard Business Review, 54:4 (July—August 1976), pp. 115:124. 2Donnelly Marketing, Third Annual Survey of Promotion Practices (July 1981). 3Strang. 4Strang. 5Cotton, B.C. and E.M. Babb, "Consumer Response to Promotional Deals,” Journal of Marketing, 42 (July 1978), pp. 109-113. 6Chevalier, Michel and Ronald C. Curhan, "Temporary Promotions as a Function of Trade Deals: Descriptive Analysis,” Working Paper, Marketing Science Institute, Cambridge, Mass., May 1975. 7Corstjens, Marcel and Peter Doyle, ”A Model for Optimizing Retail Space Allocations,” Management Science, 27:7 (July 1981): pp. 822-833. 8Bowersox, Donald J., ”Planning Physical Distribution Operations With Dynamic Simulation” Journal of Marketing, Vol. 36 (January 1972), pp. 17-25. 9Bowersox, Donald J. et. a1. Dynamic Simulation of Physical Distribution Systems, (1972): pp. 199, 202-204. 10Chevalier and Curhan. 11Bowersox, Donald J., Logistical Managgment, Macmillan Publishing Co., Inc. New York (1978), pp. 19-21. 12Blattberg, Robert 3., Gary D. Eppen, and Joshua Lieberman, ”A Thoretical and Empirical Evaluation of Price Deals for Consumer Nondurables,‘I Journal of Marketing 45 (Winter 1981): pp. 116-120. 13Quelch, John A., ”Trade Promotion by Grocery Products Manufacturers: A Managerial Perspective,” Marketing Science Institute Research Program, (August 1982), pp. 5. 23 14Bliemel, Friedhelm, ”Inventory Decisions by Trade-Off Analysis: A New Approach in Product-Oriented Marketing Strategies," Journal of Business Logistics, Vol. 1, Number 2. 15Strang. 16Strang. 17The author was employed by Ford Motor Company at the time administering rebates and incentives. 18Strang. 19 . Donnelley Marketing. 20Kingsbury, Steve, ”Industry Trends in Promotion," presentation to SPAR Annual Seminar, Garden Island, Vermont, June 23, 1980. 21Bowersox, Logistical Managgment, pp. 255. 22lemont, F.L., ”Room at the Top in Promotions," Advertising Age, March 23, 1981. 23Haugh, Louis J., ”Measuring Promotions Grows in Importance," Advertising Age, October 1, 1979, pp. 58-59. 24Lemont. 25Lemont. 26Lemont. 27Strang. 28Strang. 29Blattberg, et. a1. 30Brown, Robert 6., "Sales Response to Promotions and Advertising,” Journal of AdvertisingiResearch, 14:4 (August 1974), pp. 33-39. 31Chevalier, Michel, "Increase in Sales Due to In-Store Display," Journal of MarketinggResearch, 12:4 (November 1975), pp. 426-431. 32Massy, William F. and Ronald E. Frank, ”Short-Term Pricing and Dealing Effects in Selected Market Segments,” Journal of Marketing Research, 11:2 (May 1965), pp. 171-185. k 24 Also see McCann, J.M., ”Market Segment Response to Marketing Decision Variables,” Journal of Marketing Research, 11 (November 1974), pp. 399-412: Webster, Frederick E., Jr., "The Deal-Prone Consumer,“ Journal of Marketing Research, 2 (May 1965), pp. 186-189; Brown, op cit. 33Woodside, Arch G. and Gerald L. Waddle, ”Sales Effects of In-Store Advertising," Journal of Advertising Research, 15:3 (June 1975). Also see Cotton and Babb, op cit. 34Massy and Frank. 35House, Robert G. and Jeffrey J. Karrenbauer, "Logistics System Modelling," International Journal of Physical Distribution and Materials Management. Vol. 8 Number I: 36Bowersox, ”Planning Physical Distribution Operations With Dynamic Simulation.” 37Bowersox, ”Planning Physical Distribution Operations With Dynamic Simulation.” 38Bowersox, et a1., pp. 29-30. CHAPTER II REVIEW OF THE LITERATURE This chapter examines the theoretical base of the present research and the applicable literature. The theoretical base is examined from the context of price theory as applied to channels and from the concept of division of labor as it relates to the cost structures of institutions within a channel. Price theory enables the reader to understand the pricing logic and competitive environment of the channel structure under study. The examination of division of labor notion provides a framework whereby the relative cost structures for various channel members can be developed and envisioned. Relevant literature is examined in four areas. First, a review is made of the historical and current calls for distribution efficiency. In order to answer that question as it relates to trade incentives, a review is made of the literature addressing single echelon models. Following that a review is made of multiple echelon problems and research designs. In order to understand the market impact of trade incentives and their related promotional activities a review is made of that body of literature. The chapter concludes with a discussion of the contribution of the present 25 26 research based (N) the literature cited, and the questions left unanswered. Price Theory Applied to Channels The nature of competition as it relates to channel structure has been examined by Mallen1 who has suggested the application of price theory to this area. He suggests the use of traditional economic terminology to describe a classification of economic markets. The classification schema proposed is based primarily upon the degree of concentration and number of suppliers on the selling side, and the degree of concentration and number of middlemen on 2 the buying side of the market. The classification schema is based on the following principles: 1. The smaller the number and the greater the degree of concentration on the selling side, the less purely competitive and the more monopolistic that market becomes. 2. The smaller the number and the greater the degree of concentration on the buying side, the less purely competitive and the more monopsonistic that market becomes. 3. Where this diminishing number and increasing concentration are working on both sides of the market, the closer that market moves to bilateral monopoly. His classification of economic markets is as follows: Market Situation Middlemen (buyers) Suppliers (sellers) Pure Competitor Oligopolist Monopolist Pure Competitor Pure Competitor Oligopoly Monopolist Monopolist Pure Competitor Pure Competitor Pure Competitor Oligopsonist Monopsonist Oligopsonist Monopsonist Monopolist Pure Competition Oligopoly Monopoly Oligoposony Monopsony Bilateral Oligopoly Bilateral Monopoly Successive Monopoly 27 In his description of *vertical price relationships, Mallen views the channel member who holds monopoly power as the one who determines the profits for the channel, and who - reaps those profits. For example, he states: Assume the retailer is the monopolist and the others (wholesalers and manufacturers) are pure competitors, as for example, a single department store in an isolated town. Total cost to the retailer are composed of the total cost of the other levels plus his own costs. No pure profits of the other levels are included in his costs, as they make none by definition (they are pure competitors). The retailer would be in the same buying price position, so far as the lack of suppliers' profits are concerned, as would the vertically integrated firm. Thus, he charges the same price as the integrated monopolist and makes the same profits. If the manufacturer were the monopolist and the other channel members pure competitors, he would calculate the maximizing profits for the channel, and then charge the wholesaler his cost plus the total channels' pure profits, all of which would go to him since the others are pure competitors. The wholesaler would take this price, add it on to his own costs, and the result would be the price to retailers. Then the retailer would do likewise for the consumer price. In the context of grocery product channel market structure, an examination can be made of the number of and relative size of channel members. A typical grocery channel consists of a large number of manufacturers that sell to wholesalers or distribution centers that are of a relatively small number. They in turn sell to a large number of 3 In some channel structures, the distribution retailers. center is owned by the same firm that owns and operates the chain of retail supermarkets that are sourced from the 28 distribution center. In others, the distribution center is owned and operated independently of the retail stores, which may be a chain or be independently owned. The manufacturer competes with a large number of other manufacturers for the distribution center's business. This holds true for most product lines where several manufac- turers' brands are available to the distribution center. The level of competition in many cases takes the form of monopolistic competition due to brand loyalty or product differentiation. The distribution center can be seen in this context also as in most markets the number of indepen- dent or vertically integrated centers is relatively small.4 Therefore, the distribution center may be seen from the point of view of the manufacturer as an Oligopsonist, having the ability to influence price. From the perspective of the retailer, the distribution center may be its only source of supply, and thus represent a monopoly to the retailer. In other cases the distribution center may not be the only source of supply, and thus represent an oligopoly situation. Because of this primarily monopolistic relationship between the distribution center and the retailer, the operational difference between an independent channel and a vertically integrated one becomes minimal. The retailer competes with other retailers for the final consumer's business. Most products sold are not differentiated between stores, however, some attempts are made on the part of the retailer to differentiate store 29 brands, meat, and produce items.5. In theory, the grocery retailer is in a purely competitive market, although it may be argued that due to location and other factors, the retailer does in fact have a degree of differentiation, and 6 For the pur- thus is actually a monopolistic competitor. poses here, given similarity of goods and the mobility of most consumers, the individual retailer is considered to be in a state of pure competition. Given the relationship to the distribution center, however, most large markets are increasingly becoming oligopolistic in nature with minimal 7 The retailer, given its relatively price competition. small size and dependence on the distribution center has relatively little influence on price. The market structure as described can be seen in Table 2.1 below: Table 2.1 Competitive Market Structure Channel Member Selling Market Buying Market Manufacturer Monopolistic See * Competition Distribution Center Monopoly/Oligopoly Oligoposony Retailer Pure Competition No control of price *The manufacturer undoubtedly possesses power to influence price, however, this element is outside the domain of the present study. 30 In the context of Mallen's framework, this channel structure would approach that of a simple monopoly, however, the channel member that has a channel monopoly/monopsony, the distribution center, does in fact compete with other channel structures within a given market. The nature of competition between distribution centers in a given market can be viewed as oligopolistic in nature, although they compete indirectly through retailers. Competition across retailers can then be viewed as a competitive oligopoly, where the competitive state is limited to the degree of competition between distribution centers, that set cost (and selling price assuming a direct pass through) for the retailer. The role of price/incentive activity within this framework can be seen as follows: The distribution center receives a price incentive from the manufacturer. This incentive is based on the desire of the manufacturer to increase sales and market share at the expense of other 8 A performance requirement on the part of competing brands. the distribution center/retail network may or may not be required.9 The incentive based on the performance require- ment may or may not be passed along to the retailer and consumer. For the purposes of this study, it is assumed that the incentive is passed along at both levels. The above discussion examined the competitive structure of grocery channels citing the nature of competition within and across channels based on the application of price theory 31 as proposed by Mallen. In order to understand the nature of the channel in terms of channel member size and cost structure, and the underlying rationale on the part of the manufacturer for price incentives, the theorem on division of labor as it relates to vertical integration as proposed 10 by Stigler can be utilized. Division of Labor - Channel Structure In 1951, George J. Stigler published ”The Division of Labor is Limited by the Extent of the Market." He proposed. -that the concept of division of labor could be applied to the understanding of vertical market structures. Stigler viewed the firm as engaging in a series of activities, each of which may have a different cost relationship. These activities, when seen in terms of different operations within a channel, would in many cases result in non-optimal operating volumes for certain channel members. It would therefore follow that the size of the institution would eventually conform to the optimal output level for that activity at that point in the vertical channel structure. As Stigler pointed out, however, this might not always be the case. A monopoly situation or public regulation might interfere with the natural workings of the price system. In addition, new firms might require services that are below the level of optimum output for another channel member. Therefore they might be required to engage in the activity to ensure the function being carried out. Firms in mature or declining industries might also find themselves required 32 to engage in activities that previously had been performed by other firms for the same reason. In this case the firm that needed the activity performed would be forced to vertically integrate to assure a source of supply or other function. In the context of channel structure in the grocery industry, the division of labor and its resultant cost structures results 1J1 three distinctly different sizes and types of operations. The manufacturer typically has great economies of scale, and therefore operates a limited number of processing plants. The wholesaler or distribution center has relatively large operating economies, however, it must exist at a point between the producer and the market, and due to market size and market share, its operating level may be less than it would be excluding this fact. The retailer has the lowest relative operating level due to the necessity of being close to its consumer market. Because of the decreasing cost function of the manufac- turer additional output is a desirable objective. This can be accomplished through a trade incentive, that to a certain extent reflects the production cost saving in price reduc- tion. This incentive may have a much different effect on the distribution center. Because of the fact that a distri- bution center may not have the same falling or relatively constant cost function as the manufacturer, the incentive may simply increase those costs. In addition, those increased costs may not be offset by increased total gross 33 margin during the incentive period. Given that the manu- facturer may be offering the same incentive to other dis- tribution centers within the same market area, increases in volume for the retailer may not be significant. - In addi- tion, the increase in sales may be at the expense of other products carried by the distribution center and retailer. The difference in the cost structures between the channel members may make use of an incentive beneficial to the manufacturer only. Potential benefits to the distri- bution center and retailer may be negated because of different and increasing cost functions in addition to a highly competitive market structure. The hypothesized relationship between these cost structures is seen in Figure 2.2. At point 01 the level of sales is depicted in a non-incentive situation. Both manufacturer and distribution center/retailer are operating at a profit. At 02 which represents an increase in volume due to the incentive, profit level for the manufacturer has increased. Despite the leveling of the total revenue curve due to the lowering of price by the manufacturer, overall profitability has increased as costs continue to increase at a relatively moderate rate due to continuing economies of scale. Thus, for the manufacturer, the incentive increases volume and profitability. 