MANAGING BRANDS AS A RESOURCE: A STUDY OF HOLLYWOOD FILM FRANCHISES By Thomas Daniel Chaffin A D I SS ERTA T I O N S ubm i t t e d to Mi c hi g a n S tate Uni v e rsity in pa r t i a l fulfil l ment of the re qui r e ments for the d e g r e e o f Business Administration Strategic Management Doctor of Philosophy 2015 ABSTRACT MANAGING BRANDS AS A RESOURCE: A STUDY OF HOLLYWOOD FILM FRANCHISES By T homas Daniel Chaffin In spite of the extensive research that has been carried out on the resource based view significant omission since the brands that are developed by a firm clearly fit the cri teria valuable, rare, difficult to imitate and with no substitutes that define resources that can form the basis of a sustainable advantage. Furthermore, RBV research has not given much consideration to the different strategies that a firm can deploy t o manage its brands as a resource. This dissertation draws on the literature in marketing to identify three different strategies that are most commonly used by firms to derive value from their brands. Brand extensions have received the most attention as it can be applied to a wide range of industries that offer products or services that consumers need to acquire on a regular basis. Beyond this, there has been some ts or services on a regular basis in order to incorporate the development of new features and to accommodate changing consumer preferences. The subject of this dissertation is brand re - creation, which has received little attention as it is mostly l imited to industries where each product represents a different brand and each of these have a relatively short life cycle. This is most commonly observed in the entertainment and leisure industries, such as motion pictures, video games and mass market book s. Because of the need to regularly introduce new products, firms can draw on a brand associated with a prior successful offering through the use of brand re - creation. This leads to the creation of what is referred to as franchise, which can extend the lif e of a brand acro ss several consecutive products . iv To Jill whose strength, faith and insight have made irrational ventures like this possible . v ACKNOWLEDGMENTS I wish to thank my committee: Bob Wiseman, Clay Voorhees and Don Conlon for their guidance, support and feedback. In particular , I would like to thank Jamal Shamsie for his help in the development of this manuscript and for his commitment to pursuing this line of research. Finally, I would like to thank my parents for their constant encouragement, listening ear and unfailing wisdom. vi TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ........... viii LIST OF FIGURES ................................ ................................ ................................ ........... ix CHAPTER 1 ................................ ................................ ................................ ....................... 1 INTRODUCTION ................................ ................................ ................................ .......... 1 CHAPTER 2 ................................ ................................ ................................ ....................... 5 THE EVOLUTION OF THE RESOURCE BASED VIEW ................................ .......... 5 Resources are tied to opportunity ................................ ................................ ................... 5 Resources are tied to sustainable competitive advantage ................................ ............... 7 Debate around the theory ................................ ................................ ............................... 8 Empirical approaches to the RBV ................................ ................................ .................. 9 Conclusions ................................ ................................ ................................ .................. 11 CHAPTER 3 ................................ ................................ ................................ ..................... 12 BRAND AS A RESOURCE ................................ ................................ ........................ 12 Characteristics of brands ................................ ................................ .............................. 12 Brands as valuable. ................................ ................................ ................................ .... 12 Brands as rare. ................................ ................................ ................................ ........... 13 Brands as inimitable. ................................ ................................ ................................ . 14 Brands as non - substitutable. ................................ ................................ ...................... 15 Managing Brands ................................ ................................ ................................ ......... 16 Extending brands to reach heterogeneous consumers. ................................ .............. 17 Revitalizing brands to satisfy changing consumer preferences. ................................ 20 Re - creating brands for single purchase cycle goods. ................................ ................ 23 CHAPTER 4 ................................ ................................ ................................ ..................... 27 RE - CREATING BRANDS IN THE MOTION PICTURE INDUSTRY .................... 27 Use of Sequels ................................ ................................ ................................ .............. 28 Timing of Sequels. ................................ ................................ ................................ ..... 29 Hypothesis 1: Time elapsed between sequels is negatively related to sequel performance. ................................ ................................ ................................ ........... 31 Hypothesis 2: There is a curvilinear shaped relationship between time elapsed between sequels and sequel performance such that the relationship begins negative and turns positive. ................................ ................................ ................................ ... 32 Number of Sequels. ................................ ................................ ................................ ... 32 Hypothesis 3: Sequel number is positively related to sequel performance. ........... 34 Hypothesis 4: The negative relationship between time elapsed and sequel performance is moderated by sequel number such that the performance will be much weake r for earlier sequels than for later sequels. ................................ ......... 35 Leveraging Product Level Similarity ................................ ................................ ........... 35 Hypothesis 5: The use of previous lead actors used in the prior movie is positively related to performance. ................................ ................................ ........................... 37 vii Hypothesis 6(a): The positive effect of previous acting talent decreases with time lapsed. ................................ ................................ ................................ ..................... 38 Hypothesis 6(b): the effect of previous actors on performance decreases with the number of sequels. ................................ ................................ ................................ .. 39 Leveraging Brand Level Consistency ................................ ................................ .......... 39 Hypothesis 7: The use of production companies used in the prior movie is positively related to performance. ................................ ................................ .......... 41 Hypothesis 8(a): The effect of previous production companies on performance decreases with the time elapsed between consecutive films. ................................ . 42 Hypothesis 8(b): The effect of previous production companies on performance decreases with the number of sequels. ................................ ................................ ... 44 CHAPTER 5 ................................ ................................ ................................ ..................... 45 METHODS ................................ ................................ ................................ ................... 45 Sample ................................ ................................ ................................ .......................... 45 Dependent variable ................................ ................................ ................................ ....... 46 Domestic box office receipts. ................................ ................................ .................... 46 Independent variables ................................ ................................ ................................ ... 47 Time e lapsed. ................................ ................................ ................................ ............. 47 Sequel number. ................................ ................................ ................................ .......... 47 Previous actors. ................................ ................................ ................................ .......... 47 Previous production company. ................................ ................................ .................. 47 Control variables ................................ ................................ ................................ .......... 48 Previous f ilm performance. ................................ ................................ ....................... 48 Production budget. ................................ ................................ ................................ ..... 48 Ticket price. ................................ ................................ ................................ ............... 49 Book based franchise. ................................ ................................ ................................ 49 Character based franchise. ................................ ................................ ......................... 49 MP AA rating change. ................................ ................................ ................................ 49 Star power. ................................ ................................ ................................ ................. 50 Unrelated sequel title. ................................ ................................ ................................ 50 Critic reviews. ................................ ................................ ................................ ............ 50 CHAPTER 6 ................................ ................................ ................................ ..................... 53 RESULTS ................................ ................................ ................................ ..................... 53 CHAPTER 7 ................................ ................................ ................................ ..................... 70 DISCUSSION ................................ ................................ ................................ .............. 70 Theoretical implications. ................................ ................................ ........................... 71 St rategic implications. ................................ ................................ ............................... 73 Managerial implications. ................................ ................................ ........................... 76 Limitations ................................ ................................ ................................ .................... 77 Conclusion ................................ ................................ ................................ .................... 82 REFERENCES ................................ ................................ ................................ ................. 83 viii LIST OF TABLES Table 1 Contextual Factors for Effective Brand Management Strategies ........................ 23 Table 2 Data Measures ................................ ................................ ................................ ...... 51 Table 3 Correlations and Descriptive Statistics ................................ ................................ 57 Table 4 Maximum Likelihood Regression Coefficients and Robust Standard Errors ...... 58 ix LIST OF FIGURES Figure 1: Predicted values of log domestic box office according to different levels of sequel number and 90% confidence intervals ................................ ................................ ............... 60 Figure 2: Predicted values of log domestic box office receipts according to different levels of previous production company and 90% confidence intervals ................................ ........ 62 Figure 3: Predicted values for time elapsed on domestic box office receipts with 90% confidence intervals shaded ................................ ................................ ................................ .......... 64 Figure 4: Predicted values of dome stic box office receipts based on varying levels of time elapsed and at low and high levels of sequel number shaded areas represent 90% confidence intervals ................................ ................................ ................................ ................................ ......... 66 Figure 6: Predicted values of deomestic box office receipts based on previous actor and sequel number at both high and low levels shaded areas represent 90% confidence intervals .... 68 1 CHAPTER 1 INTRODUCTION Existing formulations of the resource based view (RBV) have helped to shift the source of competitive advantage away from industry characteristics and towards firm attributes (Hoskisson, Hitt, Wan, & Yiu, 1999) . In particular, this framework has proposed that sustainable competitive advantage (SCA) can be attributed to the possession of resources that are valuable, rare, difficult to imitate and without substitutes (VRIN). Although resources generally lie within the firm, the benefits that can be obtained from any of these must be driven by their value in the market. Yet, most of the focus of RBV has been on supply - side resources such as human capital and intellectual property. As such, there has been little effort to focus on demand - based resources such as brands. In fact, e arly work on the RBV did suggest s can se rve as a valuable strategic resource which can enable them to successfully launch products and improve their performance (M. Peteraf, 1993; Wernerfelt, 1984, 1985) . However, m ost of the research on the use of brands as a resource has been carried out in marketing and has focused mainly on the development of brand extensions (e.g. Aaker, 1990; Aaker & Keller, 1990; Boush & Loken, 1991) . Brand extensions refer to leveraging an exi sting brand to introduce new products t hat may attract different market segments . Such a single focus on brand extensions has hindered attempts to identify different brand management strategies and to tie each of these to specific industry characteristics. In particular, little work has been carried out on how brand management strategies would differ across industries that exhibit differences in the purchasing patterns of consumers (Pringle, Wilson, & Brody, 1982) . For example, most consumers would tend to stick with a given brand in their 2 purchase of products such as soft drinks or laundry detergents which they buy on a frequent basis. However, it is not clear that these consumers would show similar loyalty to brands for products such as automobiles and smart phones which they buy on a less frequent basis. In this paper, I explo managed properly. As such I identify three different brand management strategies and tie their benefits to particular types of industries with an emphasis on demand driven differences i n consumer purchase patterns. Finally, the main purpose of my study is to focus on industries such as movies, books and video games where most consumers would only purchase a brand once. In these conditions, firms cannot maintain a continuous presence for all of their brands and must re - launch their brands on a regular basis in order to continu e to derive benefits from them. By doing this, I hope to help develop a better understanding of the strategic challenges that can be associated with managing brands across different contexts. In this work, I make three contributions. First, to research on the RBV I establish the brand as a key strategic resource by illustrating how brands possess VRIN based characteristics. To date , the RBV literature has largely focused on production based resources such as human and technological resources and has largely neglected how brands can serve as a strategic resource (see for example, Coff, 2002; Hall, 1992; Ployhart, Nyberg, Reilly, & Maltarich, 2014) . As a market based asset, brands can enhance firm performanc e (Capron & Hulland, 1999; Morgan & Rego, 2009; Rajendra K. Srivastava, Shervani, & Fahey, 1998; R. K. Srivastava, 2001) and by considering the VRIN based characteris tics of brands I begin explore how brands can be a source of advantage. Second, I make a contribution to the marketing literature on brand management by focusing on the use of brands as a strategic resource. In particular, I develop three different brand 3 management strategies a nd demonstrate how the effectiveness of each of these strategies is closely intertwined with demand - based characteristics of industries such as differences in consumer purchasing pattern. In this way, I show how the appropriate deployment of brands as dete rmined by industry context can play an important role in their effectiveness. Third I expand on the deployment of brands in the motion picture industry by expanding on a brand re - creation strategy which can be used in industries where consumers typically purchase any given product only once. I show how firms in such industries can introduce new products that carry over the brand from a previous product in the form of a franchise. In developing the specifics of a brand recreation strategy within the motion picture industry I manage to show the importance of the timing between the introduction of successive products and the number of films to date in the franchise. I also touch upon issues associated with the concepts of product level similarity and brand con cept consistency in maintaining the value of the brand in such industries. I find support for the notion that the timing of sequel products plays an important role in how brands impact performance in the motion picture industry . Moreover, I find that quic kly releasing sequel films can enhance performance early in the franchise while more time between sequels can be beneficial as the franchise matures. Secondly, I find that product level similarity in the form of the use of consistent lead actors and brand concept consistency in the form of the use of the same production companies can enhance brand performance across sequels. Finally, due to satiation and brand wear out the positive impact of th ese forms of similarity and consistency is contingent on when the sequel is released and where it falls within the franchise such that their positive impact is reduced as the time between sequels increases and the franchise matures. 4 In summary this work suggests that brand management strategie s can play an important strategies is contingent on demand - based industry characteristics such as consumer purchase patterns . Brand re - creation specifically is an appropriate brand management strategy in industries where pro ducts have a short product life - cycle and consumers generally purchase the product only once. Furthermore, managers in these types of industries can expand upon the benefits of a brand re - crea tion strategy through decisions regarding the time between different products and the number of new products to introduce. The organization of this dissertation is as follows. Chapter 2 provides a review of the evolution of the RBV. Chapter 3 considers br and as a resource which can be managed through extensions, revitaliz ation and re - creation. Chapter 4 applies brand re - creation to the Hollywood f ilm industry and provides arguments for the h ypotheses that will be tested. Chapter 5 outlines the methodology for testing these hypotheses. Chapter 6 includes the results from the analysis and Chapter 7 includes a discussion of the results and conclusions. 5 CHAPTER 2 THE EVOLUTION OF THE RESOURCE BASED VIEW The resource based view represents a collection of literature within strategic management which theoretically focuses attention inside the firm for creating and maintaining value. Early work on the RBV focus ed more heavily on an understanding of how resources can help firms to pursue various opportunities (Penrose, 1959; Wernerfelt, 1984) . Later work focused on how specific resources could be viewed as a basis of sustainable competitive advantage (J. Barney, 1991; M. Peteraf, 1993; M. A. Peteraf & Barney, 2003) . Over time, t he RBV has undergone both criticism and refinement (J. B. Barney, 2001; Foss & Knudsen, 2003; M. A. Peteraf & Barney, 2003; R. L. Priem & Butler, 2001a) as well as nume rous empirical tests with somewhat mixed results (Arend, 2006; Armstrong & Shimizu, 2007; Crook, Ketchen Jr., Combs, & Todd, 2008; Newbert, 2007) . Each of these developmental phases has played an important role in the formation of the RBV as a dominant perspective within strategic management research (Nag, Hambrick, & Chen, 2007) . The following sections provide a review of these phases illustrating how each of them shed light on differing aspects of firms, their resources , and how the firm creates value within a market. Resources are tied to opportunity Early work in the RBV was pioneered by Penrose (1959) . Trained in economics , Penrose focused on how a better understanding of its resources could allow a firm to figure out how they could be deployed to pursue various opportunities. As such, her work was designed to investigate through its ability to pursue these opportunities . Becau how managers could understand the different ways in which resources could be combined to 6 pursue opportunities. Consequently, Penrose laid considerable emphasis on the lear ning of managers so that they would be able to link resources to opportunities (Penrose, 1959, p . 