34 Figure 2.1 Relationship Between Manufacturer and Retailer/Distribution Center Cost and Revenue Curves Manufacturer A Total Revenue /Cp———- Total Cost \\ Retailer/Distribution Center Total Revenue Total Cost 35 The retailer/distribution center faces a different cost curve as seen in the lower panel. As with the manufacturer, revenue declines on a relative basis due to the lowered price during the incentive. Total costs, however, follow a different pattern. The retailer/distribution center has costs that differ in structure and pattern from the econo- mies of scale of the manufacturer. It is hypothesized that these costs increase (N) a relative basis and result in a decrease in overall profitability to the retailer and distribution center. This study is designed to examine the exact nature of those costs as they result in increased or decreased profitability for the retailer and distribution center. The overall impact on profitability is dependent also on the slope of the total revenue curve. In the present study, the slope of the revenue curve is depicted by differing levels of input parameters including sales response increase, payback, and sales revenue per unit. Figure 2.1 depicts this relationship including those areas of cost and revenue lines for the retailer/ distribu- tion center that this study has examined. Importance of Distribution Efficiency The general issue of distribution efficiency has been of concern to marketing scholars for the better part of this century. Early studies within the discipline and courses instructed carried the term ”distribution” prior to the .11 common acceptance of the term "marketing. Early descrip- tive studies of distribution practices in relation to farm 36 12 products were done by James E. Hagerty and John Franklin 113 in the first decade of the twentieth century. The Crowel depression of the 1930's brought forth concern with market and distribution efficiency and produced major studies such as the Does Distribution Cost Too Much? report prepared by The Twentieth Century Fund in 1939.14 The period following the Second World War saw further developments including the 15 whereby development of the systemic approach by Bryer selling activities within a channel of distribution were considered to be one of several ”disposal” activities for a channel member. Distribution efficiency as it relates to total channel performance was examined in the years following from a total cost and systems approach. Lewis, Culliton, and Steel16 examined air freight from a total cost approach, including all costs required to carry out a logistical task. The systems approach, as applied to logistics by Shycon and Maffei,17 Parker,18 and others allowed for the examination of tradeoffs within a logistical system. Within this framework, the field of logistical management and the measurement of distribution efficiency has developed at a rapid rate. The specific issue relating to the impact of incentives on channel performance has not been examined to any great extent in the literature, although there have been calls for it, some reaching back to the early 1950's.19 The following 37 two sections examine the relevant literature that has addressed this and similar problems. Single Echelon Models This literature deals with endeavors to arrive at decision rules relative to the use of an incentive. These typically involve criteria that a channel member may use to arrive at the quantity that should be- ordered in an incentive scenario. This literature essentially deals with profit maximation at one level of the channel. It does not. incorporate multi-echelon considerations or any factors that are dynamic in nature. Within these limitations, this literature is directly concerned with cost efficiency for a channel member as impacted by an incentive activity. One such approach to the decision criteria for the 20 He acceptance of volume discounts has been made by Weiss. suggests using a return on investment criteria within an economic order formulation. This formulation is then modified for a volume discount situation. The basic formula that he proposes is: l D = [W + “1 ‘C’ ”300) Where: Order quantity producing the desired ROI Minimum inventory units to usage Carrying cost D M C This formula, with a carrying cost of 20% and a ROI of 15%, will (cause the traditional EOQ to be reduced by approximately one third. To examine order quantity under 38 condition of an incentive, Weiss suggests the following fomulation: Order the larger quantity if the following is true: A(E-G) + AK[-%)- - 11“] R01 + c< M .. MA (3-3) Annual usage in units Smaller order quantity Unit price of smaller quantity Larger order quantity Unit price of larger quantity Cost of ordering Where: A D E F G K This formulation, in addition to utilizing the concept of return on investment where the inventory acquisition is required to generate a given return, also incorporates a decision rule. This rule, based on the cost versus benefit of taking a larger quantity, allows a purchasing manager to assess the relative benefit of the larger order, within the constraints of the formula. The higher the result of the formula is above the figure [ROI + C] the more advantageous it is to take that quantity at that price. Although used by FMC Corporation, the decision rule relates to one level only of a multi-echelon channel structure, and does not include many factors that are subject to change with larger order quantities in an incentive situation. Another formulation to determine the proper quantity to 21 His approach is based on a order is offered by Silver. case lot scenario, whereby a buyer has the option of pur- chasing in single or case lot quantities. If the purchase 39 is made in case lot quantity, a discount is offered. Although this is not technically an incentive, the same basic principles apply. Silver's approach is based initially on the determination of what he terms ROSA, or return on supplementary asset. In this computation, a comparison is made between the savings associated with the larger purchase quantity, and the added carrying costs on carrying the additional inventory, assuming a normal pattern of sales. This savings (or added expense) is then compared to the additional investment required to finance the increased inventory. Finally, this figure is adjusted to reflect the debt/equity relationship within the organization and becomes what he terms ROSE, or return on supplementary equity. The decision to accept or reject the larger quan- tity and lower cost relates to the threshold level of managerially acceptable ROSE. As with the formula proposed by Weiss, Silver's approach is single echelon in nature and excludes many dynamic elements. Multiple Echelon Models This area of the literature consists of examination of related problems within a multi-echelon structure utilizing relatively advanced research designs. This literature, although addressing problems only related to the present study, provides a background of research within the general area of distribution channel performance and efficiency. Literature that deals with multiple echelon problems to 122 the present one includes an approach by Blieme who 40 proposes using a trade-off approach in making inventory decisions. His position is that a firm must go beyond attempting to minimize cost at a given service level. He states that the real challenge is to maximize benefit to the finm by making trade-offs between market response and costs associated with inventory decisions. To illustrate his position, he utilizes examples of trade-off decisions as they relate to capital tied up versus incremental sales Volume, and capital tied up in inventory versus spoiled goods. Although Bliemel's intent to illustrate the use of trade-off analysis in inventory decisions apart from the impact of incentives on channel performance, his use of trade-off examples closely parallels the format of results to be generated in this study. An example of this can be seen in Figure 2.2. In this example, Bliemel shows the relationship between the contribution margin of the product, the cost of capital, and the change in sales volume associated with one week's increase in inventory. For example, with the data given in Figure 2.2, given a contribution margin of 30%, a cost of capital of 15% and an increase in the level of inventory by one week, a .7% increase in the level of sales would be required to make this a profitable trade-off. The form of this presentation, showing multiple trade- offs in an inventory decision environment is very similar to the type of trade-offs that are examined in the present study. 41 Figure 2.2 Trade-off Limits: Contribution from % Change in Sales Volume vs. Carrying Cost of l-Week's Inventory Contribution opportunity of % changes in sales Contribution - . . I - Margin of volume equnalent to 1 week's Inventory wnen Product cost of capital :5 10% 15% 20% 10% 1.85 2.70 ' 3.60 20% .80 1.20 1.60 30% .47 .70 .93 40% .30 .45 .60 50% .20 .30 .40 60% .13 .20 .27 70% .09 .13 .17 80% .05 .02 .10 EXAMPLE: A.7%changeinsalesvolumecasedbya 1 week'schangein inventory would be a limit" 109a protitdale trade-off between contribution opportmity and carrying costs, given that the collItSr‘itiutimmrgindtheprowctismwcostoicmital I: Source: Bliemel, Friedhelm. ”Inventory Decisions by Trade- Off Analysis: A New Approach In Product-Orientated Marketing Strategies,” Journal of Business Logis- tics, Vol. 1, No. 2. An approach utilized by Mohantry and Chandrashekhar23 examined a problem similar to the present one. Their con- cern was that of co-ordination between production and distribution subsystems, the primary concern being that ordering policies in the distribution subsystem may affect the way demand arrives at the production subsystem, and that scheduling policies in the production subsystem can affect supply to distributors. Their examination of this problem utilized simulation modeling whereby a single manufacturing 42 plant producing Inultiple products through Inultiple ware- houses was replicated. The factors examined included pro- duction rates, work force levels, and inventory levels in the system at different points in time. Their system allowed for the smoothing of production to account for discrepancies in production level versus forecast produc- tion. The forecast was the sum of the orders received from the distribution subsystem. Production smoothing was achieved through a mixed strategy of inventory holding and changes in production rates and overtime. The causal loop structure of this system is seen in Figure 2.3. Within this system it was found that the optimal policy was to hold 17.5% of lead time requirements as safety stock. This level would result in the least amount of total cost for the system when applied to each warehouse as operating policy. In addition, in order to examine the effects of interactions between the production and Idistribution systems, a partial factorial design was utilized. This research, although dealing with a different problem, utilized similar methodology to the present research. In addition, it dealt with dynamic interactions within a pro- duction and distribution environment that are closely related to those of the present study. Source: Mohanty, R.P., and Chandrashekhar, V., “Computer Simulation Study for a Production-Distribution System,” International Journal of Physical Distri- bution andiMaterials Management. Vo1. 13, No. 3. 24 examined the problem of channel Corstjens and Doyle optimization utilizing signomial. geometric programming to solve three distribution decisions simultaneously. One element of the three, the use of price to influence the performance of the units within the channel, has signifi- cance for the present study. Their methodology was designed to solve simultaneously problems of distribution strategy, intensity, and management. Distribution strategy is defined as the selection of channels to serve end-markets. Distri- bution intensity relates to the number of outlets to be operated within each channel utilized. Distribution manage- ment relates to the use of price and other variables to influence the performance of units within the channel. The business that their approach was based upon was a manufacturer of high quality candy that operated its own 44 retail network in addition to selling to other stores that sold their candy cut a franchise basis. The company also sold to distributors in European countries and to domestic retailers utilizing the retailer's brand name. The problem facing the company was which channels to concentrate on, the number of outlets to have in each channel, and what margins to seek in each channel. The model developed in the study was based on a series of demand and cost functions. These demand and cost functions were derived from observed data and management judgement, and resulted in suggestions for increased prices, concentration on larger company owned stores and export and private label business. Corstjens and Doyle, although utilizing a new and different methodology from the present study, addressed a conceptually similar problem. The cost and demand functions utilized to determine optimal strategies relate closely to the theoretical base of this research. This research, dealing in a fixed channel structure scenario, has cost structure variations between manufacturer and distribution center as the proposed basis for the research hypotheses. In addition, differing levels and patterns of demand are utilized to partially determine overall levels of channel performance. These investigations of related problems in a multiple echelon structure provide insight into comparable research designs related to logistical performance issues in a multiple echelon setting. The discussion which follows 45 examines the relevant literature on market responses to trade incentive and promotional activities. Market Response to Promotions and Incentives This area of the literature relates to the impact of incentives on consumer purchase behavior. The impact of incentives on intermediaries within the channel is not well known, however, its impact on the ultimate consumer has been examined extensively in the literature. This information provides insight into the sales response function within the channel, which is predicated on the consumer response. In addition 1“: provides insight into incentive activities at the retail level. Literature on the market response to incentive acti- vities covers a broad range of topics including the consumer rationale for accepting an incentive, the nature of the consumer response, the nature of products and prices as they relate to incentives, and the type of consumer segments that are likely to respond heavily to incentive activity. The research that has been conducted in this area provides a background for several of the assumptions that are made in Chapter III. IAs trade incentives often result in consumer related promotional activities, including special displays, advertising, and price discounts, this literature provides information on several aspects of trade incentives as they relate to market demand. 25 Blattberg, Eppen, and Lieberman examined price deal- ing on consumer non-durables in the context of a model that 46 describes consumer and retailer welfare. The model for the retailer examined quantity sold on deal, and the frequency and magnitude of deals. The consumer model examined the quantity bought on deal and the time between purchases. They concluded that price dealing exists because consumers engage in stockpiling activity. Both the retailer and consumer have inventory holding costs that are comprised of two components. These include a cost of capital and a cost of storage. The cost of storage is less for the consumer, thus a retailer through a deal can shift inventory to the consumer in return for a decrease in revenue associated with the deal. An alternative theory that was not supported by the data collected suggested that dealing exists as a means to reduce the cost of a consumer's experimental purchase with a new brand. The implication of their findings is that the benefit of price deal may be overstated as consumers are merely stockpiling the goods and not switching brands. (Cotton and Babb26 utilizing consumer panel data on purchases of dairy products made a number of findings related to deal purchases. These include that the response to a promotional deal is much greater than a reduction in price alone. In addition, the response to a deal was less for products more familiar to the consumer. The lowest response was for fluid milk whereas response rates for products such as yogurt were much higher. They also determined that in store specials were less effective in increasing purchases than other types of specials. Finally, 47 it was determined that substantial increases in sales occurred during the promotion, however carryover effects in later periods were much less. Massey and Frank27 utilizing consumer panel data made a number of findings relative to dealing effects on food and household products typically sold through grocery stores. They determined that family units not loyal to brands had a much higher reaction to changes in price than brand loyal families. In addition, price elasticity was found to be higher for larger sized containers. Finally, price elasti- city was found to be higher in chain stores as opposed to independent dealers. Utilizing a similar data base Webster28 determined that consumers prone to dealing activities tended to buy more brands and devote a smaller share of their purchases to favored brands. He determined that deal proneness is inversely related to the amount purchased. Also, older housewives appeared to be more prone to dealing than younger ones. Webster believed that the older housewife may be a more expert shopper or has a more flexible budget. The latter explanation parallels the findings of Massey and Frank, as they found a positive relationship between income and size of package purchased, the larger size having a greater price elasticity. Both findings would also tend to support the consumer stockpiling notion advanced by Blattberg, et al. 48 McCann29 determined that medium to light users of a product responded less to price dealing than did heavy users which would also support the stockpiling notion. Households paying a higher price were also more responsive to price dealing than households paying a lower price. McCann indicates that the household paying the higher price may believe that it is "splurging", therefore they may be sensitive to price changes within their evoked set. A household paying lower prices may be satisfying their price consciousness simply by paying the lower price initially. A further explanation might be found in the previously cited notion of financial ability to stockpile, as indicated by the pattern of higher priced purchases. McCann also determined that innovators were more responsive to price dealing than non-innovators, and that households that limited their purchases to a small number of brands were less sensitive to dealing activities than those that purchased a wider variety of brands. Store loyalty was seen to have the opposite effect. The higher the loyalty to the store, the greater the response to dealing activities within the context of that store. Brown30 determined further the relationship between brand loyalty and responsiveness to dealing activity for purchasers of instant coffee. Utilizing a telephone survey it was determined that promotional activities yielded a faster response than did advertising. The promotional activities did not, however, produce long-term buyers. 49 A similar finding was made by Sexton3‘1 whereby in a study of the relative effects of pricing, dealing, and print media versus long term advertisimg, it was found that the former activities have a greater impact on short-term brand share. In a comparison of pricing changes to point of sale promotions, Woodside and Waddle32 found that for instant coffee ixIIa supermarket setting, consumers responded at a higher level for point of sale promotion than for price reduction when each was taken separately. When both were used in combination, the sales increase proved to be the sum of the two independent increases when taken separately. These findings are similar to those of Cotton and Babb in this regard. Chavalier33 using an experimental design, with eight products determined that competitive market structures where no one particular product had a clear market share advantage resulted in increased market response to in-store displays. This result is similar in nature to those regarding brand loyalty by Cotton and Babb, Massey and Frank, McCann, and Brown cited previously. In addition, Chavalier determined that products in a mature stage of the product life cycle had a higher sales response than products in an earlier stage. This study also determined that previous advertising activity had little effect on the degree of sales response, and that the size of the price reduction likewise had minimal effect. It would appear that the tradition of 50 special displays being associated with price reduction caused consumers to react in a pre-learned manner. These results closely follow those of Cotton and Babb, and Woodside and waddle in the comparison of pricing to promo- tional activities on sales response. The literature on sales response to promotional acti- vities at the retailer-consumer level yields an overall pattern of behavior on the part of the consumer. Consumers appear to be engaging in a stockpiling as opposed to a brand switching activity. This would indicate that promotional activities are at best transferring inventory to the con- sumer on a short-term basis. It would logically follow that a sales payback function at the end of the promotion would be relatively high. Brand loyalty appears to be inversely related to response to promotional deals, and at the same time consumers who are more adventuresome have a higher response rate to promotions. Package size appears to be related to response to promotional deals, as larger sizes yield a higher response rate. In a similar fashion, higher income and older consumers were more prone to respond to a promotional deal. These findings would also support the stockpiling notion. Finally the degree of price reduction associated with the promotional deal appears to have a minor effect relative to the other activities associated with the promotion. A summary of the literature on market response to incentive activities appears in Appendix A. 51 Contribution of the Present Research Based on the review of the literature in this chapter the following conclusions are warranted: 1) The competitive and price environment in a grocery products distribution system is that where there is a strong inducement for a manufacturer to utilize trade incentives. 2) Because of variations in cost structure these incentives, although producing positive bene- fits for the manufacturer, may be done so at the expense of the distribution center/retail network. 3) Single echelon models have addressed this issue from the perspective of one channel member, however they do not allow for many variables, including the dynamic nature of a multi-echelon trade incentive activity. 4) Multiple echelon models have addressed related problems; the need remains, however to utilize this type of research methodology to investi- gate the logistical impacts of trade incentive activity. 5) A broad base of literature exists that provides a background of knowledge on the market reaction to trade incentive/promotional activities. This background can be utilized to operationalize a multiple echelon model to investigate the present problem. The present research is designed to address the issue of distribution performance measures and overall efficiency as it is impacted by trade incentive activity in a dynamic multiple echelon environment. Despite calls for this type of research, and the existance of research methodologies capable of addressing the problem, this issue has not been previously investigated. This research project is designed to address that need. 52 This chapter has examined the theoretical base for the present study, and the related literature. The chapter which follows details the model structure to be utilized and the assumptions developed to test the research hypotheses. 53 FOOTNOTES-CHAPTER II 1Mallen, Bruce, ”Introducing the Marketing Channel to Price Theory," Journal of Marketing, Vol. 28 (July 1964) PP. 29-330 2 Ibid. 3Marion, Bruce W., et al. The Food Retailing Industry, (Praeger Publishers, New York, 1979), pp. 6. 4Daft, Lynn M., "Competition in the Grocery Retailing Industry,” Ph.D. Dissertation. Buchigan State University, 1969. pp. 84. 5Gold, Faye, et a1. Modern Supermarket Operations Third Edition (Fairchild Publications, New York, 1981), pp. 17. 6Chamberlain, E.H., The Theory of Monopolistic Competition (Cambridge: Harvard University Press, 1957). 7Daft, pp. 88. 8Blattberg, Robert B., Gary D. Eppen, and Joshua Lieberman, "A Theoretical and Empirical Evaluation of Price Deals for Consumer Nondurables,” Journal of Marketing, 45 (Winter 1981), PP. 116-129. 9Refer to Appendix B for types of trade incentive performance requirements. 10Stigler, George J., ”The Division of Labor is Limited by the Extent of the Market,” Journal of Political Economy, 1951. 11Bartels, Robert, The History of Marketing Thought Second Edition (Grid Series in Marketing, 1976), pp. 21-22. lzIbid. pp. 142. 13Crowell, J.F., ”Report of the Industrial Commission on the Distribution of Farm Products (washington, D.C.: Government Printing Office, 1901). 54 14The Twentieth Century Fund, Does Distribution Cost Too Much? (New York, 1939). 15Breyer, Ralph E., Quantitative Systemic Analeis and Control: Study No. l-Cfignnel and Channel Group Costing. (Philadelphia: College Offset Press, 1949). 16Lewis, Howard T., James W. Culliton, and Jack D. Steel, The Role of Air Freight in Physical Distribution, Graduate School of Business Administration, Harvard University 1956. 17Shycon, Harvey N., and Richard B. Maffei, "Simulation- Tool for Better Distribution,” Harvard Business Review Vol. 38. November-December 1960. 18Parker, Donald 0., ”Improved Efficiency and Reduced Cost in Marketing," Journal of Marketing, Vol. 26. April 19Smith, Carles W., "Increasing Distribution Efficiency by Better Organized Research," The Journal of Marketing (January, 1953). 20Weiss, Fred R., ”When do Volume Discounts Pay?" Purchasing Magazine. 21Silver, Alan. "When to Buy in Case Lot Quantities,” IQ_(November 1983). pp. 125-134. 22Bliemel, Friedhelm, "Inventory Decisions By Trade-Off Analysis: A New Approach in Product-Oriented Marketing Strategies,“ Journal of Business Logistics Volume 1, Number 2. pp. 103-119. 23Mohanty, R.P., Chandrashekhar, V., "Computer Simula- tion Study for a Production Distribution System,” Inter- national Journal of Physical Distribution and MateriaIs Management Volume 13, pp. 51-67. 24Corstjens, Marcel and Peter Doyle, ”Channel Optimi- zation in Complex Marketing Systems," Management Science Vblume 25, Number 10, October 1979. 25 Blattberg, et al. pp. 116-129. 26Cotton, B.C. and E.M. Babb, ”Consumer Response to Promotional Deals,” Journal of Marketing, 42 (July 1978), 27Massy, William F. and Ronald E. Frank, ”Short Term Pricing and Dealing Effects in Selected Market Segments,” Journal of Marketing Research, 11:2 (May 1965): PP. 171-185. 55 28Webster, Frederick E., Jr., ”The Deal-Prone Consumer,” Journal of MarketingiResearch, 2 (May 1965), PP. 186-189. 29McCann, J.M., "Market Segment Response to Marketing Decision Variables," Journal of Marketing Research, 11 (November 1974), pp. 399-412. 30Brown, Robert 6., ”Sales Response to Promotions and Advertising,” Journal of Advertising Research, 14:4 (August 1974), pp. 33-39} 31Sexton, D.E., Jr., ”Estimating Marketing Policy Effects on Sales of a Frequently Purchased Product,” Journal of Marketing Research, 7 (August 1970), pp. 338-347. 32Woodside, Arch G. and Gerald L. Waddle, ”Sales Effects of In-Store Advertising,” Journal of Advertising Research, 15:3 (June 1975). 33Chevalier, Michel, "Increase in Sales Due to In-Store Display,” Journal of Marketing Research, 12:4 (November 1975), pp. 426-431. CHAPTER III RESEARCH METHODOLOGY Introduction This chapter examines the research hypotheses and the structure of the model utilized to investigate them. The hypotheses are developed based upon possible negative impacts of incentive activities on measureable outputs of distribution channel performance. The simulation model is described, including the system configuration, operational policies, and the levels of parameters to be tested. The method of analysis, utilizing chi-square and analysis of variance is detailed, indicating the nature of the output and the means of analysis. Development of Hypotheses Trade incentive activity has increased at a rapid rate 1 and in all probability will continue to do in recent years so. Despite the widespread use of trade incentives, little is known about their overall impact on channel performance. This research is designed to provide insight into this problem with the hypotheses providing as wide a perspective as possible on the issue at hand. The general nature of the individual hypotheses is the impact of trade incentive 56 57 activities on specific measures of distribution channel performance. Prior research has demonstrated impacts of ordering policies on multi-echelon systems. Forrester2 has shown the impacts on a production-distribution system of changes in the pattern of retail orders. Figure 3.1 shows the impact of a 10% unexpected rise and fall in retail sales over a one year period. As shown in the figure, substantial changes occur across the channel that clearly have performance implications. Changes in retail orders are similar in nature to the use of a trade incentive, as they can both cause changes in volume of goods, and the pattern that they follow through a distribution channel. This change in volume and pattern was shown by Forrester's experiments to result in substantial fluctuations in inventory levels, production output, and order filling delays. These changes, it would logically follow, would have a negative impact on channel performance. Orlicky3 has suggested in the context of a manufac- turing environment that end orders can create "lumpy demand" through the use of order point logic. In his example, seen in Figure 3.2, the demand for and inventory levels of items further back in the system show marked discontinuity. This is due to the timing of orders on an irregular basis and also due to the difference in the lot size at each level. Although Orlicky's presentation is in the context of a manufacturing environment, both conditions of irregular 58 .mch can >oafi3 ccon can mmopm a 52 a :2. a 32 a 30m B:4IH 32 a 52 a :3. 5 R32 a Row B :2. a 52 a 32 5 Sq co 8 2. 8 I 3 o. R 8 o. ma .xuow 3oz .ocH .H.H.: use .moMEm: o Howuumsccm ..z >mn .uoumouuom "mousom O: OO— O _ \Ildl'l I. .. w \\ «.3- II 2.3130 .38 .. ,. \\ u - I 05:... .003 16.26 I II. _ «one: \ \. Coin-a /.p/ 3.655... :63: :3: o~ ....... x . .. zH./ V \.\LIIII .. . . \ 2U . I (.11. \ I/ . . \ // :wulnm \\ 00mm I \\ 7:5; to: . li- So...3 22:5 , 4. On... :aoszol few-C Q-NT .253. ud. 33:02: :32... 1.33:6... xxx . . 3.3 .531 ..ve! \ :63 oz: 326.9. :3: / > $33 ....CI. /.19~I\:_¢3 uKm :005 3.19.:30 01. .32.)... 25:6 :22... \ :0»! x :23 3:52.20 \ {own + . mt: «5:52.46 55.. 200.0 :ozou / .\ ...-v 2.23 no: :23. a. :35 05.33.23! oom~ .... .. I. —. ,9 FW .1. 32:3 :1. :35»... .33: . canon umowloco m uo>o moamm meumm cu Hank can omwm oouoo xoca com a cu Eoumxm cowuscfiuumMoIcofiuoscOEm m we uncommom H.m whamfim 59 Figure 3.2 Order Point and Dependent Demand End product Many snzl' inde;endenl Comonds Iron Cusrorrcrs Invenfory J F M ’ I ComponenI of end producf Few large demands dependenI on producl monufacIure l 0’ .v vac... ;, 3.4.: 3'§-::::.:3: ’ “ white" 0 ‘5 o oo ‘aawfi%*¥5s§€§ 5 0 $3 ' 06090900000 fififih&&&&&&50o_ 03$ .0 O 0 , ° 4944 .... O o v Mmflfiwaxamt .......... u Raw material . Few large dc , ands dependen! on component manufacture I I I I I i A Source: Orlicky, Joseph, Material Requirements Planning McGraw Hill Inc. 1975. pp. 26. 60 demand and difference in lot sizes exist in a distribution environment under an incentive. The examples cited would imply, if transferred to a distribution renvironment with incentive activity that I\) cycles of inventory levels and other channel performance criteria would develop and B) flow of merchandise could occur (”1 an irregular basis. The hypotheses which follow are thus based on the negative impacts of activities associated with trade incentives. on_ distribution channel performance. General Research gguestion: To reflect the overall nature of the problem at hand, the general research question is stated as: What is the impact of variables associated with a manufacturer sponsored trade incentive on a company owned distribution center and retail store network in terms of measureable outputs of logistical performance which in turn impact the overall financial performance of that network? The above question contains three basic elements: 1) variables associated with a manufacturer sponsored trade incentive, 2) measureable outputs of logistical perfor- mance, and 3) financial performance of the distribution center/retail network. Thus, the nature of the general research question and specific hypotheses take the form of variables associated with incentive activity as the independent variables impacting measurable outputs of logistical performance as the dependent variables. These outputs of logistical performance then in turn impact the profitability of the distribution center/retail network. 61 The following hypotheses reflect these relationships: H The use of an incentive increases inventory levels for the products under the incentive at the distribution center. 2 The use of an incentive decreases the overall customer service level from. the Idistribution center to the retailer for products under the incentive. H3 The use of an incentive results in increased variation in shipment sizes between the manufacturer and the distribution center. H4 The use of an incentive results in an increase 1J1 the total number of shipments from the manufacturer to the distribution center. H5 The use of an incentive results in a decreased level of financial performance for the distribution center/retail network. Model Development Overview: In order to test the research hypotheses a grocery product distribution system was modeled utilizing the Simulated Product Sales Forecasting (SPSF) model. The grocery product utilized was a typical packaged goods product. The product modeled packed twelve to a case, and weighed one pound per unit. Initial non-incentive unit cost for the distribution center was $1.00. This product was simulated in a mix of ten products produced by a single manufacturer. Each of the ten products was configured with the specifications of this product. System Cbnfiguration: The simulated system configura- tion consisted of three echelons, a single plant inventory location, five distribution centers, each with five retail locations. The plant inventory location carried five 62 products that were subject to a trade incentive, and five other products that were not subject to the incentive. The distribution center and retail locations were configured as being owned by the same concern, which is typical of large 4 The echelon structure is depicted chain operations. in Figure 3.3. Incentive Description: The incentive that was utilized in the testing environment was initiated by the manufacturer in order to increase production volume and to move inventory to lower points in the channel. The incentive consisted of a price reduction in return for a larger order commitment by the distribution center. The incentive was accepted by the distribution center with cost and volume changes as detailed in Table 4.59. The sales response of the market was modeled independently of the acceptance of the incentive by the distribution center. The retail locations were assumed to promote the incentivized products and to either pass or not pass along the price reduction to the consumer depending on the scenario. The specific actions taken by the management of the retail locations is outside of the domain of this research, however the results of those activities are implicit in the sales response levels modeled. The cost information utilized in testing hypothesis five considers this activity. The duration time of each incentive period was two months in length. There were two incentive periods modeled in each year, with each simulation run spanning three years. 63 @iEEE@€EE€€E_Efih@ l€.:::.sagas Luucou souemu Loucmu toacou goucmu 8:32;; 85323:. 5.53235 8.53:: .5 8:32pm; _ _ L :o_uau04 xtou=m>cn weeps cofiunu: Hucou Eoum m Hoccmco n.m wusmwm 64 The time frame of the incentives are depicted in Figure 3.4. At the conclusion of the incentive, under two of the three levels for this factor, it was assumed that there will be a reduction in sales due to stockpiling activity on the part of the consumer.5 In the third level, no reduction in sales occurred at the end of the incentive period. These payback levels and periods are illustrated in Figure 3.5. Inventory Policies: Initial demand for products was based on daily sales demand by consumers at the retail level. In the case of a stockout, the order was permanently' lost. Replenishment of inventory for the retailer was obtained through orders placed with the distribution center. The retailer ordered in variable quantities based on previous and anticipated demand sufficient to carry through to the following shipment. The retailer ordered in minimums of single case size quantities and attempted to maintain a 6 minimum of stock above that on shelf display. If an order was out of stock at the distribution center or plant inventory location , it was not backordered.7 The distribution center acquired inventory by orders placed with the manufacturer. The distribution center utilized an order point, order quantity method with the order quantity predicated on forecasted demand. Inventory levels and safety stock were maintained based on previous experience with fluctuations in demand from the retail locations and orders from the plant inventory location. 