85) . At the same time, her work placed considerable emphasis on the bundles of resources that could be created by the firm rather than on any single resource. Wernerfelt (1984) followed up on these ideas several years later. Based on the deployment of resources to pursue specific o pportunities, he suggested that they were fundamentally connected to a firms choice of products . Or , managers , resources and products are two sides of the same coin (Wernerfelt, 1984, p. 171) . Resources in this sense offer heterogeneous benefits based on their differing characteristics. Therefore, according to the RBV, firms differ in their approach to opportunities and growth because of their heterogeneous resource base and the benefits that t hey choose to obtain from them. Early conceptions of t he RBV therefore emphasize the interplay between resources and markets in understanding how firms exploit opportunities . B oth Penrose (1959) and Wernerfelt (1984) a lso acknowledge the importance of time in both the development and exploitation of resources. As such, res ource development and exploitation was viewed as an emergent process resulting from an interplay of market opportunities and productive resources to develop unique knowledge within the firm (Garnsey, 1998) . For example , Wernerfelt (1984) illustrated the RBV through a market sequence matrix in which firms developed capabilities in one market which were then exploited in new markets. Early work on the RBV th erefore emphasized how firm growth resulted from the accumulation of firm based knowledge that allowed them to continue to link market opportunities to their resource deployment over time. Notwithstanding these remarkable 7 contributions , to this point , the RBV remained largely amorphous and without specific parameters for theorizing. Resources a re tied to sustainable competitive advantage While early conceptions of the RBV focused on the application of resources to pursue various opportunities, subsequent formulations f ocused on how resources serve as a source of sustainable competitive advantage (SCA). Building on early developments with the RBV, Barney (1991) began parameterizing the perspective into a theory of SCA. Barney proposed that characteristics : they were valuable, rar e, could not be imitated and did not have substitutes. had been developed by Penrose (1959) and Wernerfelt (1984) . The key attribute of a resou lack of imitability of any resource was tied to the specificity of a resource or its ties to a given firm. Beyond this, it shifted the focus of the RBV to the attributes that were tied to each resource. Closely associated with his work , Peteraf (1993) further developed the parameters by connecting the RBV to the notion of Ricardian rents. Specifically , she elaborated on focus on the advantages that a firm could derive from its possession of particular resources. Clearly a SCA was derived from a heterogeneous distribution of resources across firms. Furthermore, this heterogeneity could only be preserved under cond itions where firms did not face competition in securing the resource and could subsequently prevent its rivals from obtaining or developing similar resources . , Peteraf (1993) also argued that a SCA would depend on relative lack of mobility of resources between firms. This extended the focus of the RBV 8 beyond single resources to a complementary set of resources, as Pet eraf suggested that a particular resource may be more valuable to a given firm because it could match it with other resources that it also possessed. For the most part, t he later contributions b y Barney and Peteraf moved the RBV toward more formal theoriz ing, in the form of specific parameters for formal propositions and testing. In addition to formalization , these researchers focused on SCA, outlining not only how resources enable the exploitation of opportunities but also how resources can be a source fo r continuing to secure benefits from exploiting th e se opportunities. Debate around the theory The formalization and parameterization of the RBV was not without its critics. In their critique of the theory , Priem and Butler (2001a) argued that the RBV had fallen short of being a theory due to a lack of law - like statements, overly general parameters and tautology. In addition, Foss and Knudsen (2003) suggested that the RBV as developed by Barney (1991) and Peteraf (1993) suffere d from overly broad definitions and a lack of parsimony and clarity. These critiques prompted a lively debate among organizational scholars based on the merits and usefulness of RBV as a theory of competitive advantage (J. B. Barney, 2001; M. A. Peteraf & Barney, 2003 ; R. L. Priem & Butler, 2001b) . Each of these developments led its proponents to clarify the theory, definitions, dependent variable and scope of the RBV. clear definition of value. In particular, there has been a lack of sufficient distinction between the attributes and the benefits of resources, suggesting that the theory cannot be refuted (Kraaijenbrink, Spender, & Groen, 2010; R. L. Priem & Butler, 2001a) . It has been proposed that this issue could be addressed by relying on different measures of the value of a resource based on 9 its abi lity to generate customer perceptions (Bowman & Am brosini, 2000; Hoopes & Madsen, 2008; Schmidt & Keil, 2013) . A related criticism of the RBV focuses on its assumption that possession of VRIN resources is sufficient to provide the firm with a SCA. Unlike Penrose (1959) , Barney gave little consideration to the role of mangers in identifying and deploying resources. In fact, the ogenous factors has been criticized (Makadok & Coff, 20 02) . Subsequent work has incorporated the role of managers in generating re venues from resources (Holcomb, Holmes Jr., & Connelly, 2009; Kor & Mahoney, 2005; Sirmon, Hitt, & Ireland, 2007) . Some recent contributions have pushed for more work on the conditions which allow a firm to use its resources to create a SCA (Becerra, 2008; Foss & Knudsen, 2003) . Uncertainty, for example, is essential for the limiting of competition which has been proposed by Peteraf (1993) . Similarly, the many factors that increase the immobility of resources must be given more consideration , since they may be a key deter minant of SCA. In general, further work on the RBV has been shifting to how resources are deployed by firms in a manner that can provide them with a SCA. This has led to a growing interest in how combinations of resources can create resource complementary or resource specificity. It has also , in fact , help to establish how firms may use their resources to generate value. Empirical approaches to the RBV Amidst the continuing t heoretical refinement , the RBV also underwent significant empirical testing. Broadly, one of the fundamental questions focused on the locus of the variance in firm performance. There were a number of studies which partitioned firm variance into 10 industry an d firm components (S. - J. Chang & Singh, 2000; McGahan & Porter, 1997; Schmalensee, 1985; Short, Ketchen, Palmer, & Hult, 2007) . In general , these studies align with the RBV , suggesting that most of the variance in firm performance is at the firm level when compared to industry or strategic group. Although many different studies have been carried out to test the RBV, support for it has been somewhat mixed as reflected in recen t surveys of this research (Acedo, Barroso, & Galan, 2006; Armstrong & Shimizu, 2007; Crook et al., 2008; Lockett, Thompson, & Mo rgenstern, 2009; Newbert, 2007) . To a large extent, these mixed results reflect several underlying issues which have continued to confound researchers . To begin with, there are differences in the manner in which resources are defined. According to Newb ert, (2007) , researchers focused on different aspects of resources in their assessment of their links to firm performance. To begin with, there was no clear distinction between what could be defined as a resource and what co uld be defined as a capability. Next, many studies focused on specific resources rather than on a broad set of resources. Finally, studies differed on the specific attribute value, rareness, inimitability that they used measure the effect of a resource. In this regard, a review of research by Crook et al . (2008) was much more helpful. It found more support for the RBV when they focused on studies that focused on resources that more clearly conformed to the specified criteria. In other words, they found the results to be weaker when studies f ocused on resources that were not clearly valuable, rare, inimitable or without substitutes. Furthermore, the support was stronger for results that had been unaffected by issues that were tied to appropriation of revenues (Coff, 1999) . 11 However, the most serious problem with the RBV lies with its lack of attention to the manner in which SCAs are created as the result of the processes that are used by firms to create and deploy their strategic resources. Dierickx and Cool (1989) , for example, advanced some ideas about the factors around resource creation and deployment that could explain their lasting effect on performance. In their review of RBV studies, Armstrong & Shimuzu (2007) have suggested that the processes that tie resources to performance must be more carefully studied in order for us to gain a better understanding of the contribution of RBV. Conclusions In spite of these different issues, the RBV has moved the field of strategy into several new directions. These can be broken down into two different categories: resource - centered and capa bility - centered. In the resource - centered stream, RBV has led to a growing stream of research on the contribution of different types of resources. In this regard, there has been growing interest in knowledge - based resources, otherwise known as the knowledg e - based view (Grant, 1996) . As Kraaijenbrink et al . (2010) have proposed, however, there are many different ways in which resources can be characterized, based on their relevance to the RBV. The most commonly used have been criteria such as financial, reputational or human; tangible or in tangible; fungible or non - fungible. In the capability - centered stream, RBV has led to a greater focus on firm capabilities, including dynamic capabilities (Teece, Pisano, & Shuen, 1997) . This research has been probing into the way that managers in organizations may learn how to combine resources and deploy them to pursue specific opportunities. Together, these two streams of research may allow us to combine the perspectives of RBV that have been advanced by Barney (1991) with the one that had been laid out earlier by Penrose (1959) . 12 CHAPTER 3 BRAND AS A RESOURCE According to the RBV, firms derive advantages from the possession and utili zation of their resources. In particular , RBV scholars suggest that these advantages stem from the heterogeneity of resources held by different firms. To the extent that specific resources are attached to different firms on a semi - permanent basis , they can be the source of SCA . However, Barney (1991) suggested that resources can provide such advantages only when they can be considered to be valuable, rare, inimi table and without substitutes. Characteristics of brands Brands as valuable. ident ity for the products and services of a particular firm that also separates them from those offered by its rivals (Aaker & Keller, 1990; Boush & Loken, 1991; Sood & Drèze, 2006) . Generally brands serve as a valuable resource because they help the firm to develop a degre e of loyalty from consumers. Brands serve to develop loyalty by performing three important functions namely, making consumption more efficient through search processes, user skills and by shaping consumer preferences. First, brands create loyalty by facili tating efficient consumption through lowering search costs for consumers. Consumption markets can be highly complex and evaluating all products and services on all characteristics can be cognitively demanding, particularly for those products that consumers buy on a frequent basis (Shamsie, 2003) . For example, consumers would not have much moti vation to engage in extensive search every time that they buy soap, toothpaste or toilet paper. Therefore , brands can serve to speed up consumer decision making based on the 13 prior performance of their brand, thus making evaluation more efficient for consum ers (Nelson, 1970) . Second , consumers are more likely to stick with a given brand as they learn more about its uses and benefits. Consumption requires consumer effort to produce benefits (R. L. Priem, 2007; Ratchford, 2001; Stigler & Becker, 1977) . As consumers become more familiar with a , they can learn more about how to derive the maximum possible benefits from its use. For example, repeated use of a particular spreadsheet software can provide consumers with mastery of its use , allowing them to perform many different functions. Firm specific user skills can therefore enable consumers to engage in consumption in such a way that they are more efficient and more likely to realize benefits from the (Wernerfelt, 1985) . Third, brands are valuable inasmuch