65 Id m < h w Z < 2 w W LOCOS L _ ._ NIIImponom o>HDCmocH.11\\\\\\ uncommom can mcflewa o>wucoocH e.m magmas 66 m w - H I 1' u-- e .m>Hu:ooaH uwum< xomczmm. same o>aucoocH mo Rom u< xomnamm modem o>wucoocH m.m ouamwm sucoz moamm 67 These policies were designed to yield a customer service level at or around 92%. The manufacturer produced the products on a continuous basis utilizing historic data. Production volume levels were constant with enough inventory front loaded to allow for incentive shipments from the plant inventory location. Sales volume simulated was within the capacity of the manufacturing operation, and the availability of raw materials was considered infinite. It was assumed that the manufacturing operation could adjust production levels to match normal changes in demand. In the case of a variation in demand that was above a forecasted level that was positive, stockouts would occur. In the case of an unusual decrease in demand, it was assumed that the excess production would be held in plant inventory. Uncertainty: Uncertainty within the channel configu- ration was produced through stochastic lead time and demand variables. Lead time uncertainty was replicated by the use of a gamma distribution. The gamma distribution produces 8 Uncer- non-negative values that are skewed to the right. tainty of demand was also created through the use of a gamma distribution. [Mdiizing the gamma distribution, a random number generator within the demand module of the SPSF generated daily demand.9 Transportation System and Cost: The transportation system between the manufacturer and distribution center consisted of common carrier truck and rail shipments with 68 freight rates as seen in Table 4. Outbound shipments from the distribution center Ix) the retail locations were made via a private fleet owned by the retail/distribution center concern. Cost of outbound freight was a cost per hundred- weight basis, with no other rates or breaks applicable. Shipments inbound Ix) the distribution center were made as required, whereas outbound shipments to the retail locations were made on a predetermined daily schedule. When neces- sary, additional shipments were scheduled to the retail locations on a temporary basis. Given the above model configuration and assumptions, the discussion which follows details the variables and levels employed in the testing environment. Research Desigg In order to structure the research design, variables are defined and abbreviations assigned for each of the independent and dependent variables. Table 3.1 lists the variables to be utilized, and a description of each follows. 69 Table 3.1 Testing Environment Variables SPSS Integer Code Variables Abbreviation (Independent) Sales Response Increase Increase of 0% SRI00 Increase of 20% SRIZO Increase of 40% SRI40 Sales Reponse Uncertainty Low Uncertainty Level SL501 High Uncertainty Level SLSUZ Sales Response Payback Zero Payback SRPOO 50% Payback SRP50 100% Payback SRP10 Incentive Level Non-Incentive Scenario INC00 20% Incentive Increase INC20 40% Incentive Increase INC40 Product Incentive Products (Products 1-5) PRODl Non-Incentive Products (Products 6-10) PRODZ (Dependent) Inventory Level TTINV Shipment Size Manufacturer to D.C. SHPCAT Number of Shipments Manufacturer to D.C. TOTSHP Customer Service Level D.C. to Retail Location SVC Total Financial Performance TOTCOST UNH WNH NNH Independent variables that are associated with 70 an incentive activity were tested in all possible combinations. These variables are as follows: Sales Response Increase SRIOO SRIZO SRI40 Sales SLSUl SLSUZ Sales SPAOO SPA50 SPAIO No increase in sales during the incentive period; the non-incentive scenario. An increase in sales during the incentive period of 20% above the normal level of sales. An increase in sales during the incentive period of 40% above the normal level of sales. Response Uncertainty A relatively low level of sales uncer- tainty. Coefficient of variation = .3 of daily demand. A relatively high level of sales uncer- tainty coefficient of variation = .7 of daily demand. Response Payback .A reduction in the normal level of sales at the conclusion of the incentive period at a level of zero. A reduction in the normal level of sales at the conclusion of the incentive period equal to 50% of the sales increase during the incentive period. A reduction in the normal level of sales at the conclusion of the incentive period equal to 100% of the sales increase dur- ing the incentive period. Incentive Level INCOO The non-incentive scenario whereby reorder logic remains the same during the incentive periods as non-incentive periods. 71 INC20 .A 20% incentive whereby shipments are made at the beginning of the incentive equal to the normal sales level prior to the incentive plus 20% for the incentive period. INC40 IA 40% incentive whereby shipments are made at the beginning of the incentive equal to the normal sales level prior to the incentive plus 40% for the incentive period. Dependent variables that measure cost and customer service ramifications of the various combinations of the above variables are as follows: Inventory Level TTINV The inventory level of products at the distribution center location. This is measured in average inventory per product per distribution center for each report period. Shipment Size SHPCAT ‘The shipment sizes between the manufac- turer and the distribution center. This is measured by the number of shipments that fall into the following rate breaks for each reporting period: 4000 LBS; 20,000 LBS: 30,000 LBS: 40,000 LBS; 100,000 LBS. Number of Shipments TOTSHP The total number of shipments from th manufacturer to the distribution center. This is measured in total for each reporting period. It was assumed that incentive activity on a product will not increase the number of shipments to the retail location as the one product represents a minimal percentage of total sales. 72 Customer Service Level SCV The percent of orders shipped on time from the distribution center to the retail location. Total Financial Performance TOTCOST The summary of financial performance based on cost and revenue assignments made to measures of logistical perfor- mance. This is based on: TOTCOST = REV - (SHPCOST + INVCOST + SVCCOST) .A total of twenty six simulation runs was required to test all relevant combinations. An incentive scenario is depicted in twenty four runs and a non-incentive scenario is depicted in two runs. For the non-incentive scenario, the only variable that is tested at different levels was sales uncertainty. A summary of incentive run combinations appears in Table 3.2. In addition to testing the research hypotheses, the simulation runs provided a wide range of information on the impacts of various factors associated with incentive activities. The discussion which follows examines the structure of the simulation runs, the output generated, and the method of analysis. Simulation Runs: Each of the twenty six simulation runs replicated the distribution channel activities of the network for a three year period. The three year period was of sufficient length to allow for changes in outputs during incentive periods and to capture their effect on subsequent periods of time. Measures of outputs were made at the end 73 TABLE 3.2 RUN COMBINATIONS TESTING ENVIRONMENT SIMULATION LEVEL OF INDEPENDENT VARIABLE S SRI INC SRP TESTING CONDITION 1.0 1.0 1.0 001 002 1.2 1.4 003 004 1.4 1.0 1.0 1.2 1.0 005 006 007 1.4 1.0 1.4 008 1.4 1.4 009 1.4 010 1.0 1.0 1.0 1.4 011 012 1.4 1.0 1.4 013 014 1.2 1.2 .8 015 016 1.4 . .6 1.0 1.0 1.0 1.2 017 1.0 018 1.2 1.2 019 1.4 020 021 022 1.4 1.4 1.4 1.4 1.4 023 024 1.0 1.0 1.4 1.4 025 026 1.4 74 of each year's activity. Output was measured from the plant inventory location and distribution center, as appropriate. The total sample which resulted from the simulation run consisted of five measures for each product and distribution center at the end of each year. Thus, for a single measure of performance each of the simulation runs produced 150 measures of output (5 distribution centers x 10 products x 3 years reports). The study in total produced 19,500 measures of output (150 per measure per simulation x 5 measures x 26 simulation runs). In addition to this, shipment summaries for the system in total for each report period were generated. Method of Analysis: To test for the relative impacts of incentive versus non-incentive conditions (N1 channel performance, an analysis of variance or chi square analysis was performed between appropriate testing conditions. Analysis of variance is described below: ”Analysis of variance is a statistical technique that assesses the effect of one or more categorical independent variables (factors), measured at any level upon a continuous dependent variable that is usually assumed to be measured at an interval level. Conceptually the cases are divided into categories based on their values for each of the independent variables, and the differences between the means of these categories on the dependent variable are tested for statistical significance. The relative effect upon the dependent variables of each of the independent variables, their cthined effects and interactions, may be assessed." The $953 ANOVAl 1 procedure is based on the least-squares general-linear hypothesis approach and as noted above measures differences in the means of interval 75 level dependent variables. Outputs of the testing environment simulation runs are in the form of interval level. Interval level measurement is described as: "The interval level of measurement has the property that the distances between the categories are defined in terms of fixed and equal units."12 Testing environment outputs met thisI criteria, as unit counts, dollar levels, and percentages are interval level. Mean scores were generated for each output variable from 75 observations across the three year simulation runs. These observations include those prior to, during, and after incentive periods. They thus measure average levels of performance for the dependent variable given the levels of the independent factors. In the SPSS ANOVA procedure, independent variables may be all nonmetric or combinations of nonmetric and metric variables. In the testing environment for the present research, the independent variables are nonmetric, categorical variables and are thus referred to as factors.13 In order to test the specific research hypotheses, factors that are secondary to the hypothesis were held constant, allowing the Idifference in the factor Iiirectly related to the hypothesis to be examined. For the present research, incentive levels (incentive versus non-incentive) were examined, with the other factors held constant. This procedure was performed for each of the dependent variables as they relate to each specific hypothesis. In addition, an analysis of variance was performed within each incentive 76 scenario to determine the relative impacts of each of the other independent variables on each dependent variable. In order to test the hypotheses related to shipments, 14 was also utilized to determine the chi-square analysis significance of shipment changes between testing conditions. A final analysis of the testing environment output was made whereby a dollar cost was assigned to each level of dependent measures. Comparisons were then made to determine the overall cost effectiveness of incentive versus non-incentive activities. In this procedure a total financial performance figure was developed for each testing condition. The above analysis of variance procedure was then followed to determine the impact of the various factors on total financial performance. Model Validation The SPSF model was originally validated through an extensive procedure whereby the model was tested by each individual module for logical program statistical output, face validity by review groups, model stability, and model 15 Program validity consists of testing the sensitivity. model's reaction to various input parameters to determine that the model operates according to design specifications. Face validity is more qualitative in nature and deals with the assumptions in the model reflecting actual business operations, understandability of the reports generated, applicability to various situations, and the fit of the results to what would be expected in a business situation. 77 Prior to the initialization of the present testing environment, the model was tested for stability and for reaction to input variables. Customer service level was utilized to test for stability within this procedure. Simulation runs were made on testing conditions 001, 010, and 013. These conditions represented a non-incentive scenario, an incentive scenario with sales increase with no payback, and an incentive scenario with sales increase with payback. Utilizing an analysis of variance procedure, customer service levels were compared for each of the products across the three year time duration of the simulation. F values ranged from .683 to 1.973 with the significance levels of the F values ranging from .1465 to .5084, indicating no statistically significant difference across time at the ot= .05 level. The model was also tested for reaction to the input variables. A comparison was made for each dependent variable based on changes in the INC, SRI, SRP, and SLSU variables. The result of these initial runs and the analysis of variance tests on customer service levels may be seen in Appendix C. Summary This chapter has developed and structured the research hypotheses. Independent and dependent variables have been developed and defined. The structure of the model and the assumptions within it have been presented. Finally, the nature of the output and the method of analysis has been 78 identified. The chapter which follows describes the output of the testing environment and the statistical analysis of the results generated. 79 CHAPTER III - FOOTNOTES 1Strang, Roger A., ”Sales Prmotion-Fast Growth, Faulty Management," Harvard Business Review, 54:4 (July-August 1976): pp. 115-124. 2Forrester, Jay W. Industrial Dynamics, The M.I.T. Press and John Wiley and Sons, Inc. New York. 1961. 3Orlicky, Josephy, Material Requirements Planning: McGraw Hill Inc. 1975. pp. 26-27. ' 4Marion, Bruce W., et al. The _Food Repailing_ Industry-Market Structure, Profitsy and PriCes, Praeger Publishers, New York (1979) pp. 12-13. 5Blattberg, Robert B., Gary D. Eppen, and Joshua Lieberman. ”A Theoretical and Empirical Evaluation of Price Deals for Consumer Nondurables," Journal of Marketim, 45 (Winter 1981), pp. 116-129. 6Gold, Faye, et a1. Modern Supermarkpt Operations, Fairchild Publications, New York (1981) PP. 87. 7This is the normal procedure in the grocery industry. For pharmaceutical type items, backorders are usually accepted. Reference discussion with Richard Fezette, Distribution Manager, Dow Chemical Company, Consumer Products Division, Indianapolis, Indiana. 8Basic, E. Martin, ”Development and Application of a Gamma * Based Inventory Management Theory" (Ph.D. dissertation, Michigan State University, 1965). 9Bowersox, Donald J., et a1. Simulated Product Sales Forecasting, M.S.U. Business Studies (1979) pp. 14515. 10Nie, Norman H., et a1., Statistical Packa e for the Social Sciences, McGraw Hill Bock Company, New York 91975), pp. 9. 11 Ibid. 12Ibid. pp. 5. 13Ibid. pp. 399. 80 Ibid. pp. 223. 15Bowersox, Donald J. et a1.,‘Simulated Product Sales Forecasting. MSU Business Studies, 1979. pp. 199-271. CHAPTER IV RESULTS AND CONCLUSIONS This chapter details the findings of the present research. Hypotheses are presented in the order previously listed with output data presented in tables, graphs and in statistical analysis reports“ Each hypothesis is examined first, then a review is made of additional input variables and their impact on the dependent variable related to that hypothesis. Hypothesis One: The use of an incentive increases inventory levels for the product under the incentive at the distribution center. The results of the testing environment support this hypothesis. As seen in Figure 4.1 incentive conditions resulted in an increase in inventory level over the non incentive condition. Testing condition 001 represented a non-incentive scenario whereas testing condition 006 and 013 represented incentive scenarios with 20% and 40% sales increases respectively. The presence of an incentive increased dramatically the average inventory. In non- incentive scenario average inventory was 1,295 cases. This level increased to 2,437 and 2,687 cases during the incen- tive scenarios for the higher level of sales uncertainty. A 81 82 27 ' Low Uncertainty Inventory Level High Uncertainty 16 - 15 - 13 - 12 - Non-Incentive 20% Incentive 50% Incentive 001 006 013 01h 019 026 Testing Condition Figure 4.1 Comparison of inventory levels for incentive and non incentive testing conditions 83 lower level of uncertainty also resulted in a higher level of inventory. To test this hypothesis, a one-way analysis of variance was performed to test for the impact of incentive level with an equal sales increase without any payback. These results are seen in Table 4.1. This resulted in an F-value of 6537.8 with the probability of F at zero for the lower level of uncertainty. This would clearly indicate that there was a major and significant difference in average inventory level during incentive testing conditions as opposed to the non-incentive scenario. At a higher level of uncertainty, the difference in inventory level was not as great, however they were also significantly different at a F probability of zero. These results are seen in Table 4.2. These and the results which follow were due to the changes in independent variables, not from the random number stream, which was the same for each testing condition. These results overall strongly supported Hypothesis One. To test for the impacts of other variables on inventory levels within 20% and 40% incentive levels further compari- sons were made. Figure 4.2 depicts the impact of various sales responses without payback within a 20% incentive. Testing condition 005 represents a zero sales increase due to the incentive whereas TC 006 and 007 represent 20% and 40% sales increases. In this comparison inventory levels did decrease as the sales response becomes greater. These 84 HHOH.OHNN cu mmm~.moom mnoo.amqm cu wmca.oacm ~cam.~om~ cu ~am~.~m~u can: you ucu ucou uum mm .Aoum m coco.wmo~ ooco.mmm~ ccco.e~c~ cocc.mQMH laluxox Hem.nmmo Dayna h oooo.waHH coco.secm coco.mo- oooc.mm- Issued: neu¢.-mo mano.moo~mnac ouuasum can: m~m~.du ommm.oH o~m~.m nouum vuuvcmum caqq.no on~c.~m mcom.~m couumu>on vuuvsaum mmmm.e0o~c¢HZHm4 3o: H< 4m>mq m>HHzmqu >9 am>ma >xoezm>zH mo moz mo mHm>Aoo cam: vuuvcwum vumvcmum oooo.mofimmnmo «NN mmm¢.enmmn nmnu.m~m~mm- NNN mam.m¢¢ nma¢.nommmcwm NQNQ.QMHmmmmm N bayou m nouuaum cam: mouuzum we saw Ideml NHZHma 20H: H< 4m>m4 m>HHzmqu rm 4m>ma Machzm>zH mo moz mo mHm>Aon euavcuum vuaucoum emmn.moamo~o coe~.ocomnoH cmfim.anm~o¢¢ ouuasdm no flow ommm.o~¢~ mmmm.-m~ nemm.nmc~ scan.mmem cum: «Nu NNN NHZHm4 ggfi— H< mmmHfizmozH NON rm qm>ma >¢OHZN>ZH mo moz<~m<> ho mHm>Aun vumvcoum vuuvcmum amem.nmeammm ooe~.¢wommon amoc.accmmmm muuuaum ad How wmmc.aqm~ momm.mmam mmma.~mm~ ammo.ncm~ cum: «NN NNN szHuq =UH: H<.mmmHBzmqu NON rm qm>ma >zoszm>zH mo moz<~m<> mo mHm>Amq 3OA H< mmmHHzmqu Noe we 4m>ma rmOHzm>zH mo moz¢~m<> mo wuquon coo: mmmmw vumvcmum vuuvcoum OOQN.Ommano- «NN HauOH owes.mssm~ oo-.omseomos NNN uaaouu canon: eom.~m ooo~.o¢ommom cemm.mm~mmMHH N onoouo savanna cuunx m nouoaum coo: nouaoum no Bow IJMdml mouoom >BZHm4 :OH: H< wmmHHzmqu NOQ rm 4m>mq >¢OHZM>ZH mo moz mo mHm>Aon one: unaccoum unaccoum mmsO.~aneN~ aea mamo.cmmm scoo.NmHH¢NH med moo. Noce.wacs NcOe.mHoN a cage: m mouaawm cow: mmumoum mo Sam .m.a :zHSS—moza .mO ANS: 30.— H< mama—.— MuHH2NUZH NON rm AM>MA rmOHzm>zH ho muz hO mum>4mn can: vuwvcmum vumvcmum mmmm.mH~o¢ON med NqOO.NOOHfizquH NON we 4m>mq rmoezm>zH mo moz mo m~m>dHZHm4 SOA H< mqm>ma nNN mm mm nu acoou Hooch maoouu cusufiz maoouu coozumm ouuoom xoHezmozH NON rm qm>m4 >¢OHZM>ZH mo moz<~¢<> mo mHm>Am4 =OH: H< m4m>m4 xoHHzm02H NON rm .55: 55925:: .mO muz .mO mumwa—di OH.¢ MAOmn com: ucoou unaccoum vuuvcmum mmmm.wm~amoa med Hmuob “was.maoan mnaa.mmam~sfl men maaouu canon: oNcm. mnm. ooom.NNHm ocow.NNHa H maoouo couzuom .aoum m ofiuam m oouuoam can: mmumoum mo Bom .m.a mouoom wezHmq 304 94 mqm>mq xoHHzmozH NOe wn 4m>mq VMOHzm>zH mO moz mO mHm>Ama pudendum pumvcmum mmmm.mmmommq Neoe.ONmmomw NOOO.NMO¢N moumovm «0 Bow rfizHflhmmozs no 4m>m4 :OHI H< mgm>ma NOON.omoN OmH mnma.-- ms OOO¢.em0N mm one: ucoou ova Hobos sea masouu conga: H maoouo coozumm .m.a mouoom m>HHszZH NOQ rm Am>m4 >mo92m>zH mo muz no mmm>4ma 30A H< mqm>ma ¥oHHszzH NO¢ rm Am>mq >MOHZM>2H no muz mO mHm>AHZHm4 OOH: B< mqm>ma MoHHzmozH NOe wm 4m>m4 >xoezm>z~ mo moz mO muquoo unaccoum moumowm mo aom wthmq 304 H< qm>ma m>HszozH wm 4m>mq MOH>¢mm mszHmOO hO moz<~m<> mO mHmwama mun: H< Am>mq m>HHzmozH rm Am>m4 mo~>xmm «MXOBmOO ho moz<~m<> mo mHm>Am4 NOH>xmm ¢NZOHmOO ho moz hO mHm>Aua unaccuum nNmO. «NOO. NONO. muueowm mo Bow erHm4 ZOA 9< mmmthmozH NON ON.Q mqmds ammo. NmNm. Nawm. «Nmm. new: «NN NNN mmm annoy ma aoo ma coo ma no: acoou seaguucoo wcfiumoa Hausa aasoao cacao: maoouo cooauom uuuoom 1515 OmmO. ONON. Nmmw. HNOO. Ou Oocm. OOHO. Omen. HcOO. Ammo. Hemm. OOOO. cu mNNm. Oflea. OMHO. mmOO. ROMO. mONO. NOOO. Ou ONOO. Omma. OMNO. NNOO. nNNO. «MOO. one: new you ucoo uum no loafixaz mmmwmwm IIMMMMMI mmmmmmwmm coo: euavsuum vuovcaum NONN. QNN OHOO. OONN. NNN OOOO. NOm.Om OONO. OOmO. N «NMMMIM. mmMMMIM nouaovm can: nuuaovm no flow Ideml erHm4 :OHO H< mmmuhzmozH NON mom am>md mOH>mmm “MXOHOOO mO moz<~¢<> no mum>4HHzmozH Noe mom 4m>ma mOH>¢mm «NZOhmOO mo moz<~m<> ho mum>4oa cumucnum NowO. OOOO. OONO. moumovm mo Bow >HZHm4 30; H4 ON . e mqmfiu maqa. Nana. qua. NNOO. can: QNN NNN mNN mN mN mN ucooo :oNuuvcoo kuop mmoouu :«zuwz unsouu cmoauem ouuoom 1137 ONNO. ou OOmO. OOOO. ou ammo. NOHO. 0u NOOO. one: ace ucH «coo gum mO OOOO. .aoaa a OmmO. OONO. OmmO. OmmO. aoafixw: OO0.0N owuom m OOON. OOON. OONO. OONO. anqcaz NNOO. OOmO. mouoovm cam: cmOO. OOOO. «moo. HONO. ONOO. OONO. a; vumucaum vuovsuum ONOO. NOON. ONNO. mmumoum «o How szHmq :OH: H< mmmHHZNOZH NOQ mom Am>m4 MOH>mmm mmxoemao mO moz<~x<> mO mHm>Aoa OOOO. ONOO. OONO. OOHO. OOHO. vuuvccum nonwoum no sow erHm4 304 H< mqm>m4 OOOO. OOOO. HHOO. coo: OOH OOH .h.O MoHHzmqu NON rm Hm>m4 MOH>mmm KMSOHOOU ho muz ho mHmrHoO coo: mmmmm mmwmwmmmm Oumvcmum Oucvsmum wcHumoa NOS. «.3 H38. NOHH. OOH mqoouu :HOqu OOOO. H mnoouo cuoauom mmumovm mo Eom .h.O mmmmmm >82H<9mmoz= ho Am>mq 30H: H< m4m>ma MoHHzmozH NON HO HN>MH MOH>mmm muzOHmOU ho moz ho mHmeaa vuuvsmum unacceum OOHH. OOOO. NOOO. mouuovm mo Ham szHdfimmoza m0 AM>MA 304 H< mHm>mq ¥0HHZMOZH NON rm Hm>m4 MOH>mmm mm ON.O m4m<9 OOOO. OHOO. HNOO. NONO. coo: ONN NNN 2H mmzommmm mma m0 mHm>Hmq 20H: H< m4m>ma xuHHzmozH NON wn Hm>MH m0H>mmm «HZOHOOO ho moz m0 mHmwAoO vuavcaum mouaoum one: >H2Hmq OOH: H< O4m>mq mouaoum mo Bow NOOO. OOOO. OOOO. one: OOH OOH xOHHzmozH NOO VO Hm>uq m0H>¢mO MMtOHOOO m0 m02 m0 OHO>HHHzm02H NOO rm Hm>mq m0H>mmw OMZOHOOO m0 MOZ m0 OHO>HmO one: Ouuvcaum Ouavcmum NOOH. OOH OOOH. OOH OOOO. H moquOm no flow .m.O >HZHmq =0H= H< mqm>m4 x0HZHmq 304 9< O4m>ma OOOO. OOOO. OOOO. NOOO. cum: ONN NNN x0HquozH NOO rm HN>MH m0H>xmm OMZOHOOO m0 moz<~m<> LO OHO>HoO :moz vumvccum vuavccum «OOO. ONN nu 3 O O 11 NHOO HOON NNN OOOO. ON0.00 OOOO. ONOH. N .Ooum m oHqu m muumovm cam: nonwoum no Bow .m.O erHmq 30H: H< mHm>mq M0HHzmozH NOO rm 4m>m4 m0H>xmm OOEOHOOO m0 moz ho OHO>HfiucmocfiIcoc 0cm m>HucoocL :3 L a . «inc 9: amok. one a~o ado mac coo Loo o>«acocc~ mo; os«acooe~ vow nauseous—1:02 . u u o xucueuuouc: and: \H I sued-ounce: so; N coauao .mcoflbwtcoo MOW TUCMEHOMMTQ HMHUCMwa MO COWMHMQEOU cc.s.=coc neaaace a. a os«»ewoc~ mo: . o>qucowen mom . o>nacocm~ocnz 2. I 2- m I o o I c a I a N a N o I O m I N 4 I 4 o I o m I 0 Ca I ca Na o Na .2 I an o“ I an ad sucuouuoucz sad: I nu om I 0N mm I NN :m abououuouca :04 . 4N ”N I oN mm o 0N a a coauao -a ma. . ch mH.a ct:o_a onco>om 170 aomm.mowm~ Ou VOMMImmHMH mmeIVHm wNMMIHHNm wQQVImmmvH mNN atHOH mNHMIQviN Ou mesmIDNNMN ocvoImHmOM ooos.whbma aromIMQN ammonmvN mMmOIMvmmm m5 Mao oorv.mm¢hfi 0» omv~.o~vma ooovIvmmNN ooomewwHH omBNIva QONHINONN owoonwaH mp woo hmwaOmwN Ou MHmNINmmH OOOOIHmmh GOODImmhmI oHHmIFHN mHhHIbomH oochoHVN mp doc zMD zucwmuumoco no am>mq 304 um ~m>mq o>wucmocH >2 oocuEuOuuwm waocmcgm mo oucowuu> uo mfimuamc< ooIv mqm<9 171 ooov.hoona NNMN.¢mow 0» NehmIvNHm ooov.booaa mMFQImomN Cu owOMIMNmH oommIMBNh hwthommN 0» mHmNINmaH oooo.~moh 24m: MOm 82H hzou Hum mm IDtHXND OMqumuuooc: uo Hm>0q 304 um ~o>oq o>flucoocH >n ooanuOuuum HofiocmCMm uo oocnflum> no mwm>~wc< HwIV @4048 ostIvam NamcIvowm hmmo.vcm~ oovamva 24m! VNN NNN .h.0 mNN Aoaol oowm.mw~v NonImvw oomMIvav ommo.~ov OOOOIBQawI QMQQIwgn zatHth momma QmmD Dmucwmuumuc= uo Ho>wq so“: an ~m>cq w>fiucoocH >2 wocuEMOuumm Hofiocucflm uo cocuwuu> uo wwm>awc< Nva mamdfi mhnm.vmmm~ vw¢mIO¢mN~ N5o¢.ahmh~ OOONInvv 24m! QNN NNN .mIo mum 44909 ms oNo ms mac ms «do Hzaou up Hmuoe naaouo swap“: nazouo com3uom MUMDOm 173 onOIoovv 0» mommImNom Ou Mbom.mro~ Ou ZND Dmucfiouuooco no Ho>oa so“: an ~m>mq mmIQ mum<9 o>wucmocH >n oocuEMOuuom HqucMCwm no macawun> uo mwm>amc< HavaImahH omomI¢mmm heao.ommH ooom.mvv 24m: vNN NNN .m.o mNN 4480? m5 wNo mh mmo mh vac HZDOU UP HmuOP maaouo :Mnuwz mascuo cwczuom NUMDOm 174 financial gain or loss of an incentive. This result held true for the higher level of uncertainty also. These results by category are depicted in Tables 4.60 through 4.63. Under the condition of an incentive producing sales equal to the incentive amount, without any payback, the assigned costs subtracted from revenUes produced a positive amd significant gain. Under these conditions, hypotheses five would be accepted. Additional comparisons were made in the same sequence as hypotheses one through four, to examine the impact of sales response increases, and sales response paybacks within a given incentive scenario. Figure 4.16 depicts the impacts of various sales responses, without payback on a 20% level incentive. Under Option One, product revenue increased substantially and significantly as the level of incentive increased. All of these levels of performance are above the figure for the non-incentive scenario also. This result held true for both levels of uncertainty. Under Option Two, however, results were higher than the original level of performance only in the case of the incentive producing a sales response equal to the incentive level. This was true for a low level of uncertainty. At a high level of uncertainty, results were higher than the non-incentive scenario in testing condition 019 also. These results are seen in Tables 4.6 through 4.67. .mHm>oH wmcoammu mmamw acmumwuwo cum: m>flucwoc« wow uOu wocm5u0wuwa Howocmcgm wo cOmwquEov o~.v muaowm 532500 9.3-0... .5325qu 92.4.31. omo So as 8° 90 30 So So So So 80 «8 3.3.5.: “0.. .4332: wow 0.10.35 0.5.x. 0:20.: “C: 039.05 ‘om 3.0.2:: onus I OAI I OAI I m I Q I o I o I a I a I N I N 353.325 :3: I o I o I N I N I a I 4 5 7 3.319525 :0.— I o I o 1 I Q I O I CA I OH I NA I Nd I .1 I o: - a - s uzcw>um sag-:35: gnu: I m.— I 04 I cm I ow .3... a I. I mm a u 00:: 30.. I «N I am I ..N I .WN I 0N N .833 -.. H .838 -.. I On I an I N... I N? 1f76 mhwo.mNNH~ Ou mvmv.m-om Gehc.mmvhu 0» owv~.o~vo~ v~cv.v~m- Ou cawm.hoon 24m: mOm 92H LZOU Hum mm coco. .momm m oomw.wm~a~ oomo.mmNmN oocv.vmmN~ oo~a.o~om~ :DZHXmo omfiozdfim nmucunuuooca wo Ho>oq 30a um nomoouocH oncoomom moamm acoumuufio Lug: o>wuco0cH cow uOu cocoEHOuuom aufiocmcwm mo mucouuc> uo mamaamct vw.v mqmth mNN J4HOB mp poo mp coo ms moo azsoo up auuoe museum conga: museum coozuom mumaom 177 ovmm.womw mmwm.o~mm .mmhmImomN wmmmIVFHNI 0» Ou Ga Ga mvNN.m~N~ homm.mwmv ow0m.mmmH ommeMFmNI 24m: «Om 92H szU Pom mm oooo. .momm h .m cowuao. sucauuuooc: no nus: o>wucoocu «ow uou comm.vthH comm.vmhma oowmIMBNF ooom.w~m ZDZHX<2 mmw.o~m OHH<¢ m doom.~mm owom.hwhm oomh.omeI oochNwm mmOOIOVN Hmvm.mhom ooovIFdel thv.mHaH thVIHNN oomh.wmdwl finnv.oom nmmHIwmnH ZDIHZH: mommm ZOHHmo omoq 304 um mommouucw uncommom moamm acouwuwflo oocchOwuom HawUCMCwm uo oocowun> no nunaamc< meV mammoom oo¢m.ovm¢ mmo~.HmHHN Cu mavw.om~m~ comm.bmoom oomOIemNb bmomewv Hlo.oHNv mwa.mhomH 0» m~m~.omvma oomm.omom~ oomMIMVBV omwo.~ov Hemm.Mva onmm.o>o- 0» mmhmIFH~HH cowh.womm~ oovm.ovmv mmhm.mmm hN¢¢.thm zdmz “Om 92H mzoo Hum mo ZDZHx<2 ZDszHz mommm ZOHBmo om-w-aom Hamv.hwhv¢bma ommr.o~m~¢mamom coco. eho.mm vmon.mhammommfi meov.~mmm~aova .momm m OHH<¢ m mmm<90m 24m: mmmqumuuooc: no Ho>oq so“: an mommouucH oncoomom modem acououuwo cum: m>wucoocH «ow uOu oocuEHOMuom amfiocncdm mo oucufiuu> uo nfim>uuc< ww.¢ want? mNN 44908 mh one mh mac m5 mac 82300 DB Hubby mozouo cocoa: mosouo comsuom momDOm 179 mmmm.m>a~ wNN~.Nva mmmm.m~o~ waQIvaNI 24m! MOm 92H mzou Bum mm Ca Ca Ou 0» omwm.mn mmon.mhhm oome.~nm wmav.hmrml coco. .mOmm m .N cofiuaov sucumuuooca no cud: m>flucmucH now no“ oomh.hhmHH oomh.hhmau oovo.v~oh oomh.oow~ ZDZdez OFOIwm OHH~ cowh.mv¢ol coeo.mthl Hvom.NHv oomh.mvvol mamm.ovm ooom.~bmal whwm.omm :DSHZH: momma omnu v~or.mh~mmv0NMH mmm<30m 24m: moom.>mfiq ¢emo zam: ezaoo ow omcozo4 so“: an mommmuocH mmcoomom mmamm acououuflo oocmEuOuuom Huwococwm mo oucmfluu> no mam>amc< thv mqmfiucmocfl .2 32150 P: .30... ONO «mo :8 m3 «3 do 030.35 $0.. 0.09.05 “ON 0.00.55 chow - - c . .wH0>m~ vasommmu Aug-:02... :3: ausacuuouca :3 N cofiuao wow MOM wUCMEOwth HfiwOCMwa uO cOmwthEOU £2158 9:23... ONO nNo ..No mac Nae .30 0.00.35 Io; 0:00.55 *8 030.35 Sun I S. . . . . . I I w I I o I I .— I I ... I I o . I N I I __ I I o I I m . I 2 I . ma I I ...— I I 9 I I a. I I 8 3503025 :3: I I wm I I ..N 3530025 30.. I I .X I I an I I am I .I a 838 - p~.v whamwm OAI 09c0>0m 182 Homm.¢w~ma Ou mmmm.mmmh~ homm.mmm moah.m¢mv mo z.mmv vaVINcchomwowu hmwm.¢movuomva N Mano .momm m OHHucfimuuooca uo Ho>oa 304 an momcoomom moacm ucmumuuuo cu“: o>uuc00cu oov uOu oucoEHOuuom Hnwococum uo oucuflu0> uo nw0>~mc¢ mev mamdfi mNN 4(909 ms mHo mh NHo mh HMO PZDOU UH AmuOP museum cacao: no cuozuem mumDOm 183 ooov.howaa oomm.~wwwl NHGHIMHMH NNMN.vmow Ou ~o>w.v-m ooov.howua cowh.NMVI wab.ovN aomm.mmo~ Nam¢.¢owm mmMMINHmH Ou ovno.vmo~ oovm.awmm oomNImcmml vmom.m- ovm~.mwm~ mwo~.nwvH hamo.mbhml Ou va~.H~mml OONH.HMMH cowQI~wwwl vv-.hma ~w~m.-mu mvmo.mvaI zmo zucfimuumuc: uo Ho>0q 30¢ um momcoommm mwamm acouOMuwo cud: 0>flucmocu woe new oucmEuOMuom Hmfiocucflm uo moonwum> mo mwm>amc¢ mw.v mqmdfi mNN A¢POH mh m~o ms Nae ms uao HZDOU UH annoy museum cacao: monouo coozuom mumDOm 184 Hmom.mmmvm Ou hvmw.vme~ bfiah.HNmH~ 0» MVOHIaNoom HmNH.wmnm~ Ou MFFNIvamH zmQ GMdDZde avom.mbooowmwmh mmmv.mmmmhwhmov waco.mNHhNamth wufianm m0 tam mmmm.vmmm~ vwvn.ovm- omevahaom NMOh.mmmvH z¢mz vNN NNN .m.o AH newuoov >ucNmuuouca mo ~o>oq no“: at mmmcoomom mwaom acoumuuwo nu“: 0>Hucoocn woe uOu 00cn5u0uu0m Hmflocmcflm uo 00cmfium> mo mfim>~mc¢ Oh.v mam<9 mNN dflHOP ms muo mh mmo m5 vNo HZDOU UH annoy monouo cwzuuz nasouo cmo3uom Homoom 185 ooow.vmeH ooom.mnmaal mo~o.wovv Ou voom.momm oooo.vmwaa oowm.m5m~HI mmmw.hmm vmmo.h~mm 0» wcmOINmNH oomm.mr~o~ oow~.v~whl whHBINmm OFMFIFmHNI Ou meOIFbmml oomMIvam oovv.mmmoHl hNHN.Hbm 24m: MOm 92H mzoo 80m mm ZDZHXmo DmuCamuuooca mo ~o>0q no“: as mmmcoomom madam ucououufio cud: o>wucoocu woe uOu oocoEHOuuom Hawococfim no 00cmfiu0> mo nfim>~mc< ah.v mqmfiucoocfl mom coauascoo 0:.0.09 «so ado woo 80 Juan»!— iON 30330..— chow abounuuuucz :04 Auc«00uou:: gnu: N coauao .mH0>mH xomo>mo ucmumuuflc cums ommouocfl uOu oucmsu0uumo Hmfiucmcflu no comwumosou co.sascno 0:.suoa mac ado N8 80 noon»...— iON 103»... 9.0”». . . . . I OAI I m I o I a I N I o I N I .— I o I o I as auc«0uuou:: 304 I Na I 4‘ I 2 ./ sucuauuuuc: sad: I ma I ON I mm I am I cm I mm .W H couuao I .. m~.e «panda 0A N.— 5 Q4 NN R N? 03c0>0m 187 the 0‘ = .01 level at both levels of uncertainty. In the lower level of uncertainty results were higher than the non-incentive scenario in the case of a sales payback equal to the incentive amount. In the case of higher uncertainty this did not hold true. Under Option Two, the presence of a payback function had a substantial and significant negative impact on performance. These results are seen in Table 4.72 through Table 4.75. This procedure was followed for a 20% incentive with a 40% sales increase at different payback levels as depicted in Figure 4.19. Under Option One, all testing conditions produced results greater than the original testing condi- tions, for both levels of uncertainty. Under Option Two at a low level of uncertainty this held true for only the testing condition where no payback occurred. At the high level of uncertainty, results were below the non incentive level only in the case of the full payback level. The impact of sales payback within this incentive and sales response comparison was significant in all cases at the uI= .01 level. These results are seen in Tables 4.76 through 4.79. An additional comparison was made within a 40% incen- tive with a 20% sales increase at a zero and 20% payback level. These results are depicted in Figure 4.20 and Tables 4.80 through 4.83. Again Option One produced results higher than the original testing condition under both conditions of uncertainty. Option Two produced results below the original 188 .mH0>0H xomn>mo ucmumuufic nu“: ommmuOCN 0mcoommu mwamm mov cufls 0>NucwoCN «on new wocMEuOWuwo Hmfiocmcwu mo cOmNumoEou mH.v musuwm 53.38 9.3-0... 83.38 33...... So o8 08 So So 08 nag... 8.. .3”..sz SN gosswwoouuN new”? 3.. loam”. “ON fiafiflwooufl . _ . . . I 0.. . . . . . I SI I c I a afliuuuoucz find: I q I o I ._ . I _. I N I ... . o I 6 353.506: :0.— I m I a I .. I I. I o I o I o I o I 3 sued-0.50.5 30... I 3 I NN I S I ... I ... I 9 3503025 :3: I 3 OS¢O>QM I o. I o4 I 8 I 8 I NN I «N I ..m I aw I we I on I am I cw N .838 -.. N .833 - a I S. I «m. 189 OOOVIVQmNN mwmw.mmNNNH Ou mmeIwmvam oONBImmmmH corvIMMQFH Ou oovHIONvoa ooov.vmmNN zwH ooom.mmm- cthIva mONH.NONN :DIHZH: momma ZOHHmo am¢oz<fim am¢02¢8m m¢hm.bvmowmmmma vommImmmowmm wmmv.HNmNvawm mwNmImthvvmmm momm.mwhhvvmmm mmm0A 304 00 m~0>0q xomoamm acououufin zuwa ommmuocH mwcoomox moamm wow 20“: 0>Nucmoca now uOu oocuauOuuom Nowucmcfim uo macawuu> wo mNm>H0c< NFIV mam<fi 1590 commIMBNh ooow.mmoml VNGFImmoHI 0» oomOIbomml oooo.mom~ ooowIQmoml mvrv.wwa mmomImomN 0» ooOMIMNQH coomIMNNh oocv.hwmul thv.HNN zND Om mmm0q 304 00 0H0>0A xunoxmm ucououuao saws omomuocH mmcoomom moaum wow zuwz 0>NucwocH NON uOu 00cmEuOuumm HmwocncNm uo mocuNuu> No mNm>H0c< mn.v mqmv oooo.Hov avwm.mrcm totHzH: mommm ZOHHmO omucNmuuooco uo Nm>oa soda um m~0>oq xuoo>om ucououuao nag: monouocm omcoomom modem com nu“: 0>wucooca now new ooanuOuuom Nowocucwm mo macawuo> mo mNn>Hoc< vs.v m4m<fi oma ddHOH ms m.o mp m.o 92:00 on .muoa monouo :.:uNs monouo c003uom mumoom 1192 oovoIvNoh NNNh.o~mI 0» NHHoIMNhal oomm.owom mommImNoN Ou oooQIHho oovoIVNoh 24m: mow 82H mzou Hum mo zazuxmo nadozucwmuuooco mo N0>oq :oN: um m~0>0q xonoamm acououuwo :uNz unmouocH oncoomom madam «ON nu“: 0>Nucmucu new new oocmEuOuuom NmNocmcfim mo 00cmwuu> uo mfim>amc< mh.v mamdfi omH 44808 ms m.o ms m.o azoou up .0009 maaouo 2.20.3 monouu coosuom mUMDOm 193 hmm~.momwa Ou qumIOMNHH oroo.mo~o~ 0» hth.oNNm~ ohoo.mNNHN Ou mvvaMNHON 24m: mOL 82H mZOU Hum mm oooo. Imomm m oomoIomNmN oooa.mmmh~ ooNFIoNoHN oooo.ommon ZDZHXMO Dmucfimuuooco no ~0>0q 304 um m~w>0q xumoxmm acououuflo nu“: oommouucH omcoomwm moamw woe nufiz 0>fiucoocH «ow uOu oocmEHOuuom HmNUCQCum no oucwfiuo> uo mwm>amc< oh.v manta mNN Josooo mmm¢DOm zmD QMCDzdhm MNBBIomAmmohth mMNoIVMN¢NHomb oov~.-mvmmmaou mmKCDOm m0 tam bmvaoNoH «NQNImomHI moom.mo~H oooHINvmm zucNmuuouco uo ~m>04 304 um m~m>oa xomn>mm ucmuoumwo :ufi3 oomoonocH omcoomom madam mow nu“: o>wucouCH aoN new oucaEhOuuom Hufiocucfim mo oucmwun> mo mwmxamcd FBIV month mNN QmD nmucwmuuouco uo No>oq now: an m~0>0q xumnamm ucmuouuwo.nuN3 commouucu omcoowom mmaum oov nu“: 0>Mucoocn mom new oucusuONuom HmNococh uo oocnwuu> mo 0am>~ac< oh.v mqm<9 mNN 44809 mo hao mo ago mh oNo 92:00 U9 «MUCH monouo oocu.z museum coozuom mumoom 196 oomv.omo~I Cu mooNIoonI .NoNoIMoMN Ou mmm~.hoo oNNHINva 0» mmoh.ohh~ zMD Dmucfimuuouco mo No>0q zoo: um ma0>0q xumoxmm ucououmwo 20w: oommouocu omcoommm moamm wow so“: o>wuc00cn wow you mucMEHOuumm HuflococMm uo oocnwuo> mo mwm>amc< mBIV mam<9 197 0wm0hOCH mmamm MON SUH3 0>wucwocw w0¢ co.a.ocoo u:...oh ago «mo ooo «No Joanaum *oa mudnhum anus . o . Aucuouuouc: 304 sued-0000c: gnu: N :oNuoo .mHo>0H xomoxmo ucmuoumfio nufiz new oucmEHOMumo H0a0CMme mo comwumoeou oo.a.ocoo no.0aoo ANO «NO moo «No nuanacm mom nuananm ouow . u n o I CA I I o I a“. I a I w I O I m I 4 I o I o I OH I Na xucuouuooc: 304 I 3.— I on i Q.— I ow aucuauuouco gnu: I mm I ..N I m« I ON -a a no.0oo I a... o oN.o muoo.o Odo NON-'00 .3 onco>0m 1£98 ommmImava 0» mmav.voom~ ~wo~.oo~m~ Ou NmHh.NmHoH ZCMZ MOm HZH mzou Hum mm oooo. .momm m oooo.ommnm oooBIooohH oooQIommnN SDtdez mooImHN OHH‘M m oooHIvooou oooHIvoooH voovaoH mooo.~vo~ ooooINHBNH omo~.vm~ mmoNINON~ totquz momma ZOHHNQ amucwmuuooco uo ~0>0q 304 um m~0>oa xouo>mm acououufio 2093 00000009 moanm ao~ £093 0>Nucoucu aov uOu 00:05h0uuom Nuwococfim no 0000900> no ngm>~0c< oo.v mamtfi om. 44909 m9 ooo m9 NHo 9zooo o9 Nauo9 maoouo cozu.z 003000 cooauom momaom 199 oohmINHmHI Ou ammo.vmmwl ooMMIN~mH Ou oeho.