as they shape consumer preferences. Consumer preferences to some degree are ambiguous. In some markets , consumers may be uncertain as to which factors should play an important role in purchasing decisions. Firms can try to reduce this ambiguity by emphasizing those characteristics that make their products more attractive to consumers. In this way , firms can persuade consumers to make their purchase decisions more easily by drawing them to the spe cific characteristics their brand. For example, Heinz has emphasized the thickness of its ketchup to differentiate it from other competitive offerings. These associations between brand and product factors can serve as important mechanisms shaping the consu mers (Fischer, Völckner, & Sattler , 2010) . Brands as rare. Although firms have the capacity to create many brands , it is difficult to develop a loyal customer base for each of them. Brands that have value because of the following 14 that they have already developed in the market tend to b e rare. In other words, while brands in general do not appear to be rare, those that have already established themselves are certainly heterogeneously distributed, and names with high levels of favorable recognition are certainly rare (Capron & Hulland, 1999) . In addition, brands are rare because of the limited number of brands in a consumer s consideration set. Purchase decision making is conceptualized as a multi - stage process where consumers restrict the number of brands to be considered in a purchase. A consideration set represents the brands a consumer is willing to evaluate when making a decision (Roberts & Lattin, 1991) . Consideration of a number of brands requires cognitive effort and , because consumers seek to economize cognitive effort , they are likely to reduce the total number of available brands to a select few (Roberts & Lattin, 1991) . Because the potential benefit of additional consideration of brands offers diminishing marginal utility , consumers are likely to be highly selective in the number of brands und er consideration for a given purchase (Roberts & Lattin, 1991) . These market con straints enhance the rareness of brands as a strategic resource. Brands as inimitable. makes it a strategic resource. In accordance with the RBV, brands are difficult to imitate because t hey are usually connected to the idiosyncratic resource base of a particular firm. Brands represent an asset that is created and supported by the various competencies of a firm. In other and the resources needed to produce these products makes brands idiosyncratic and therefore more difficult to imitate. In addition, brands benefit from time compression diseconomies (Dierickx & Cool, 1989) . Building brands requires repeated investments over a long time period (Anand & Delios, 2002) . Establishing a positive image in the minds of a high number of consumers requires large, 15 repeated investments. The accumulation of positive brand images , therefore , can generate time compression diseconomies b ecause other firms cannot quickly imitate them with their own set of resources and capabilities (Dierickx & Cool, 1989) . These time horizons , coupled with the repeated investments , also enhance the inimitable nature of brands as a resource for firms. Brands as non - substitutable. Brands have been shown to have a number of characteristics that make them non - substitutable. The use of brands can also serve an important source of information to reduce the uncertainty when purchasing new products. In order to reduce the risk disappointment associated with pur chasing new products , consumers often rely on the brand associated with the product (Claycamp & Liddy, 1969; Milewicz & Herbig, 1994) . Associating a product with the brand can enable consumers to select products that are more likely to meet their needs and to have the appropriate expectations (Basuroy & Chatterjee, 2008) . Other appr oaches such as product features or distribution are unlikely to have similar risk reducing properties. In addition, brands can serve as an efficient market signal, in that they can communicate a high level of complex information very concisely (Wernerfelt, 1988) . As a market signal, brands allow a consumer to associate a product with specific attributes. Furthermore, by continuing to purchase the same brand , the consumer is assured of the consistency of these attributes. This feature of a brand, particularly one that is already well known to the con sumer, makes it difficult to find substitutes for it. Ba sed on the RBV resources are strategic only when they demonstrate VRIN type characteristics in that they are valuable, rare, inimitable and with no - substitutes. As noted above, brands offer a number of strategic characteristics; however strategic management research has 16 also discussed that competitive advantage arises not only from the possession of resources but also from their appropriate management (Sirmon et a l., 2007) . Managing Brands Early conceptions of the RBV generally focused on the characteristics of resources, which assumed that the mere possession of strategic resources made their appropriate use self - evident (J. Barney, 1991; M. Peteraf, 1993) . Later work challenged this assumption, suggesting that how these resources are managed, combined and deployed played an important role on their leading to SCA (Sirmon & Hitt, 2003) . Therefore, current work in the RBV suggests it is not only the possession of strategic resources, but also their appropriate management which influences SCA (Sirmon et al., 2007) . Even though some of the RBV research has identified brands as a strategic resource (i.e. Helfat & Pete raf, 2003; Wernerfelt, 1984) and there have been frequent calls for integrating strategic management and marketing (J. Barney, Wright, & Ketchen, 2001; R. K. Srivastava, 2001) there has been little theoretical and empirical work on the use of brand as a resource carried out to date. Combining the RBV literature with the marketing literature is critical to un derstanding how firms can manage their brands for SCA. While there has been substantive work about brand management strategies in the marketing literature (Aaker, 1990; Kapferer, 2008; Keller, 1999, 2008; Low & Fullerton, 1994) to date this work has not fully considered how the effectiveness of these brand management strategies vary by contextu al factors like industry characteristics like product life - cycle. In terms of brand management, the marketing literature has mostly concentrated on the use of brand extensions (Aaker, 1990; Keller & Aaker, 1992; Spiggle, Nguyen, & Caravella, 2012; Völckner & Sattler, 2006) . Brand extension as a brand management strategy refers to 17 using established brand names to introduce new products to reach new consumers. The core product of a brand may not appeal to all consumers therefore to broaden the reach of the brand; brand extensions enable firms to reach new consumer groups with new pro ducts. In addition, there has been some growing reference to the concept of brand revitalizatio n (Berry, 1988; Brown, Kozinets, & Sherry, 2003; Gilmore & Wansink, 1999; Thomas & Kohli, 2009, 2009) . Brand revitalization refers to reinvigorating the br and for existing consumers. Brands can lose their consumer appeal as brands and consumers age (Bivainiene, 2010; Simon, 1979) . Therefore revitalization represen ts the firms efforts to maintain the efficacy of the brand for existing consumers (Berry, 1988) . In spite of this progress, there has been little effort to link different brand management strategies to specific industry contexts with different consumer purchasing patterns. Furthermore, these existing strategies do not cover all types of industries such as movies, books and video games where managing brands can be challenging given that most consumers do not purchase any single product more than once. The following sections provide a more in depth assessment of various b rand management strategies. Extending brands to reach heterogeneous consumers. Across a wide range of industries that offer products consumers buy on a regular basis, firms build on the appeal of their most popular brands through the use of brand extensio ns (Aaker, 1990) . Brand extensions are most commonly used to reach new consumer segments that are not attracted to its existing brand. For example, Coca - Cola has expanded the appeal of its flagship brand to other segments that prefer drinks that are lower in sugar or in caffeine content. Such a strategy allows the firm to draw on the strengths of its core brand resources while modifying the product to appeal to a wider range of consumer preferences. 18 When effectively executed, the attributes of th e parent brand are transferred to the new product extension, thus enhancing its position relative to competitor products (Farquhar, 1989; Sood & Drèze, 2006) . Strategically, brand extensions provide an opportunity for firms to utilize their market based resources to efficiently expand their customer base. To accomplish this, brand managers navigate the difficult balance of product similarity, in order to leverage and maintain brand identity, while differentiating the product su fficiently, so as to secure a net increase in sales rather than cannibalizing existing products. When an extension product is too similar to its parent brand, brand extension strategies can backfire. For example, when Gillette sought to offer a lower end s having cream to compete with Barbasol they regrettably used the Gillette name, simply opted for the Gillette product at the lower price (Aaker, 1990) . Thus brand extension strategies are challenging as firms expand the consumer base through brand extensions while maintaining the identity of the parent brand. Brand extensions can also be used by firms to offer products that fall into other industry categories. Like traditional extension, in cross industry brand extensions, firms seek to transfer some of the positive characteristics or associations to products that they can introduce into other industries. In order to be successful in cross industry brand extens ion, parent brands should have brand associations that are salient and favorable in the extension context because even if the parent brand is strong, extensions are often negatively evaluated when similarity between the extension and parent brand is low (Keller & Aaker, 1992) . This can be exemplified in the efforts of Zippo to diversify away from smoking centered products. Their introduction of fragrances was a flop because of the negative association of lighters t o perfume (Hagerty, 2011) . Therefore 19 successful brand extensions need to have some level of similarity between the parent brand and extension. In addition to brand characteristics and fit, effective brand extension strategies also require effective market communication. Both marketing support and distributor acceptance play important roles in the effectiveness of brand extensions. Marketing support refers to advertising and distribution strategies which can enhance the perceptions of fit while enhancing cooperation among distribution partners (Collins - Dodd & Louviere, 1999; Reddy, Holak, & Bhat, 1994; Völckner & Sattler, 2006) . While brand extensions lower the overall expen ses associated with launching new products when compared to launching a new brand, extensions also require effective marketing support. Therefore allocating the necessary resources to launch the brand extension is critical to its effectiveness. Brand exten sions provide both risk and opportunity for brand managers. Brand extensions provide an opportunity for firms to both reinforce and extend the reach of their brands, therefore successful extensions can actually strengthen the parent brand. However, unsucce ssful brand extensions can also dilute the parent brand (Keller & Aaker, 1992) . For example, when Sprite introduced other related products such as Sprite Orange, the parent brand was significantly weakened (J. W. Chang, 2002) . Thus brand managers seek to choose brand (Spiggle et al., 2012) . In summary, brand extension strategies are of particular value to firms transacting with heterogeneous consumer markets whereby the positive attributes of parent brands are leveraged to introduce new products. In general, brand extensions are more prevalent in industries where it is not too costly to introduce new products and where consumer preferences are varied but stab le. 20 The key issue in the use of brand extensions is to find the balance between the need for similarity with the core brand while creating some meaningful differences in the brand extension. Revitalizing brands to satisfy changing consumer preferences. Con textual factors such as changing consumer preferences can play an important role in how firms strategically manage their brands (Carpenter & Na kamoto, 1989) . Consumer preferences refer to the priorities consumers place on certain benefits from the products they purchase. These preferences are likely to vary over time for certain product categories due to changes in cultural values, consumer li fe stage and technology. The degree to which consumer preferences are stable can play an important role in how firms manage their brands. For some product categories, such as products have a long stable value proposition. When consumers have stable preferences and technology plays a lesser role brands can be established and exploited over longer periods of time. A brand revitalization strategy is effective in product markets wh ere consumer preferences change over time. As consumer preferences change the focal brand can undergo a life - cycle, much like products, where brands emerge, become well known and then begin to age (Bivainiene, 2010; Simon, 1979) . In this context can be critical for firms to engage in a brand revitalization strategy where the brand is consistently renewed. Brand revitalization can take many forms such as new produ ct models to satisfy changing needs or marketing tactics such as promotions and advertising to re - instill the brand values. Not only can revitalization strategies vary in their tactical approach but also in their temporal pacing. For automobiles, these ch anges are carried out on an annual basis in order to incorporate the most recent features and reflect style changes. In other industries, such as clothing, changes may be carried out two or three times each year in order to match the offerings 21 to the seaso n or time of the year. Brand revitalization strategies then incorporate marketing tactics such as new products or advertising to reconnect the brand with consumers and ensure that consumers have positive experiences related to the brand. Brand revitalizati on differs from extension as extension focuses on reaching new consumers whereas revitalization is designed to hold on to current consumers. Brands with a long period of market exposure can begin to decline when consumers experience dissatisfaction with th e brand. Differences between the brand promise and consumption experience can generate dissonance for consumers, thus weakening the meaning of the brand within the market (Anderson, 1973) . To rejuvenate brand meanings, firms engage in product innovation to reinforce brand meanings and enhance consumption experiences. For example, while Cadillac historically enjoyed strong brand meanings, over time it became less impactful in the market as GM struggled to compete with Japanese and German competitors. GM, however, introduced new models such as t he CTS, the STS, and the DTS, which all helped to revitalize the Cadillac brand (Ireson, 2008). In large part, t he impetus for brand revitalization arises from changing consumer preferences which can be catalyzed by moves of competitors (Thomas & Kohli, 2009) . In many industries, competing firms develop a pattern of updating their products or services on a regular schedule. For example, Apple and Samsung introduce new smartphones in order to catch up with each other in terms of their offerings. In other industries, firms may need to re - orient their brand occasiona lly, rather than regularly, in order to accommodate what may be longer term trends in consumer preferences. The dimensions by which consumers distinguish value can vary over time, which can require that firms make changes to their offerings to respond to t hese shifts (Adner & Zemsky, 2006; Danneels, 2011; Tripsas, 2008) . Similarly , the growing interest of 22 many consumers for fresher and healthier food has pushed McDonalds to add salads and offer fruits. By moving in this direction, the firm is trying to maintain its appeal among the largest possible consumer base . In some cases the firm may even change its brand in order to avoid losing consumers because of their changing preferences. As an example , Kentucky Fried Chicken repositioned its brand by going with the Finally, brand revitalization may be useful to develop new demand for produ cts, especially when the firm may need to rely on new uses of the product in order to maintain growth. The characteristics of a given product can often result in value for other consumers or consumption contexts (Adner & Snow, 2010) . For example, Johnson & Johnson suc cessfully targeted its baby shampoo to adults by emphasizing its gentleness as an attribute (Keller, 2008) . Alternatively, fi rms may decide to communicate new uses for its products in order to keep exploiting its brand. For example, Arm & Hammer brand faced significant decline when fewer consumers cooked at home. To combat this, they advertised baking soda as a deodorant for ref rigerators, revitalizing the brand to apply in new consumption contexts. In summary, in those industries where consumer preferences change over time brand revitalization strategy can be an effective way to maintain the appeal of the brand to its market. Un like brand extension, however, this form of brand management tends to rework existing products rather than introducing new ones. In other words, the latest model of Toyota Camry is simply an updated version of the earlier car with improvements that can all ow it to more effectively compete with the new offerings of its rivals and to offer more value to consumers. In addition, brand extensions focus on reaching new consumer segments whereas brand 23 revitalization focuses on reconnecting the brand with existing consumers. Therefore brand extensions enhance brand performance in markets with heterogeneous consumer preferences and revitalization enhances brand performance in markets where consumer preferences are changing. Both extending and revitalizing brand stra tegies focus on brands which maintain their market presence with continual transactions. For brand extension strategies the continued market presence of the parent brand is critical to keep the brand name salient for potential extensions. For revitalizatio n strategies, the continual deployment of new product models and marketing strategies maintains the brands continual market presence. Extensions and revitalization then effectively address consumer heterogeneity and changing preferences but they do not con sider the complexity of marketing a brand that is only temporarily present in the market. Re - creating brands for single purchase cycle goods. In contexts where products are only temporarily present in the market firms may need to re - create their successful brands in order to exploit their value. An important factor influencing how firms manage their brands is the purchase cycle for goods within the category (Pringle et al., 1982) . For goods with multiple Table 1 Contextual Factors for Effective Brand Management Strategies Brand Extension Brand Revitalization Brand Re - Creation Purchase Cycle Multiple Multiple Single Customer Mix Heterogeneous Heterogeneous or Homogeneous Homogenous Consumer Preferences Stable Changing Changing Target Market New Existing New & Existing Relevant Industries Consumer Packaged Goods Consumer Packaged goods & Consumer Durables Cultural Goods 24 purchase cycles firms are able to communicate their brand values through repeated purchases of the branded product. However other goods such as books, motion pictures and video games, are largely purchased within a single purchase cycle and thus sales of any new product tends to decline as consumers are not as likely to keep consuming the same product (Holbrook & Hirschman, 1982) . This effect is strongest in motion pictures, as most consumers tend will not view the same film more than once. But it also affects video games, where consumers will exp erience decreasing value from the same experience. For single purchase cycle products consumers tend to satiate when consuming a given product therefore firms need to introduce new products on a regular basis (McAlister, 1982; Redden, 2008) . For cultural products such as films, art and books consumer demand is largely satisfied with a single purchase. In addition, the sales of cultural goods is often driven by information casca des where the consumption response of friends strongly influences sales (Pollock, Rindova, & Maggitti, 2008) . But just as demand grows through information cascades it also quickly declines as the goods are consumed. As a result of this satiation effect, firms will be constrained in the extent to which each new product can resemble previous ones and each new product gets recognized as a separate brand. Firms therefore face challenges in draw ing attention to and generating interest in a new brand with the launch of each new product. presence firms employ a brand re - creation strategy. This strategy, which has becom e known as franchising, enables firms launch new products that are tied to a previously successful brand after sales have significantly decreased. These new products stay close to the concept of the previous product but will change some of its attributes. In this way, firms can continue to capture benefits from the brand of the preceding product by capturing some of its core appeal. 25 Brand re - creation differs substantially from both brand extensions and brand revitalization because of the temporary presence of the brand. Neither brand extensions nor brand revitalization consider the issue of intermittent periods of the brand being abse nt from the market. For example, in brand revitalization strategies firms continually sell a relatively high number of old models while new models are being launched therefore while sales may vary across the transition, the company can largely maintain its operations across various models. Furthermore a brand extension assumes the consistent presence of a parent brand in the market to launch new products. These periods however play a critical role in products with a single purchase cycle because they face t he challenges of weakening brand associations (Alba & Hutchinson, 1987; Basuroy & Chatterjee, 2008) effectively spacing new products and managing the extent to which re - creation of a brand is associated with previous products of the same brand (McAlister & Pessemi er, 1982; Sood & Drèze, 2006) . The issue of temporary market presence also complicates how firms manage brand continuity. Continuity plays a critical role in brand re - creation strategies because intermittent brand deployment provides fewer opportunities for reinforcing brand values. Alt hough brand extensions and revitalization consider brand continuity, they consider these issues in the context of repeated purchase cycles in which firms are able to repeatedly communicate and reinforce the brand ideals which can be critical to managing a brand (Campbell & Keller, 2003) . Brand continuity becomes a challenge in brand re - creation because the need for continuity is complicated by fading memori es and associations from a lack of exposure. Thus while the context requires a totally new sequel product the absence of the brand from the market increases the need for brand continuity in order to be identified with the franchise. 