¢mo~ zdm: MOm 82H @200 Foo mo oooo. .momm m oovo.moom oonmIomm oovo.moom SDZHXmo om~ vam.ovmoommo¢ ovoo.o~mmbomam ovooIonohoon mmmdbom 24m: mmm0q 304 um m~0>0q xuooxmm ucououuflo 2093 0mmmuucH mwamm aoN nufia 0>oucoucu now new mocoEHOuuom Hmoocmcwm uo mocmfluo> mo mwm>~0c< HQIQ mam<fi omH Q¢HOH ms moo mh Nao HZDOU UH quack maoouo conu.3 mascuo coozuom mumDOw 200 oovHIthom momm.¢omo~ Ou 9mmHIBovo~ ooo8Ivao~ baobIHNmHN 0» mvoHIGNooN oov~I~hhom z‘m: mom 82H mzou 80m mm toruxdz oooo. oNHImN .momm m 0H8<¢ m oovo.vo99 oovoIVohh ovvm.oom oooN.vomv ooNH.mnoo mmvoIVBV HMN~.MH~V toIHsz momma 20H84H>mo omtoz<8m omucwuuuooco mo Ho>0a :09: um m~o>oq xonn>0m ucouwwufio :09: 00009009 monm mom £093 0>9ucoocw oov new oucthOuuom unwococfim uo oucu«uu> uo wwm>anc< uoIv mqm<8 omH 44809 m9 ~No m8 mNo 92:00 U8 Hau08 museum coou.3 monouo 000300m NUMDOm 201 whooImoH Ou 900m.vrv~l vomo.9~o~ Ou owoo.NmNH zmo zucflmuuouco no ~0>0A so“: an m~0>oa xumo>0m acououuwo :09: 00000009 mmamm «on 2093 0>9ucoocu aov uOu 00005009000 anwucoc«m uo oucowun> no wwm>auc< mo.v mqm<8 202 level in all cases except TC025. ‘ When a payback occurred these levels were negative at both uncertainty levels. In the absence of payback, the low level of uncertainty produced a value less than the non-incentive scenario, whereas the high level of uncertainty produced greater amount. Vfithin this comparison, the impact of payback on financial performance was significant in all cases at the = .01 level. A final comparison was made for a 40% incentive with a 40% sales increase with different payback levels. As seen previously, under Option One improved performance over the non-incentive scenario occurred. These results were significantly different from each other at the d = .01 level. Under Option Two, under a low level of uncertainty, these results were above the original level, only in the condithmu of a zero sales payback. Under a high level of uncertainty these results were above the original level for both a zero and 20% payback level. The results are depicted in Figure 4.21 and Tables 4.84 through 4.87. The comparisons made within this hypothesis would indicate that 1.) Under ideal conditions of a sales response increase equal to the incentive level and without any sales payback occurring that an incentive would improve financial performance. 2.) Under Option One, where the incentive discount is not passed on to the consumer performance is improved. 3.) A sales increase not equal to the level of the incentive may reduce financial performance. 4.) The 203 omnouocH mmHmw 333:8 05 039 $0 NNO 80 08 m8 So 3385 new no}? 9.3 I o o .23»... 8.. - auchuuouco 304 AucHauuouco :uH: IAIIIIIIIIHHHH N coHuoo .mH0>0H xomo>mo ucouowwHo coo; woo £093 0>90¢00¢9 woe 000 mocmEuOMumo HmHocmCMm mo COmHumoEou 83.2.3 9.389 m8 NNO 0&0 oHo moo 20 100950“. wow x0093...— 090s . o 003?. 00.. . 0A 4A 0A QA mm AucHouuouco 304 >00H009000: :nH: H coHuoo HN.o muoon O.- Na an 0::0>0m OH 2(14 mmhh.mmova Ou 9MM8.No8ma Noom.ooHoN Ou omHm.oonH om~m.oovw~ Ou ovho.bhmm~ 24m: MOL 82H 0200 800 mm oooo. .m0m0 m oovoIoHoom oomNI¢mooH ooo¢Iomomu oovoIQHoom IDSHXm0 0muchu00oco uo H0>00 304 00 0H0>0q x000>00 0:0000000 0003 00000000 m0Hmm now 2003 0>Muc0ucH vow 2003 00005000000 900000000 00 0000000> uo 000>H00< v0.0 @4048 mNN a<909 m9 oHo m9 ooo m9 mHo 92:00 09 H0009 003000 09:09: 003000 0003000 MUGDOM 205 oooQIFooaa moomI99qu Ou ooov.vaNI ooOMImooH omom.¢m9~ 0» o0v~.omo~ oooo.0mom NNMN.voo0 0» No90IQNHm ooov.9o0HH z00 0m~mvv- mmmtoom 00 tom momo.HooN mmN 00908 mmvH.ommHI m9 oHo mvooI9mm~ m9 ooo NHmv.voom m9 mHo 240: 92:00 08 «mm «0008 NNN monouo chqu N 003000 0003000 .m.0 mumoom .N 0000000 >0000000oco uo H0>0u 300 00 0H0>0q xo0o>00 000000000 0003 0000000H 00H0m wow :00) 0>900000H aov 0093 00005000000 Hufiococum mo 0000900> no 000>900< mo.v 04048 206 oom0.om>hd Ou mom0.~wmma mmmv.¢m¢- Ou bomm.vao~ Hmwm.m~NeN Ou h¢~m.vmwam zMD 00”.ovvam vocm.ovm- z0a :00: um 0~0>0q x000>00 0:0000000 2003 00000020 00000 ao¢ £003 0>00200¢0 new :00: 00:05000000 Hawocncfim 00 00:0000> 00 000>Hoc< om.¢ mam<9 mNN 44908 m5 MNO ms NNO mu wwo 82000 08 H0009 003000 00:00: 000000 c003000 000000 207 ocow.vmo- omam.aOVI 00 vm0~.hcoul 0000.005h hmv0.mn~n 00 ammo.0hm~ cowh.nom- oo~o.wovv 0» v00~.ncn~ coco.v00~— 2‘0: 000 920 0200 800 no [Druxtt coco. 0vm.m~ .0000 0 00840 0 comm.nhm-I co~0.mmoai uth.oom ocoo.hooml ovmm.b~v 000m.MFa-! omoo.h~m 00:0200 00000 nztoztkm anom.ohauhmv~ nvoo.oowuhmcnv 0004000 240: m~m~.n~m~ mNN 0mo 200: 02:00 up nzuC000000c= 0o ~0>0q :00: 00 0~0>00 0000>00 0:0000000 :00: 000000:0 00000 «0' £003 0>0ucoocu 00¢ 0003 00000000000 ~000c0c00 00 00:0000> 00 000>~0c< 50.0 mam<fi 208 presence of sales payback at the conclusion of the incentive may reduce financial performance and 5.) passing on the incentive discount may compound the effects of number three and four above, resulting in financial performance substan- tially below the non-incentive scenario. Results of all testing conditions are depicted in Table 4.88. Overall, hypothesis five was rejected, however, it must be noted that under less than ideal conditions, in many cases financial performance was diminished. Clearly the cost impacts of logistical changes due to an incentive can be substantial. This chapter has described the results and conclusions of the research. Chapter V which follows, contains the implications of the research findings. 209 oh.oomv eo.~o¢m ow.v-m ho.v~Hm ma.ofiom mo.ooom oh.mhmm mo.mvm~ o¢.momm mH.Hoom om.ohom om.o¢om mo.o¢n~ oN.mwoN ma.moma NM.HN©H mv.bmma mv.m~bH ¢>.mmma vm.oho~ oH.onH mH.om>H Ho.¢hma mo.omo~ mo.mm¢~ 5H.hooam .D.m Hm.vmmm mm.vmo~ Amm.bmmmo Aom.mmHHo mo.movm Amv.v¢mv H¢.momm Ho.omMH AmH.ohomo Amm.omm~o om.mom~ Abm.H~oHo om.mvv m¢.voom H~.mova Aoo.ovamv A¢H.omofio mo.hmmN Aab.mmNNv oH.Nvmm eo.¢om~ Amm.vhmmo Ao~.momav oo.mo- on.HhmHo mv.waemm z¢m ow.¢hmm mm.~vom hm.mmom hm.oNom mo.ovn~ mo.omv~ hN.~oNN ho.oom~ mv.ovo~ om.~vo~ oo.HvoH mo.¢mm~ NH.NoNN wo.~mm~ mm.NNoH No.mvHN vh.oboa 5H.hoofiw .D.m mm.ovo- m¢.mhmom ob.omova vo.~moo~ oH.o¢vHN Hm.mm¢h~ mo.ooHoN he.ommha N¢.¢moHH om.¢om- vv.moth Nm.oo~va om.m¢v oo.mvom~ H¢.omoo~ on.moo- mm.oo~v~ wm.omwma hm.H¢o¢H mN.¢hwoN Ho.ommoa mo.moo- vo.o¢-~ ov.¢HbmH o~.~hoHH ov.w~¢mm 24m: H cofiumo MUZ¢ZMOLMmm 4zm UzHBmmH oo.¢ wanmh Hv-Ov-iv-iv-Cv-{F-lHHHHHHNNNNNNNNNNNNN swam H m U) .m>Hm mHmmmeomwm lHNMMHNMNMMHNMHNMMHNMNMMHNM O: m U) lHNNMHHHNNMHHHHNNMHHHNNMHHH U 2 H lHNNNNNNMMMMMMHNNNNNNMMMMMM mmo mNo «no mmo NNo Hmo oNo mHo oHo hao oao mao «do mao NHo HHo oHo moo woo hoo moo moo voo moo Noo Hoo 210 CHAPTER IV - FOOTNOTES 1For a review of inventory costs and costing metho- dology see: Lambert, Douglas M. and James R. Stock, gtrategic Physical Distribution Management. Richard D. Irwin, Inc. 1982. pp. 241-271. 2Rates for this calculation was based on a relative rate per hundred weight as follows: Cat 1 4,000 LBS $12.87 per hundred wt (LTL) 2 20,000 LBS 10.29 per hundred wt (LTL) 3 30,000 LBS 6.06 per hundred wt (TL) 4 40,000 LBS 6.06 per hundred wt (TL) 5 100,000 LBS 4.56 per hundred wt (CL) Adapted from: Donald F. Wood and James C. Johnson. Con- temgorary Transportation. Penwell Publishing Company. 0 pp. - 0 3Based on: Harvey M. Shycon, ”How Customer Service Impacts Sales Growth and ROI and How Distribution Can Help Increase Corporate Success.“ Proceedings of the National Council of Physical Distribution Management, Annual Meeting, San Francisco, California, October 10-13, 1982. Volume II. 4Based on discussion with Michelle Green, Promotions Manager, Dow Chemical Company, Consumer Products Division, Indianapolis, Indiana. Also refer to Appendix D for a summary of margins by products in the supermarket industry. CHAPTER V IMPLICATIONS OF THE RESEARCH This chapter discusses the research findings in terms of conclusions drawn by hypothesis, implications of the research in total, limitations of the research, and finally, future research that is warranted. Review of Hypotheses Hypothesis One: The use of an incentive increases inventory level for the product under the incentive at the distribution center. This hypothesis focused on the impact of incentives on inventory level. The findings of the present research would strongly suggest that inventory level is dramatically affected by the presence of an incentive. Inventory levels were seen to increase with the introduction of an incentive, and to be relatively less sensitive to other variables such as the degree of sales increase or payback. Inventory level was seen to increase with the use of an incentive, even in the case of the incentive having a sales increase equal to the incentive, and no sales payback whatsoever. The degree of increase in inventory level was significant when non—incentive and incentive scenarios were compared. 211 212 Within incentive scenarios, differences in sales response levels caused inventory levels to be significantly different. As the level of sales response increased, the level of inventory decreased. Differences in sales response payback generally did not cause a significant change in inventory level, which was higher as the level of payback increased. The increase in inventory level between incentive and non-incentive scenarios would indicate that even in the best possible set of circumstances, where the level of sales increase equalled the incentive level and no payback took place, a higher inventory cost would occur. The significant change in inventory level that occurred when the sales response level was changed would also indicate that this factor would result in substantial cost changes. The level of inventory change due to the different levels of payback which were not significant would imply that this factor would have considerably less impact on cost changes for the firm and channel involved. The difference in the sensitivity of inventory level to these two factors may be due to the fact that the payback factor occurred further away in time from the initial inventory loading, thus the impact of that change was introduced after the inventory level was modified by the degree of sales increase. The research findings indicate that the use of incent— ives should be monitored closely by firm and channel manage— ment as they do result in dramatic inventory changes. The 213 areas that need to be examined in this context include the degree of initial incentive inventory loading, and the degree of sales forecast accuracy for the incentive period. The importance of accurately assessing the degree of sales payback, based on the research findings, should carry a lower level of managerial concern. The industry examined had a relatively fast inventory turnover, causing the incentive impact on inventory to be less than what it would be in a slow turnover situation. In an industry with slow inventory turnover an incentive inventory loading might take months to deplete if sales did not occur at the anticipated level during the incentive. The industry examined had a relatively low cost of inventory causing the financial impact of inventory changes to be less than they would be in an industry with higher inventory costs.1 The financial disadvantages caused by a higher inventory level associated with incentive usage would also be greater in industries with a higher degree of shrinkage or obsolescence.2 Overall, the research findings strongly indicate that management considering involvement in an incentive with a forward buy should carefully consider the inventory load effect. This effect was found to be detrimental to the financial performance of the distribution center/ retail network studied even under optimal conditions. These optimal conditions were where the sales increase 214 was equal to or greater than the incentive level and no reduction in sales occurred following the incentive. Non— optimal conditions, which are likely to occur in an actual situation, produced results with an even larger negative impact due to the forward buy. Firm and channel management should carefully measure these impacts for their own operating systems and act accordingly. Hypothesis Two: The use of an incentive decreases the overall customer service level from the distribution center to the retailer for products under the incentive. The research findings indicated that the customer service level at the distribution center was improved when an incentive was introduced. This was due to the larger inventory carried during incentive scenarios. The comparisons between non—incentive and incentive scenarios resulted in a significant increase in customer service level. Within incentive scenarios, the degree of sales response increase caused a significant decrease in customer service level as that factor increased. The impact of sales payback was also significant, with customer service level increasing as the level of payback became larger. The changes in customer service level, although positive when taken alone, occurred due to the increase in inventory level in conjunction with the sales patterns examined in the study. Thus the cost of the inventory loading may more than offset any benefit derived from the customer service level 215 increase. In addition, the presence of an incentive could result in a decreased customer service level. This could occur where there was a sales increase exceeding the incent- ive level. Two scenarios examined in this study where this was the case did not produce this result, however. These scenarios depicted a 20% differential between the level of sales increase and incentive level. The level where negative customer service level ramifications occur must therefore exist beyond this point. In the industry examined in this research project the cost of stock-outs was considered relatively low, as consumers would purchase another product in most cases without any undue negative impact on the product stocked out. In other industries, where a relatively higher cost of a stock—out would be expected3 the cost impact might be substantially higher, if this were to occur.4 Under the conditions tested in this research, the use of incentives resulted in an improvement in customer service level. This would indicate that a limited degree of managerial concern and action should be placed on this logistical impact of incentive activity. This holds true under the conditions tested, however in other circumstances this might not be the case. Additionally the increase in customer service levels is reflective of the increase in inventory level, which should be a major concern of manage— ment . 216 Hypothesis Three: The use of an incentive results in increased variation in shipment sizes between the manufacturer and the distribution center. The research findings indicated that there was a significant change in the pattern of shipments as incentive activity was introduced. Incentive versus non-incentive scenarios resulted in larger shipments occurring as the incentive shipments were released. Within incentive scenar- ios the degree of sales increase generally did not produce a significant change in shipment size. Changes in the level of payback in incentive scenarios did in some cases produce significantly smaller shipment categories. The larger shipment sizes observed from non-incentive scenarios to incentive scenarios were due to the incentive shipments taking place at the beginning of the incentive periods. These larger shipments resulted in lower transportation costs overall as the lower rate breaks were encountered. This reduction in transportation costs due to the incentive would serve to offset the increase in inventory carried observed in Hypothesis One. The level of sales response within incentive scenarios in most cases did not significantly change the pattern of shipment sizes. It would appear that the initial inventory load due to the incentive made the major difference, where— as the change in sales response had little impact. This could be due to the fact that the initial inventory load was adequate to minimize shipment size differences due 217 to sales response within the incentive. The impact of payback levels was significant in the majority of cases. This response, which is opposite that observed with sales response, would also appear to be due to the timing of the change in sales level relative to the initial inventory loading. In the case of payback level the change in sales level occurred further out in time from the inventory load- ing than did the initial sales response. It would follow that given the depletion of inventory within the incentive, the change in sales level during the payback period would have a larger impact on shipment size, as the inventory level at this point would be running at a closer to normal level. The research findings would indicate that management should be concerned with the cost ramifications of shipment size due to incentive usage, which in this case was seen to result in lower costs. Possibilities exist for further cost reduction within incentive usage. These include the co- ordinating of incentive shipments either from the same manufacturer or from different manufacturers where the possibility exists for order consolidation. The level of sales response, based on the research findings, would be of lesser concern to management as its impact on shipment size within incentive scenarios was not as substantial. The degree of sales payback, however, should be of concern to management as its impact was substantial and resulted in negating the cost savings made possible by the incentive 218 shipments originally. In the industry examined in this research, transport- ation costs were relatively high compared to the product cost. Inventory costs were relatively low. In this situation the tradeoff between transportation and inventory costs was relatively larger favoring incentive usage. In a sit- uation where transportation costs are relatively low, this positive tradeoff would not be as great. Also, in the sit— uation where there are no cost reductions associated with larger shipments this benefit would be eliminated.5 Hypothesis Four: The use of an incentive results in an increase in the total number of ship— ments from the manufacturer to the distribu- tion center. The research findings indicated that although there was a substantial shift in the size of shipments, the total number of shipments remained relatively constant. The testing environment included a mix of products, some on the incentive, and others not subject to the incentive. As order sizes were increased for products under the incentive, in some cases regular shipments were included in this larger shipment. After this shipment took place however, reorders on non-incentive products occurred, resulting in close to the same number of total shipments. In the testing environ- ment, no attempt was made to plan ahead on non-incentive products in an attempt to consolidate shipments within the incentive shipment. In industries where incentive and non-incentive shipment order consolidation is possible, substantial cost 219 savings could be realized, given the results seen in this Hypothesis and in Hypothesis Three. This would have the result of increasing shipment size and reducing the total number of shipments at the same time. Although not observed in the present research environment, it is possible for the use of incentives to increase the number of shipments from the distribution center to the retail locations. If this were to occur it would be of additional concern in the cost ramifications of incentive usage. Hypotheses Five: The use of an incentive results in a decreased level of profitability for the distribution center/retail network. The findings of the research project indicated that the use of an incentive resulted in improved profitability when the market response to the incentive was optimal. This included testing conditions where the sales response was equal to or greater than the incentive level, and where no payback took place. This held true under both options test- ed. The first option depicted a scenario where the incentive discount was not passed along by the distribution center/ retail network. In the second option, the incentive discount was passed along. In the remainder of the scenarios tested, a dramatic change in results occurred. As seen in Table 5.1, when the incentive discount was not passed along, under the first option, profitability was still improved in all scenarios tested. When the incentive discount was passed along, under the second option, profitability was decreased in 2220 Table 5.1 PIIANCIAL PIIIOIIANCI SUHHARY Percent Change in Profitability From Non-Incentive Scenario OPTION ONE Testing Condition Low Uncertainty nigh Uncertainty LOU Uncertainty 2 OPTION THO High Uncertainty 20‘ Incentive, Zero Payback, 40‘ Sales Response Increase TC007/TC020 20! Incentive, 20. Payback. 