26 In summary, brand re - c reation is essential in those product markets where the consumers engage in a single purchase cycle, each sequel product has a relatively short life - cycle separated by time when the brand is largely absent from the market. Within this context, a re - creatio n strategy can enable a firm to draw on the core characteristics of a brand that have been established by the previous offerings to introduce new products. 27 CHAPTER 4 RE - CREATING BRANDS IN THE MOTION PICTURE INDUSTRY The U.S. motion picture industry is o ne of the oldest, dating back to the start of the 20 th century. It also generates the largest revenues and represents a significant contribution to the overall US economy. In 2013 , the industry recorded $130 billion in sales and 1.9 million jobs. Known as Hollywood, based on the early location of most of the film studios, it presently consists of six large firms and several smaller ones. All of the major studios, with the exception of Disney, can trace their origins back to the beginning of the U.S . film industry. Since 1980 , the six major revenues and profits. They primarily focus on fewer higher budget movies that they release widely, generally on 2,000 or more screens across the U.S. and in almost all markets outside the U.S. Over the last decade , they have released about 100 movies annually, with an average production budget of around $70 million and a marketing budget of around $35 million. In addition, over 500 movies are release d each year by smaller studios, often referred to as independents. Most of these cost less than $10 million to make and are released in no more than a handful of theatres. Three of the major studios also have subsidiaries that collectively release about 40 - 4 5 films per year that fall into this category. A couple of these independents, such as Lionsgate, could be considered large enough to be considered to be a major film studio. All films, especially those distributed by the major studios, are first released in to theatres. Increasingly, studios are releasing many of their films in many markets around the world on the same date. After their theatrical run, these movies are offered on DVD, to cable channels and for live streaming. DVD sales have declined, particul arly with the growing use of streaming video. In spite of the many different sources of revenue, theatrical box office receipts have remained a 28 The popularity of a movie during its run in theatres helps to create the mark et for its subsequent sources of revenue. Use of Sequels Film studios generate a large part of their revenues from a slate of films that they release each year. Because most consumers will not watch a movie more than once, each of the films has a relatively short life cycle, generating almost all of their revenues within a year of their release. Furthermore, since consumers will not be drawn to another movie that is exactly like one they have seen before, each film can be considered to represent a new brand which has little association with the prior offerings of the same film studio. As such, the studio takes a risk with launching each of its movies. In large part, this risk tends to be high because audiences and critics can be fickle i n their preferences. Therefore it is generally accepted that many new films fail to make money while in theatres (Vany & Walls, 1999; Vogel, 2010) . Therefore , film studios have relied increasingly on the use of brand re - cr eation. In motion pictures, re - creation consists of turning the brand associated with the movie into a franchise which can allow them to release sequels (M. A. Desai, Loeb, & Veblen, 2002; Eliashberg, Elberse, & Leenders, 2006) . By establishing a franchise , a studio reduces the risk of f uture offerings by tying it to a film brand that has already been successful (Landro, 1989) . With each s equel , therefore , the studio seeks to re - create the brand that was associated with the ti tle of a previous movie. T he use of sequels has grown considerably over time. In 2014, 12 of the top 25 grossing films were sequels. In addition, IMDB currently lists 226 anticipated sequel films between 2015 and 2020. The use of sequels as a strategy offers a number of advantages , s uch as reducing market risk. Because the franchise brand provides information about some key film characteristics , 29 consumers can more confidently select the film prior to consumption (Basuroy & Chatterjee, 2008; Keller, 2008) . For example, franchises such as Lethal Weapon and Mission Impossible integrate certain storylines and characters into the film offering (Basuroy & Chatterjee, 2008) . Associating the film with a franchise name serves as a useful signal to consumers making purchase decisio ns (Wernerfelt, 1988) . Essentially brand re - creation in the motion picture industry reduces the risk for consumers by providing useful information regarding whether t he film will satisfy their desires . Effective brand re - creation in the film industry requires that firms re - create a brand concept t hat sufficiently links the new product to a prior film in order to leverage favorable consumer memories. According to the branding literature , brands can be considered living entities that require both identity and vitality (Kapferer, 2008; Keller, 2008) . There fore , in the motion picture indu stry, firms effectively create sequels by leveraging components of the previous film which communicate the identity of the brand in the minds of consumers without (e.g. Aaker & Keller, 1990; Erdem & Swait, 2004; Park, Milberg, & Lawson, 1991; Sood & Drèze, 2006) . In other words, cons umers must believe that their experience with the sequel will be comparable to the previous film without being an exact replication. Timing of Sequels . The timing of sequel products represents a critical strategic decision for firms seeking to effectively manage the brand value associated with a franchise (Basuroy & Chatterjee, 2008) . In the movie ind ustry, franchises often vary widely in the timing of sequels with some franchises offering sequels in quick succession while others take longer between sequels. For example, Back to the Future P art III (May 1990) followed only 6 months after Back to the Fu ture Part II (November 1989). In contrast, Herbie Fully Loaded (2005) came 25 years 30 after the previous Herbie Goes Bananas (1980). The strategic decision of when to deploy sequel products can have important implications for the effectiveness of brand re - cr eation strategies. Time elapsed between the launch of sequel products can influence the degree to which subsequent offerings are able to draw upon positive brand associations related to the franchise. By strategically timing sequel products , firms capital ize on positive brand associations (Basuroy & Chatterjee, 2008; Wyer & Srull, 1986) . The release of a sequel product within a franchise refreshes the brand within the market, establishing the brand ideals and t he potential value of franchise products. By q uickly releasing sequel products , firms are more effectively able to leverage the franchise brand to promote each of their products. This is largely because the short time between sequels ensures that brand ideals are salient and accessible to potential consumers when making their purchasing decisions (Alba & Hutchins on, 1987; Basuroy & Chatterjee, 2008; Lynch & Srull, 1982) . As such, by releasing sequels in quick succession firms are able to draw on the benefits of a salient brand that can enable them to differentiate their films from others. The more closely a fi rm can link its sequel products to a familiar franchise, the more effectively it can allow consumers to relate to and fully appreciate the various attributes of the film (R. L. Priem, 2007; Ratchf ord, 2001; Stigler & Becker, 1977; Wernerfelt, 1985) . When consumers can employ the familiarity ac quired from previous consumption, the economies of brand familiarity make franchise branded products more valuable. In contrast to rapid sequel product dep loyment which leverages and influences brand ideals delaying the launch of sequel products can weaken relevant brand associations. The impact of a firms brand on the market is partially contingent on the relevance, salience and strength of brand associatio 31 are not stable over time but are prone to decay without reinforcing stimuli (Alba & Hutchinson, 1987) . Therefore, absent other factors to reinforce the brand such as marketing or fan clubs, when sequel products are separated by longer time periods the impact of brands on consumer preferences and product choice decreases. Thus franchise brands have a more positive effect on sequel product performance when time between products is short rather than long. Hypothesis 1: Time elapsed between sequels is negatively related to se quel performance . Although time lapsed is expected to be negatively related to sequel performance this negative relationship may attenuate and possibly reverse beyond some point. As argued above the impact of a franchise brand on sequel performance is cont ingent on exploiting vivid consumer memories. Although memories related to a given product are likely to fade over time, firms may be able to revive those memories through a lens of nostalgia (Schindler & Holbrook, 2003; Stern, 1992). Through evoking nosta lgia by delaying the release of a sequel , firms enhance earlier point in their lives. Delaying sequel products is likely to draw upon previous experiences and thus enhance nostalgia (Brown et al., 2003) likely to be experienced by consumers wh o have earlier associations related to the brand or product. These associations are more likely to enhance brand attitudes and purchase intentions when these associations are positive and founded on attachments to the product formed either during childhood or while they are growing up (Mueling, Sprott & Sultan 2014). By delaying sequel products firms have an opportunity to capitalize on these earlier memories and product associations formed at a time in life when emotional attachments to a brand or brand id entification may be stronger. 