40. Sales Response Increase TCOOJ/TCOIG 20‘ Incentive. 4o. Payback. 40‘ Sales Resoonse Increase TC°°‘/rc017 20. Incentive, Zero Payback, 20‘ Sales Response Increaee TC006/TC019 20‘ Incentive. Zero Payback. 20‘ Sales Increase TCOOZ/ICOIS 20‘ Incentive, Zero Payback.~ Zero Sales Increase TCOOS/ICOIO 40‘ Incentive. Zero Payback, 40! Sales Response Increase TC013/T0026 40. Incentive, 20‘ Payback, so. Sales Response Increase IC009/T6022 40‘ Incentive, cos Payback. 406 Sales Increase TCOIO/Tcozl 40‘ Incentive. Zero Payback, 20‘ Sales Response Increase rcoxz/rcozs 40‘ Incentive. 20‘ Payback. 20. Sales Response Increase ICOOB/TCOZI 40‘ Incentive. Zero Payback, Zero Sales Increase recii/rcozs 0055.70 +650.6 0502.9 +700.6 eb91.3 +650.0 O906.9 +012.5 9508. 1 +776.0 +501.2 *530.0 1. Incentive discount not passed on. 2. Incentive discount passed on. +6550.70 03951.00 02797.7 *3900.3 93206.0 .260b.9 +5170.J 06039.7 *3003.0 ‘6736.7 +3969.2 03349.6 #221.10 (56.1) (166.0) (2.2) (101.6) (206.5) 0232.0 (3.2) (100.1) (30.6) (192.6) (230.3) 0812.6 *360.0 (677.6) #306.7 (330.5) (794.0) «757.1 t562.63 (360.2) (559.1) (265.4) (763.0) 221 all less than optimal scenarios. In any situation where the sales response was less than the incentive level and/ or a sales payback occurred, a reduction in profitability took place. Even in one optimal scenario there was a slight reduction in profitability. This indicates strongly that the use of incentives, given the conditions tested and cost assignments made, are detrimental to profitabil— ity when less than optimal conditions exist and the incen- tive discount is passed along. This would further indicate that a firm being offered a trade incentive would be inclined to not pass on the discount. If the past experience of the firm were similar to the results identified in this research, it would real— ize that by passing on the discount, it would gain finance— ially only under optimal conditions. By not passing on the discount it would be assured improved financial performance. The results of the testing environment indicated that in addition to the significant difference seen in profitab— ility between incentive and non-incentive scenarios, there was a significant difference in profitability due to changes in the level of both sales response increase and payback. This would indicate that a great deal of management attention should be given to all aspects of trade incentive activity tested in this research project. This includes careful attention to price reduction and corresponding demand elasticity, sales forecast accuracy within the 222 incentive, and assessment of the degree of payback factor to be expected. As noted in previous hypotheses a number of considerat- ions must be made within each measure of logistical per- formance relative to their impact on profitability. A summary of these issues appears in Table 5.2. Implications of the Research Based on the findings of the present research, substan- tial logistical impacts occur when a trade incentive is utilized. These impacts have previously not been known. The results produced in this research project indicate that depending upon the exact policies employed and level of market performance attained, an incentive may or may not be beneficial to the firm and channel system involved. This represents a significant finding in view of the widespread use of trade incentives. Negative financial performance impacts as seen under many scenarios would possibly result in higher consumer prices over a period of time. This would conceivably put the channel engaging in these incentive activities at a competitive disadvantage. The short term gain in sales and market share might be totally negated by the longer term cost ramifications. The present research has indicated that under the conditions encountered in this industry setting that this phenomena may in fact occur. 223 .eeeoouao oc.ael so‘e.ueo no. e‘eeo neccecu use In.» -e30.s.vc. co co eves on cases: e.:u c.eu¢ .eucelaouueo ~e.u¢ec.. coeds. >__eo.ueaeso assaue.o es.uceuc. ecu no casescuoeeeo o» teas—es eco.uolaeee tee causes .0 _ese_ ech .n .eeeoosoo oc.nea co.e.ueo no. e.eeo neccezu ore [you neao.s.vc. ce :0 eoel en ounces eucelco‘eee «ecu .e.cu uo >u¢sou.ecee ecu uo eeseueo .euaseea useseuu.o u-«eauceueose c. u—sees v_aos euceeco.eee aeou c. eeoeecu .m .eeuao.u escesea use usbu ouc‘ teas-ocean eucelsouueo aeo.ue¢oou no eesaeees esone ecu bu sodas-see u—elesuae ees eoceluOuueo «saucec‘u queueeo .— .eoceluouueo neavsecau 053.33 ceases.- e.. 3300 euiZe no heal... «ego.- ecu .euceln.se eszceueTcoc use 2.3.39: eve-:38... bu eoes ees saloons se a. .uceIGanece arouses ecu c. aceuecou unee.ue-es vessels» ease-none uo seals: eah .— .eucesc.£e neon nosed .ueose- s¢ unseen snobs u‘ ee .oeeeeuus‘ as 0.300 avenue easy so souueo.u.les ue.ueec¢u es‘u.eon ecu .euoaQOuo segue no easesnase es.uceuc.0coc cu.s oeuec.oeoou sues encesn.ce esauceoe. u. .a .oece.c.a.o so vases avenue ea» «c co.ueo.u.seu .e.usec.u esauaeon ecu .ueso~ eues eueou co.ueuuoneceuu es.ue.eu ecu eueca say-soc. as c. .eueoo soaueuuooeceuu es.ueneu oc‘usoeu esca .ueoee- eleven eucelnage no en.e ecu eodueceUe esauceoc. secs: .~ .ueueeuo ea 3.6.. 2:232: e5 3 eat .ese- en‘saee seluesu gee-ecu se «0 seeds 7.2.2: e332... ecu neueeuo ees aeoo es» uu .ve.oaue seasons. can so. so— suesaue-es ees eusouvoue no case esp .a .ueueel.uueo uuueousecuu ea o—aou avenue sac» eusbaUOue uo seen seco.s e seas guesses. es en .usuvo 9.300 neseu euasuee selouesu c. couuusoeu e .es.uceos. no .eseu ecu sec» neueeuo snuee.uuuuse euea eeeesoc. eecoaees ee-ee ecu nu .- .ueueeuo on cases so‘ueouudleu easeluouuen ~euocec‘~ esaueoec ecu eueou ocouuueu >uOuceec. seao.c nu.) ee.uue:9c. on .n .es.uceoc. ecu a» sac eeeeuue. uuoucesc. betas: u-essueaeu e c. cusses unneAOuo cases can» uneasess. seso‘e e .uesocusu unbucesc_ sou: >~es¢ueueu we: oe.osue pageant. ech .- .Ehiifi eeeeaocu eeeesucu neuosdu ue~.l.u eeeeaoeo eaeeuoeo .‘eueso uaeueso uueueso «aeseso eeeeuueo eeeeuoeo ue~‘I.u eeeeueso eeeeuoc- seeeaocu uao._e seeeuocu eeeeuos— neu.lao ue~.s.u eeeesoeo eeeeeuea eeeeuoeu eeeeuosu ae«.l.n ue—¢I.n eeeeuueo eeeeuoeu as...» heaoe.u eeeeuuc- ue—.I.u eeeesoeu eeeesucu eeeeuocu . «euuseueosu n co.uflm - coouum OgaOuh: ~Q.0¢.C_‘ e»: .cn uo sen-:8 Iu¢flfluazu .0 Ou¢a eseJ eo.suon seIOueau Mad—Atom mmzHAmn—HDU JO easel-anus. .322...- eucsln.cm «0 sends: eucelnagu so soon gees; euusuen beluuesu seeeseesu .nua—mZOO ch QUOFUQs nacho .os.eeesoc~. husaeuaeoso no .esea ..s‘eeeuueu. esuuceocu nevus guesses .os.eeeuuc= e3 scene. . ea eeeoeees ee~eu .osaeeeuusu. esouseucu no nesea eoaueceoa es.useve. uses .es es.uceus- .sO hD¢SI~ 224 The implications of the findings of this research project are substantial given the widespread use of trade incentives. Their use has increased at a rapid rate in many industries without a corresponding increase in the knowledge of their logistical and profitability impacts. This research project has shown that the logistical impacts of incentive activity are substantial and that in many cases they result in reduced profitability for the firms involved. The primary implication of this finding is that the wide— spread use of trade incentives may be damaging to the firms and channels involved. The desire to optimize sales and market share in the short term has created a competitive environment where longer term gains may be sacrificed. These include the logistical impacts as translated into cost and profitability measures, and the expenditures and managerial effort that could be placed in other areas of the marketing mix. Based on the research findings, incentive usage would be most profitable in channels with a fast assured turn- over of inventory without any fear of shrinkage or obsol- escence. Lower costs of carrying inventory would also reinforce this improved profitability. Channels would benefit where there were cost savings with larger ship- ments, or where incentive shipments could be combined with non-incentive shipments. Favorable results of incentive usage would be more likely to occur in cases where the .5; W1»_ )3 225 transportation costs are high relative to inventory costs, thus producing a favorable tradeoff between the two. Firms that have a high cost of stockouts would also tend to realize improved financial performance with incentive usage, based on the research findings. The opposite result would hold true however if the sales response were to exceed the incentive level. In that case incentive usage would be most favorable where the potential damage of sales response exceeding available inventory would be less. Industries that have experienced relatively low amounts of consumer payback would also tend to perform better than industries with a high payback function. The accuracy of sales forcasts for the incentive are also vitally important to the successful usage of an incentive as it minimizes the costs associated with inventory load and possible stockouts. Accurate forecasting would also identify the degree of payback to be expected, which would help the firm to control for this variable, or not to participate in the incentive at all. Firms that carry pro- ducts with large margins relative to logistical costs would also tend to experience improved profitability with incentives as the potential for increased volume would more than offset the increased logistical costs. Unfortun- ately, as margins are to a large degree determined by inventory turn, this ideal firm would probably face equal negative impacts of incentive participation. Finally, 226 firms within a channel that can retain all, or a portion of the incentive discount would tend to realize improved performance. This, of course, places the cost burden on the manufacturer for a more limited sales response by the marketplace, which again raises the issue of the overall logic of incentive usage. Conversely, incentive usage would be least desirable for firms and channels which have traditionally slower moving inventory, high inventory carrying costs, little or no cost savings due to larger or combined shipments, high or unpredictable payback functions, inaccurate sales forecasts, lower margins, and the inability to retain all or a portion of the incentive discount. The research findings and their managerial implicat— ions have implications for the economy in total as well. Due to the use of trade incentives in many major indust- ries, it is entirely possible that there is an increase in total distribution costs due to this activity without any direct benefit to the economy. The findings of this research would support this possibility. Limitations of the Research The findings of the research are based on a wide range of testing conditions. These testing conditions although encompassing 78 years of simulated distribution activity are not exhaustive. Many possible combinations and levels of 227 input parameters, with their respective impacts on logis- tical performance were not examined in this research. In addition, many factors were held constant, in order to make the research design managable. These factors could be examined at different levels in various scenarios. The results obtained in this project are clearly limited to the nature and levels of the independent vari- ables utilized. These inputs were designed to provide results as generalizable as possible. Numerous other testing conditions warrant investigation. Future Research Within each of the hypotheses tested, numerous other policy and performance inputs warrant further research. This includes the levels of the variables utilized in the testing environment, in addition to the channel structure and processes which remained constant during the research. Due to the complexity of the problem these could not be incorporated in the present research. They, however, provide the basis for substantial future research. Within each of the hypotheses tested, numerous other independent variable, process, and structure inputs warrant future research. These include additional degrees and combinations of the level of the incentive, level of sales response to the incentive, level of sales payback after the incentive, and degree of demand uncertainty. A wide range 228 of process variables that were held constant during the research also warrant further investigation. These include incentive timing, inventory policies, and transportation system factors. Structural inputs that warrant further research include the channel echelon structure and the product configuration and mix. The final hypothesis relating to the financial impacts of the factors examined in this study perhaps offers the most expansive area for future research. Numerous additional cost and revenue assignments warrant future research. The possible combinations are limited only to the number of firms utilizing trade incentives. 229 CHAPTER V - FOOTNOTES 1These costs may be substantial; see: Douglas M. Lambert and James R. Stock, Strategic Physical Distribution Management. Richard D. Irwin, Inc. 1982. pp. 261-263. 2 Ibid. 3This would probably occur in higher involvement pur- chases and products with higher technical complexity; see: Harvey N. Shycon, "How Customer Service Impacts Sales Growth and ROI and How Distribution Can Help Increase Corporate Success.” Proceedings of the National Council of Physical Distribution Management, Annual Meeting, San Francisco, California, October 10-13, 1982. Volume II. 4The author has seen several instances in Ford Motor Company where an incentive with a low level of inventory resulted in severe dealer and customer relations problems. 5For example, shipping automobiles via truck. The discrete units shipped are so large that volume increases result only in additional full truckload shipments. APPENDIX A 230 APPENDIX A Trade Incentive Definitions 1. Case Allowances are discounts on each case ordered during a specified time period. Case allowance offers sometimes require a minimum purchase quantity. For example, a fifty-cent case allowance might be offered on a minimum purchase of one dozen cases. Case allowances are typically deducted from the manufacturer's invoice to the trade. Other types of allowance related to purchase quantity include: Freegoods allowances. ‘An additional amount of product is offered to the trade free, conditional upon the purchase of a minimum quantity. Carload and truckloadgprices. To reflect savings on freight and handling, a special price is offered on the purchase of the product in rail carload or truckload quantities. Often applicable to bulky products such as paper towels and tissues. Base contract allowances. A fixed off-invoice allowance is given on each order over a minimum quantity or dollar volume. Annual volume rebates. Usually designed so that larger percentage rebates are associated with larger purchase totals for the year. Independent retailers and wholesalers have formed buying groups to qualify for the higher percentage rebates. 2. Merchandising allowances compensate the retailer for prescribed merchandising support in the form of extra 231 in-store display or special featuring of the brand in the retailer's advertising. Some performance contracts specify the size, timing, and placement of the advertising feature or display required to qualify for an allowance. Like case allowances, merchandising allowances are usually offered as a percentage deduction from the list case price. Frequently, off-invoice case allowances and mer- chandising payments are usually based on the quantity of cases purchased during the promotion period. The preceding week may also be included to discourage retailers from letting shelf stocks run down in anticipation of the promotion. Unlike standard case allowances, merchandising allowances are not paid by manufacturers until they receive evidence of performance. The evidence might be in the form of a retailer affidavit, display photograph, or advertising tear sheet. .A qualifying advertising feature may range from a single line ”obituary” mention to a dominant position in the retailer's advertisement. By arrangement with the retailer, a manufacturer can sometimes have a cents-off coupon included in the advertisement. This guarantees a consumer price reduction at least equal to the coupon value. The manufacturer reimburses the retailer for the value of coupons redeemed plus the per-coupon handling allowance. Generally, merchandising allowances can be paid only to retailers since only retailers are in a position to perform the required service, whereas standard case allowances can 232 be paid to wholesalers as well as retailers. However, it is possible for merchandising allowances to be paid to a whole- saler when the form is acting as an agent of the retailer. For example, wholesalers like Associated Druggists, which puts together a common advertising program for the indepen- dent retailers it represents, can collect co-op advertising allowances from manufacturers as the -retailers' authorized representative. As an alternative or additional incentive to the retailer, a manufacturer may offer car stock, free goods which the salesforce distributes to qualifying accounts after the promotion period in return for merchandising services. 3. Financing allowances represent compensation to retailers for financial losses associated with the use of trade promotions or for their assumption of financing costs which the manufacturer would otherwise have to bear. These include: Floor stock protection. To prevent shelf stocks from running down before a promotion, the manu- facturer pays the retailer the case allowance on cases already in his warehouse at the start of the promotion period. Related to floor stock protec- tion is the count-recount procedure for deter- mining the cases which quaIify for the allowance. The retailer receives an allowance for each case of product moved out of the warehouse inventory during the promotion period, regardless of whether it was purchased before or during the deal period. Although expensive, this procedure permits rapid implementation of a promotion, since the trade does not have to await the arrival of products bought on deal. It also discourages the trade from merely buying for inventory during a promo- tion, and ensures that allowances are paid only on units sold to the consumer during the promotion 233 period. The count-recount procedure is also used when a manufacturer wishes to flush retailer inventories before introducing a new product package or before deleting a product. Early bookingand dating allowances. Used for products with seasonal demand, early booking allowances give customers per-case discounts for purchases made before a specified date. Dating lets the retailer defer payment for products ordered during the deal period, and effectively gives it an interest-free loan at the expense of the manufacturer's working capital. However, the manufacturers can produce more evenly over the year and saves on his warehousing costs once the product is shipped to the retailer. Cash discounts are usually presented as a sliding scale of off-invoice percentages, with the level depending on how soon the manufacturer receives payment. Frei ht allowances compensate retailers who are willing to collect orders direct from the manu- facturer's warehouse. Leaks and swells allowances are offered by canners to retailers as a means of simplifying faulty can claims. 4. Listing allowances for new products. Because retailers assume a risk in allocating their limited shelf space have to generate a new stockkeeping number, they may ask for a listing allowance from the manufacturer. Such allowances are usually tied to the number of cases placed on the shelves. Thus, to encourage rapid distribution, the manufacturer generally must offer a case allowance in addition to the listing allowance, because a listing allow- ance is typically a one-time offer, it does not tend to bring about competitive retaliation. 5. Sales incentives. These are incentives directed at trade personnel. They include dealer leaders for store and 234 department managers and contests for trade salespersons. A dealer leader is a sample of a premium displayed at the point of purchase until the end of the promotion, when it typically becomes the property of the store or department manager. Source: Adapted from the Grocery Products Manufacturers of Canada (1979, pp. 9-11) and the American Association of Advertising Agencies (1978). '1 APPENDIX B 235 .mLODm on» oOAwuso oumcfiofiuo Donn. owonu ecu can» ozuoouuo mmoa one 330on ou0um .3 e .um53mcoo may 0» umewEmM ouoe mum non» wuosoouo uOu nosoa ma mason o» omcoamou oak e .ocon wowuo cfl cofluosoou a Low“ coco noumouo :02: no wfimoo Hmcogoeouo cu oncoomom e .Hmoo ecu mo Uofiuoa ecu ocfiuso .ommnousa mo Hm>oH 23 c“ mommouocH Hmwucmumosm cw uaomou mason Hmcofiuoeoum e :ODDOO can comm .ocMcouwzm ocouo Eduw noumcfiofiuo ommouo new woamm on» own cozy moaned on many oasos cofluoeouo .w mo ocm on» us oowuma xomn>mm esp .>uw>fiuom ocflnoufi3m ocmun cmnu nocuou oanfiQxUODm umEDmcoo co comma on >wE monouocw moamm ecu mm .ooumumuo>o on awe Homo wooed a mo ufiuocoo one e .uo55mcoo on» can» undocumu ecu new umummuo nose no woman no uwoo onH .au0uce>cfi Dawoaon o» oouuweeoo woman no meant, on» am new about—m2; cw ooumo>cfi Hmuflmmo mo osao> ecu AH monocooeoo >umewuo ozu mm: >u0uco>cfi mo umoo one .muoEsmcoo con... mumoo 0:250: >u0uco>c« menu“: 96: muoafimuou mo mumwxo mafia—moo ooflum e ..nm. um .oumnuuoam mmcflocwm uofloz Amvuonus< mo>fiucoocH ocm mcofiuoeoum OD oncommom umxumz m xHszmm< 236 .ocmuc ce>uo e ou Hausa eues uecu emocu new cmcu eouua cu meoceco Ou :OMuoeeu necouc cone e oec moceuc ou He>oH uoc mugs: aafiEem .oeosoeu eouua eue xeaomuo co muosoouo umcu e>euaec ou eaaoea mowed mandamuo qu meouuo 30H qu couuuoeuu ecu uecu mueeooe uH .emcoomeu meaem co “woo use eowuo oaoceeucu m can Awwao use eouuo oeeo e ceesuec eoceueuwuo uCMOuMuCOum o: mu euech .maeaamwo euOue cu mo emece>wuoeuwe ecu one uosoOuo e no OMumu meuem Ou oeuwwuue>om ecu ceezuec umuxe Ou mueeaam aucmcouueaeu oz .aeuomuo euOum Cu cu emcoomeu uexueE oemmeuofi 5 muasmeu ensue—26m euecm uexumE umeuo m mec uoooouo ueH50uuueo co 0: euecs meusuosuum e>uuuueoeou .eoeum cusouo ecu aw eue uecu muosooua cecu aeaomuo encue a“ xc oemequCw meaem e>ec ou Muexwa euoe eue eaoao em: uosoouo ecu mo eoeum ensues e 5 eue uecu muooooum .00 cu eucm uoc mes couuoeouo cOucs .semucoceuu eeuuo: moceuc e oeuouuzc mo euoeomo mm ceem me3 oeumuuue>om . .mcueuueo onw>=c onu>wuooe Cu e>uuoewme ”XX. eum ecouuoaouo >Aucesvemcoo .cofiu 1050.3 m Ou ocommeu Ou xaexua uoc eue mue>sc 1..on oceum .muexso Eueu ocoH 3e: eosoouo uOc oo aecu ue>e3oc .mcfiwuuue>oe cecu memcoomeu ueueew pueu> ecouuoeoum .mue>:c HeaoHIcoc cmcu mafiaeeo Ou e>wmcoomeu weed eue mueasc Hm>oH oceum xcmum oce aemmez uewae>eco czoum 237 .moaocewsoc HemoHcoc cecu meuuw>uuoe ocwaeeo Ou e>wmcoomeu euoe ec Ou 0:30u eue3 euOum ce>uo e Ou Hm>oH euez uecu moaocemsom .moceuc mo uceEuuomme neouea e ue>o memmcousa afiecu euscuuueuo uecu emocu cecu meuuu>uuoe ocuaeeo ou e>uuumcem me uoc eue moceuc mo uecESC aHeEm e Ou memecousa puecu uwawa uecu moaoceesom .memmcousa oceuc eaouuuse acme eer uecu emocu me meuuu>uuoe ocuameo Ou e>uuwecee we ”:3. eue memecouso oceuc eaouuuss eme Eooaem uecu moaocemsom .mu0ue>occucoc cmcu new Iaeeo eouua ou e>umcoameu euos ec cu oc30u euez mu0um>0c¢~ ..uem oexo>e uuecu aucufs meocmco eowua ecu Cu e>uuumcem euomueuecu eue aecu uecu one socuousuom: eue xecu uecu e>ewaec has eouya cmwc ecu ocu>eo eue uecu endoceesoc ecu «oceuc oeouuo uesoH ecu mo c0uumeoe ecu an oeummuuee en mes oHocemsoc ocu>en eouuo 30H ecu mo mmeceogomcoo eouuo ecu uecu me>e3ec nocuse eceo .eouuo uezoa e onwaem mouocemsoc cecu ocuaeeo eouua Ou e>umcoomeu euoe euez eowuo uecouc e octane moaocemsom .muem: >>eec oo cecu oeuaeeo eowuo Ou mmeH ocoomeu uosooua e «o muem: ucOwH one sauce: .mueueeo uceoceerCu Ou cemoaoo we meMOue Cueco aw necOwc en Ou UCSOM me: auwouumeae eowum .meoexoeo oeuumlueouea uOu ueouea en Ou 0:30u mes >uu0wumeae eouum cceuoz 238 .ecoHe eco necuue cecu.He>eH uecOMc e um mu emcoameu ecu couumcficeoo cH .c0wuosoeu eouuo m.xHu cecu couuoeouo eumm mo ucuoo Ou He>eu uecouc e um ocooweu eueEsmcoo .eoceuc oeuo>eu Ou memecouoa HeuOu uuecu mo euecm neuHeEe e euo>eo oce moceuc ence mac Ou oceu eyesomcoo ecoua Heeo caum .oemmcouso uc:OEe ecu Ou oeueaeu >He>uueoec auemue>2u mu mmececouo Heeo A.meuuuuceso ueoueu euoum can emecouso Ou eace mu one ueoozc eHouerw euoe e mec emuzemsoc neouo ecu uecu no .eoeaouexuee ecu Cu maeeo uso coueem Ou 30c 9.65. can neaoocm uueoxe euOE e mu He: lou>uocH ueoao ecu moecuea uecu me>euaec nocuse ecHo .meco neocsoa cecu oeuumeo Ou ecouo euoe eum ee>u3emsoc neoao .OCumuuue>oe meoo cecu euecm oceuc co uerEu uecofic e e>ec .3er ucfia one ocuaeeo .ocfioflum .meuuu>wuoe oeuaeec Ou emcoomeu coo: uoeoeu uocHE m oec ewuzemsoc ecu uo msueum uceE>oHaEm .moaocemsoc ueouma pup cecu meuuu>uuoe oeuaeeo Ou mmece>uecoomeu uecOuc e oec moHocemsoc ueHHeEm .meuuuo oeuum sauces Cu moaocemsoc cmcu e>umcoomeu euos ewe: cuom .meuuuo eoumu aw mouocewsoc cecu meuuu>uuoe ocuaeeo Ou e>umcoomeu euoe ec Ou oeueeomm eczou Human cu moaocemaom .mmsouo eom neocso> HON cecu Amm ue>oo me>u3emsoc ueouo qu necOMc we: eeuuu>uuoe mauaeeo Ou meece>uec0omem wanna: new ocumoooz neumnez couxem APPENDIX C 239 APPENDIX C INITIAL CUSTOMER SERVICE LEVEL REVIEW FULL INCENTIVE SCENARIOS NON-INCENTIVE SCENARIOS WITHOUT SALES INCREASE ON PAYBACK PRODUCT/YR 001 (.3cfv) 014 (.7cfv) 011 (.3cfv) 024 (.7cfv) 1/1 .919 .840 .954 .912 2/1 .926 .837 .959 .919 3/1 .925 .844 .976 .923 4/1 .908 .836 .969 .894 5/1 .907 .808 .968 .891 6/1 .920 .839 .920 .839 7/1 .914 .850 .914 .850 8/1 .924 .844 .924 .844 9/1 .915 .811 .915 .811 10/1 .925 .841 .925 .841 1/2 .907 .848 .960 .906 2/2 .923 .828 .962 .901 3/2 .902 .823 .964 .910 4/2 .897 .817 .955 .894 5/2 .922 .817 .948 .893 6/2 .913 .844 .913 .844 7/2 .914 .836 .914 .836 8/2 .891 .821 .891 .821 9/2 .916 .812 .916 .812 10/2 .924 .861 .924 .861 1/3 .899 .866 .956 .912 2/3 .902 .841 .965 .909 3/3 .916 .814 .959 . .893 4/3 .906 .846 .967 .898 5/3 .909 .822 .963 .900 6/3 .896 .837 .896 .837 7/3 .900 .839 .900 .839 8/3 .912 .848 .912 .848 9/3 .891 .812 .891 .812 10/3 .910 .835 .910 .835 PRODUCT/YR 1/1 2/1 3/1 4/1 5/1 6/1 7/1 8/1 9/1 10/1 1/2 2/2 3/2 4/2 5/2 6/2 7/2 8/2 9/2 10/2 1/3 2/3 3/3 4/3 5/3 6/3 7/3 8/3 9/3 10/3 240 INITIAL INVENTORY LEVEL REVIEW (INVENTORY PLUS IN-TRANSIT) NON-INCENTIVE SCENARIOS 001 2892 3051 2957 2796 2839 2832 2998 2906 2915 2953 2639 2676 2583 2571 2646 2632 2709 2659 2691 2582 2720 2630 2718 2718 2578 2816 2703 2728 2562 2535 (.3cfv) 3711 3542 3610 3739 3578 3590 3574 3514 3602 3570 3848 3832 3796 3776 3784 3880 3760 3808 3736 3832 3832 3764 3780 3728 3820 3724 3860 3692 3800 3772 014 2740 2705 2746 2674 2592 2616 2848 2614 2564 2582 2648 2542 2497 2467 2569 2647 2563 2572 2552 2467 2661 2614 2595 2467 2452 2542 2459 2413 2450 2707 (.7cfv) 3803 3920 3908 3904 3940 3968 3823 3960 4004 3916 4004 3992 4072 4064 4060 4064 4052 4048 4052 4092 4004 3948 3940 4088 4152 3972 4032 4120 4148 3988 INCENTIVE SCENARIO WITHOUT SALES-INCREASE 011 12038 12255 12570 12096 11975 2832 2998 2906 2915 2953 11324 12263 11707 11847 12307 2632 2709 2659 2961 2585 11842 11593 11843 12094 11713 2816 2703 2728 2562 2535 (.3cfv) 3751 3695 3675 3673 3679 3590 3574 3514 3602 3570 3860 3864 3924 3940 3880 3880 3760 3808 3736 3832 3980 4048 4012 4080 4008 3724 3860 3692 3800 3772 024 10610 11250 9937 10138 10000 2616 2848 2614 2564 2582 10737 10731 10213 10166 9730 2647 2563 2572 2552 2467 10638 10659 11032 10318 9828 2542 2459 2413 2450 2707 (.7cfv) 4217 4289 4353 4337 4341 3968 3823 3960 4004 3916 4232 4424 4540 4372 4436 4064 4052 4048 4052 4092 4252 4268 4336 4344 4416 3972 4032 4120 4148 3988 241 nova. .DOBm h vmno. Mb AdHOH nooo. nvuo. Nb mmDOflo ZHBHH3 nho.~ hooo. naoo. N thOKU zmwzhun Ouhda h mmmdaom I‘m: mmfltaOm no law .h.o womDOm uUZ ho mHmhufld flUH>mum KHIOPWDU 242 voom. .nomm m ommo. ch ddHOH Nooo. ovao. Nb mmDQGU ZHIHH3 now. Hooo. nooo. N mADOGU ZENIHHO OHH¢M m mm¢ ho mumuqtzd oHo ZOHHHOZOU UzHHmmH DOHmum ”IMP um qm>mu MUH>¢mm KHIOHmDU 243 FQMN. .nOBm m hufio. vb 44909 nooo. ouflo. «h mQDOGU ZHEHHI ~0v.~ vooo. oooo. N mmDOGU ZMNBHNM Ouhdu h mnfltnOm Idflt mmm ho mHqutz< Mao ZOHHHDZOU UzHHmmH OOHmmm mink an Afl>mq fl0H>mmm KESOFmDU APPENDIX D 244 APPENDIX D NARGINS OP SELECTED FOOD ITEMS Baby Food Bakery Foods Baking Needs Beer and Wine Breakfast Foods Candy and Gun Canned a Dry Soup Canned Fish Canned Fruit Canned Juice Specialty Foods Canned Vegetables Coffee 5 Tea Dairy Products Desserts 5 Toppings Diet 8 Low Cal Foods Frozen Foods Household Supplies Nuts Paper, Plastics, Foil Pet Foods Pickles a Olives Sauces a Dressings Snacks Soft Drinks a Hixes Spices a Extracts Spreads & Syrups Tobacco Products General Merchandise Health a Beauty Aids 8 Margin 9.6 20.4 18.4 21.9 21.2 32.4 21.9 22.0 23.1 22.4 24.3 22.6 13.5 24.3 23.9 26.8 31.9 21.8 28.5 20.5 22.2 28.9 22.0 29.2 18.3 37.0 23.6 14.2 37.0 26.8 Assortment . At Warehouse 191 N/A 398 N/A 292 625 150 82 132 193 146 286 156 583 117 104 808 729 89 372 358 162 401 45 171 197 172 218 N/A 1993 6 of Super- market Volume .52 3.48 2.22 3.20 1.99 1.18 .72 .88 .57 1.32 .60 1.26 1.98 9.84 .26 .38 6.20 3.62 .39 3.70 2.13 .45 1.55 1.36 3.76 .37 .74 3.83 4.41 3.71 Source: Pro ressive Grocer, July, 1984, Vol. 63, no. 7, pp. - lye BIBLIOGRAPHY BIBLIOGRAPHY BOOKS Bartels, Robert. The History of MarketingThough . Second Edition (Grid Series in Marketing, 1976). Basic, 15:. Martin. Development and Application of a Gamma. Based Invgntory Management Theory. Ph.D. Dissertation, Michigan State University. (1965). Bowersox, Donald J., et al. Dynamic Simulation of Physical Distribution Systems, (1972). Bowersox, Donald J., Logistical Management, Macmillan Publishing Co., Inc. New York (1978). Bowersox, Donald J., et al. Simulated Product Sales Fore- casting, M.S.U. Business Studies (1979). Breyer, Ralph F., Quantitative Systegic Analysis fl Con- trol: Study No. l - Channel and Channel Group Costing. Philadelphia: College Offset Press. (1949). Chamberlin, E.H., The Theory of M_onopolistic Competition, Cambridge: Harvard University Press (1957). Forrester, Jay W. Industrial Dynamics, The M.I.T. Press and John Wiley and Sons, Inc. New York. (1961). Gold, Faye, et al. Modern Swermarket Operations. Third Edition, Fairchild Publications, New York (1981). Lewis, Howard T., James W. Culliton, and Jack D. Steel. The Role of Ai£_Freight in Physical Distribution. Graduate School of Business Administration, Harvard University (1956). Marion, Bruce W., et al. The Food Retailing_Industry-Market Structure, Profits, and Prices, Praeger Publishers, New York (1979). Nie, Norman H., et al. Statistical Package for the Social Sciences, McGraw Hill Book Company, New York (1975). 245 246 Orlicky, Joseph, Material Requirements Planning. McGraw Hill Inc. (1975)} Stock, James R. Strategic Physical Distribution Management. Richard D. Irwin, Inc. 71982). The Twentieth Century Fund, Does Distribution Cost Too Much? New York. (1939). Wood, Donald F. and James C. Johnson. Contemporary Trans- portation. Penwell Publishing Company. (1983). PERIODICALS “Channel Optimization in Complex Marketing Systems," Manage- ment Science, Volume 25, Number 10, October 1979. ”Computer' Simulation Study’ for' a iProduction Distribution System,” International Journal of Physical Distribution and Materials Management, Volume 13. "Consumer Response to Promotional Deals,” Journal of Market- ing, 42 (July 1978). ”Division of Labor is Limited by the Extent of the Market," Journal of Political Economy, 1951. "The Deal-Prone Consumer”, Journal of Marketing Research, 2 (May 1965). "Estimating Marketing Policy Effects on Sales of a Fre- quently Purchased Product," Journal of Marketing_ Research, 7 (August 1970). ”Increase in Sales Due to In-Store Display,” Journal of Marketing Research, 12:4 (November 1975). "Increasing Distribution Efficiency by Better Organized Research,” The Journal of Marketing, January, 1953. "Introducing the Marketing Channel to Price Theory,” Journal of Marketing, Vol. 28 (July 1964). ”Inventory Decisions by Trade-Off Analysis: A New Approach in Product-Oriented Marketing Strategies,” Journal of Business Logistics, Vol. 1, Number 2. ”Logistics System Modeling,” International Journal of Physical Distribution and Materials Management. Vol. 8, Number 4. "Market Segment Response to Marketing Decision Variables," Journal of Marketing Research, 11 (November 1974). 247 ”Measuring Promotions Grows in Importance,“ Advertising Agg, WA Model for Optimizing Retail Space Allocations,” Manage- ment Science, 27:7 (July 1981). ”Planning Physical Distribution Operations With Dynamic Simulation" Journal of Marketing! Vol. 36 (January 1972). "Room at the Top in Promotions," Advertising Agg, March 23, 1981. ”Sales Effects of In-Store Advertising,” Journal of Adver- tising Research, 15:3 (June 1975). "Sales Promotion-Fast. Growth, Faulty' Management,” Harvard Business Review 54:4 (July-August 1976). "Sales Response to Promotions and Advertising," Journal of Advertising Research, 14:4 (August 1974). "Short-Term Pricing and Dealing Effects in Selected Market Segments,” Journal of Marketing Research, 11:2 (May 1965). "Simulation-Tool for Better Distribution,” Harvard Business Review, Vol. 38. November-December 1960. ”A Theoretical and Empirical Evaluation of Price Deals for Consumer Nondurables,” Journal of Marketing 45 (Winter 1981). "When Do vo1ume Discounts Pay?" Purchasing Magazine. "When To Buy in Case Lot Quantities," £2, (November 1983). PAPERS Chevalier, Michel and Ronald C. Curhan, ”Temporary Promo- tions as a Function of Trade Deals: Descriptive Analysis,” Working Paper, Marketing Science Institute, Cambridge, Mass., May 1975. Crowell, J.F., “Report of the Industrial Commission on the Distribution of Farm Products (Washington, D.C.: Government Printing Office, 1901). Kingsbury, Steve. “Industry Trends in Promotion," presen- tation to SPAR Annual Seminar, Garden Island, Vermont, June 23, 1980. 248 Quelch, John A., ”Trade Promotion by Grocery Products Manu- facturers: A Managerial Perspective," Marketing Science Institute Research Program, (August 1982). Shycon, Harvey M. ”How Customer Service Impacts Sales Growth and ROI and How Distribution Can Help Increase Corporate Success.” Proceeding§_ of the National Council of Physical Distribution Management, Annual Meeting, San Francisco, California, October 10-13, 1982. Volume II.