32 When firm offerings evoke nostalgia they are likely to enhance the positive associations with a given product or brand. Evoking nostalgia can allow firms to enhance consumer interest, attention and involvement in the consumption process (Bambauer - Sachse and Gierl 2009b; Marchegiani and Phau 2005; Muehling and Sprott 2004; Muehling and Pascal 2011; Pascal, Sprott, and Muehling 2002). By enhancing attention and involvement, nostalgic brands and products are more likely t o revive positive feelings and attitudes towards the franchise brand and product ( Burke and Edell 1989; Derbaix 1995; Holbrook and Batra 1987; Moore and Hutchinson 1985). Thus by linking a sequel to nostalgic remembrances of the franchise, firms can enhanc e consumer interest and attention to influence consumer judgment (Mueling, Sprott & Sultan 2014). Increased time between sequels enables firms to reframe memories of product performance by evoking positive nostalgic memories about the franchise while avoid ing negative memories that a more recent release might trigger (Braun, Ellis, and Loftus, 2002; Moore & Homer, 2004; 2008). Increasing the time elapsed between sequels allows consumers to experience nostalgia by allowing consumers to draw upon affect laden past experiences. Recalling those experiences should influence consumer judgment through increased consumer involvement (Sujan, Bettman & Baumgartner, 1993). Consequently, though time elapsed initially undermines sequel performance, extending that time ma y eventually lead to an improved sequel performance. Hypothesis 2: There is a curvilinear shaped relationship between time elapsed between sequels and sequel performance such that the relationship begins negative and turns positive . Number of Sequels . Sequ el products serve as reinforcing mechanisms to enhance brand value. Through the release of additional sequel s firms reinforce the franchise brand by 33 partially establ ished through repeatedly creating positive associations between the consumers and the ideals of the brand. Additional sequel products enable firms to re - create similar products (Campbell & Keller, 2003) . In addition to enhancing consumer associations, each succeeding sequel may enhance the perception of brand quality (Basuroy & Chatterjee, 2008). Given that product markets can be highly complex, firms are challenged to differentiate their offerings from competitors. Brand quality becomes a signal that can differentiate a product from those of competitors by reducing support a perception of brand quality if consumers view subse quent offerings as an indication of the success, and presumably the quality, of prior offerings. Thus, a sequel product associated with a mature franchise can carry with it a signal of quality when it is preceded by other sequels. For example, the video g ame franchise Mario has been associated with over 200 sequels across multiple video game platforms. Because of its rich history as a franchise, sequel products associated with Mario often signal a high quality gaming experience. Therefore , by associating s equel products with a mature and well known franchise , firms differentiate their products by signaling quality. In addition to serving as a signal of quality, sequel products may also enhance consumer recall of the franchise (Aaker 1991; Naik, Mantrala, & Sawyer, 1998). The value of a franchise brand is partially a result of the ability of consumers to recall the brand and its ideals. Consumer recall of a given brand generally arises from repeated exposure to the brand across time (Campbell & Keller, 2003) . Therefore, by releasing additional sequel products, firms may 34 enhance consumer recall of the franchise brand and thus enhance brand value (Basuroy & Cha tterjee, 2008; Sood & Drèze, 2006). In sum, sequel performance can improve with the prior number of sequels offered. Hypothesis 3: Sequel number is positively related to sequel performance . Effectively managing a brand franchise can vary significantly ove r the age and maturity of the brand (i.e. Simon 1979). The inimitability and value of a brand is arises from how effectively a sequel product draws upon favorable consumer associations to enhance their willingness to purchase. Early within a franchise cons umers have limited exposure to the brand or its related products and thus the brand is either not well recognized or its recognition may not evoke the positive associations to signal quality (Innis & Unnava, 1991) . This lack of exposure to sequel leads to weakly held brand associations. In contrast, mature franchi ses benefit from stronger The strength of brand associations plays an important role in how the timing of sequel products may influence their performance. In particu lar, because sequel performance relies on evoking past memories the importance of timing varies according to the maturity of the franchise (Balota, Duchek & Logan 2007). Effective sequel timing capitalizes positive consumer associations which vary with the maturity of the franchise (i.e. Luan & Sudhir, 2010) . For instance, early in the life of the franchise consumer associations with the brand are likely to be weakly held and thus it may be more important for sequel products to follow in quick succession in order to capitalize on these associations. In contrast, as the franchise matures, with more sequel products, these associations are likely to be more strongly held and thus more resilient to longer time periods between sequel products. 35 Weakly held associations with a given franchise are more likely to deteriorate with time elapsed between sequel products. Weakly held associations for a given brand are more likely to be forgotten by consumers and become l ess accessible for consumption of future sequel products (Alba & Hutchinson, 1987) . Memories of brand associations are largely reinforced through repetition and for newer franchises these associations have not yet been reinforced. The lack of reinforced brand associations suggests that less mature franchises will result in lower perform ance when time elapsed between sequels is higher. Therefore for newer franchises, firms enhance performance by quickly deploying sequel products to capitalize on recent awareness and market buzz (Friedman, 1992; Lehmann & Weinberg, 2000). In contrast, matu re franchises are more likely to establish more strongly held associations with the brand. More strongly held associations result from multiple sequel products which reinforce brand ideal. These strongly held associations are more likely to be resilient to deterioration over time and thus time lags will have less impact (Joshi & Mao, 2010). A number of film franchises appear to have followed this type of strategy by quickly releasing sequels early in the franchise and then as the franchise matures allowing for more time between sequel products. For example, Teenage Mutant Ninja Turtles which released three films between 1990 and 1993 and then just recently launched another sequel in 2014. Hypothesis 4: The negative relationship between time elapsed and seque l performance is moderated by sequel number such that the performance will be much weaker for earlier sequels than for later sequels . Leveraging Product Level Similarity To maintain or enhance the value of a franchise, firms often maintain consistent use of symbols, names and designs to serve as cues regarding the potential performance 36 (Aaker, 1991; Erdem & Swait, 2004; Janiszewski & van Osselaer, 2000; Keller, 20 08). Product level similarity refers to consistency in certain product components across sequel products. The configuration of product components can play a critical role in the perceived value of the product to consumers (Cooper & Kleinschmidt, 1987) . Such a consistent use enhances the value of the franchise by exploiting the f amiliarity of consumers with these components acr oss all of the films in the franchise. For example, the Mission Impossible franchise has largely drawn on the use of Tom Cruise as a lead actor across all films. Audiences have begun to associate him as the principal character that pursues the different c hallenges that face the team in each of the sequels. Similarly, the Bourne franchise has relied on the consistent use of Matt Damon as the main actor for the protagonist Jason Bourne leveraging the familiarity with and understanding of the character in ord er to drive sequel performance. However when the film studio employed Jeremy Renner to play the role of the new protagonist performance at the box office as well as among critics significantly declined. As such , lead actor and actresses are an important pr oduct component which has been associated with higher film performance (Albert, 1998; Elberse, 2007; Wallace, Seigerman, & Holbrook, 1993). Lead acting talent often represents the face of the franchise by connecting the rs. This connection between the audience and the lead roles can endear the audience to the franchise brand and evoke positive associations. Because of the familiarity consumers may have developed for a specific actor playing a lead role this continued asso ciation is likely to more efficiently and effectively evoke positive associations with the sequel product. Thus by employing previous acting talent film studios enable consumers to rely 37 on previous knowledge making consumption less costly in terms of time and effort to learn, thus their consumption experience . Hypothesis 5: The use of previous lead actors used in the prior movie is positively related to performance . While product level similarity can enhance sequel performance this positive relationship is likely t o weaken as time between sequels increases. Consistent product components enable consumers to more efficiently engage in consumption through the economies of familiarity. However these p ositive effects may weaken as consumer memories and familiarity decline in the absence of reinforcing stimuli (Cooper & Kleinschmidt, 1987; Shepard, 1967) . Therefore the ability of product level similarity to draw upon these memories is also limited by longer time periods between sequel products. In contrast to quick succession, the connections between consumers and particular product components are likely to weaken with greater time between sequel product releases. Consumer memories connecting product level components to the franchise are likely to deteriorate and evolve over time (Alba & Hutchinson, 1987; Luo, Chen, Han, & Whan Park, 2010). With more time between sequel products the franchise brand is less prominent within the market and the connections between the franchise brand and product components are likely to deteriorate. Weakened connections between the consumer and product components also arise from shifting consumer preferences which more likely to occur with mor e time between sequel products. Consumer preferences are often changing over time based on their evolving tastes and lifestyle (Adner & Zemsky, 2006) . In particular consumer preferences for cultural goods such as films can be notoriously fickle (Pel toniemi, 2015) . When sequel products follow in quick 38 succession the likelihood of a shift in these preferences is lower than when sequel products are separated by longer time periods. Thus as memories, preferences and tastes evolve, consumers may be les s likely to find value in consistent components of sequel products. The role of time between sequel products can therefore play a particularly important role as it relates to acting talent in sequel films. In particular, the value of lead actors playing a role in a series of sequels is likely to weaken as the time between sequel films increases. Through longer time periods between sequel products, the association of lead talent with the franchise is likely to decay (Tulving & Psotka, 1971). In addition, as the time between sequels increases, both acting talent ages and the characters they represent can evolve thus potentially weakening the impact of acting talent on sequel performance. Consumers of the franchise may disassociate the actor from the character conception of the character by consumers and the portrayal by the actor leading to a devaluing of the actor within the franchise. Thus the impact of talent is likely to weaken as as sociations decay over time. Hypothesis 6(a): The positive effect of previous acting talent decreases with time lapsed . Early in the franchise, firms benefit from utilizing consistent product level similarity to increase the association of consumers with a given brand (Basuroy & Chatterjee, 2008). For emerging franchises consumer conceptions of the brand are likely to be more uncertain and weakly held (Bivainiene, 2010; Simon, 1979) . Because the brand may have a weaker salience for less established franchises it can be critical for firms to maintain product level similarity in order to leverage the e conomies of familiarity. Because the brand is not strongly recognized by consumers they rely more heavily on product component consistency to assess the value of sequel products. 39 As the franchise matures however , consumer conceptions of the franchise brand become more strongly held such that consumers become more associated with the brand and less dependent on product level similarity . Over the life of the franchise, sequel products can serve as reinforcing brand mechanisms reminding consumers of the b then are more likely to have a more established brand which is more readily recognized by consumers. Because the brand is more established, the performance of the sequel is largely due to its association with the franchise brand and may be less impacted by product level consistency (Desai & Basuroy, 2005). The role of stronger branding and satiation will similarly impact the performance of sequels in the film industry. For mature film franchises, performance is likely to be less affected by consistent lead actors due to stronger franchise brands and the risk of consumer satiation toward certain actors (Sood & Drèze, 2006) . Sequel films associated with mature franchises are likely to enjoy greater brand recognition and market buzz thus enhancing their performance despite cha nging the lead actors (Basuroy & Chatterjee, 2008) . Thus, later in the franchise history consumers may experience fewer benefits from consistent acting talent when compared with earlier in the franchis e (Brickman & Campbell, 1971; Herrnstein & Prelec, 1991; McAlister, 1982). Hypothesis 6(b) : the effect of previous actors on performance decreases with the number of sequels . Leveraging Brand Level Consistency While product level factors can play a critical role in the effectiveness of brand re - creation strategies product level similarity does not fully consider the importance of consistency in the brand concept (Park et al., 1991) . The brand concept refers to the ideals and values 40 associated with a given brand and its associated products. Consistency in the brand concept suggests that additional sequel products associated with the brand will evoke similar emotional responses from consumers. Brand concept consistency is critical to maintaining brand value for sequel products because firms capitalize on brand recognition when they ensure that consumers can easily associate the product with the brand (Park et al., 1991). Maintaining consistency of the brand concept requires a tacit understanding of the brand and the delivery of the promise embedded in the brand (Kapferer, 2008). Valuable brands within the motion picture industry are often the result of the integration of diverse business activities such as production, distribution and marketing. When these business activities reinforce a particular brand message they support the brand concept in the minds of consumers. In order to maintain the consistency in brand concept across sequel products, firms must learn to develop and utilize tacit knowledge for coordinating these diverse activities (e.g., Polanyi, 1962). For example, the Fast and Furious franchise brand promises illegal street racing, precision driving sequences, and heists interwoven with the nuances of familial ties. Neal Moritz, ing sequels while adding new features that draw consumers to the newer version. By managing a consistent brand concept, Fast and Furious fosters consumer commitment and perceived values because consumers have a consistent experience of product fulfilling t he brand promise, while also offering something to pique their interest (such as the outrageous stunts in Fast and Furious 6). Thus, by ensuring accessible and salient brand ideals, firms can use each of the sequels in a franchise to reinforce brand meanin gs and ideals. In the film industry , the production company is responsible for integrating the various activities that all eventually come together to influence the look, the feel and therefore the brand 41 of the film . They hire actors, director, special e ffects support, arrange for locations and so forth. How all these elements come together in the making of a film influences the look, feel and therefore the brand of the film. This look and feel as experienced by consumers plays a critical role in ensuring that the sequel is consistent with the franchise. By producing the previous film the production company develops the necessary tacit knowledge to effectively deliver on the brand concept for each sequel film. By enlisting the previous production company t he film studio can ensure that sequel films conform to the franchise brand concept and thus improve the performance of the sequel. Hypothesis 7: The use of production companies used in the prior movie is positively related to performance . When quickly dep loying sequel products, concept level consistency remains a highly salient indicator for using the franchise brand to enhance performance (Basuroy & Chatterjee, 2008; Naik et al., 1998). When sequel products are released in quick succession, consumers are and salience of these factors, brand concept consistency is likely to play a more important role in sequel performance when time elapsed is low because consumers are more likely to recognize and thus appreciate consistency of these factors. Increasing the time between sequel products, however, can reduce the need for consistency, making consumers more open to novelty in brand concept (Basuroy & Chatterjee, 2008; Desai & Keller, 2002; Sgourev 2013). With increased time between sequel products consumers are less likely to remember , and be influenced by , brand concept consistency factors when considering the value of the sequel product . Because these brand factors ar e often subtle, their importance to consumers evaluating the sequel product may decline with time. As time 42 between sequels increases , these factors may become elusive and thus less critical to the brand concept. Ultimately time between sequel products make s brand concept consistency less critical to the success of the product (Alba & Hutchinson, 1987; Luo, Chen, Han, & Whan Park, 2010). Part of this reduced need for consistency as time increases is that consumer memories related to the brand concept are no t stable but rather can evolve over time. The evolution of consumer memories is more likely to occur with more time between sequel products due to the re - examination of memories in light of new experiences, new information and the sharing of brand interpre tations with other consumers whose memories likely differ. By delaying sequel products, firms allow consumer associations with the brand concept to weaken (Tulving & Psotka, 1971). This may lead many consumers to be open to changes in the brand concept. A ll of this suggests that the brand concept become less important due to fading or changing memories or more openness to change. Under these circumstances a change in the production company may not be as detrimental to the performance of a sequel. In other words, the effectiveness of the continued use of a production company in order to maintain brand concept consistency may decline with an increase in the time that elapses between sequels. Hypothesis 8(a) : The effect of previous production companies on performance decreases with the time elapsed between consecutive films . Maintaining a consistent brand concept during the early stages of a franchise can be critical to the success of the franchise since the b rand concept is relatively nascent. Consistency reinforces positive associations between the consumer and the franchise brand (Luan & Sudhir, 2010) . In addition, maintaining brand concept consistency reduces the consumption ris k of consumers because linking sequel products with a consistent brand concept can ensure similar quality for sequel products (Basuroy & Chatterjee, 2008). By ensuring that the first few sequels 43 carry a similar brand concept consumers develop a clear under standing of the brand and the value proposition offered by that brand. While brand concept consistency can play a critical role in enhancing performance early in the franchise, later in the franchise brand concept consistency will play a lesser role in pe rformance because consumers are likely to be more open to changes in the brand. As consumers become more familiar with the brand they can also become more open to changes in the brand concept (K. K. Desai & Basuroy, 2005; Ho - Dac, Carson, & Moore, 2013) . With higher numbers of sequel products consumers become more familiar with the core concepts of the brand. In particular , as consumers become more familiar with the franchise brand this knowledge enables consumers to more efficiently process new information such as changes in the brand concept (Johnson & Russo, 1984) . Greater familiarity with a brand can serve as a knowledge set which allows consumers to e ffectively encode and evaluate new information. This familiarity with the brand concept can lead consumers to be less negatively affected by changes in the brand co ncept as the franchise matures. In addition to consumer familiarity, consumer satiation may also lead to consumers being more open to variation in brand concept. Within some product markets, consumers value variation in sequel products and brand concept, therefore for more mature franchises consumers may not find as much value in brand concept co nsistency (Sood & Drèze, 2006) . Additional sequel products may in fact increase the risk of consumer satiation with core elements of the brand. That is, consumers may tire of certain aspects which begin to feel dated or too repetitive (Adner, 2002; Coombs & Avrunin, 1977) . At this point, consumers may require the introduction of new concepts that update and revive the brand in ways that reawaken interest in the brand. Thus, satiation from additional sequel products in can weaken consumer preference for a 44 particular brand c oncept (Brickman & Campbell, 1971; Herrnstein & Prelec, 1991; McAlister, 1982). For franchises in the film industry the previous production company plays an important role in brand concept continuity. By maintaining the previous production company, film st udios utilize existing knowledge of the franchise brand concept for sequel films. However by changing the production company of a franchise brand can lead to changes in the brand concept as a result of new tacit knowledge . Therefore consumers may find less satisfaction with the original brand concept as the franchise matures and more open to changes with additional sequel products. Hypothesis 8(b) : The effect of previous production companies on performance decreases with the number of sequels . 45 CHAPTER 5 METHODS To study how brands are managed within the film industry the sample focuses on movie franchises within the U.S. film industry, based on the use of sequels. Sample The list of movie franchises is primarily drawn from an extensive list maintained by www.the - numbers.com . This list was further supplemented with additional franchises from the IMDB through the combined review of two researchers. The sample includes films that clearly represent sequels such as those tied to Harry Potter. However, they also include those sequels which were not as closely tied to the previous film. For example, A Fish Called Wanda had a follow up film entitled Fierce Creatures. Therefore, the sample includes sequels which are both closely associated to the original film and those which are more distant. In all, the sample for this study consists of 311 movie franchises within the U. S. motion picture industry associated with 910 total films including the first movie in each franchise. The sample then represents nearly a fully known list of franchises which have been distributed in theatres within the U.S. market. The franchises in th is sample cover 1960 to 2013 which represents an appropriate time period for this study because of the increasing use of sequels as a brand strategy of film studios over these years . For example, the number of sequel movies increased from seven sequels dur ing the 1960s to 199 sequels from 2000 - 2009. Not only did the number of sequels greatly increase during this time period, but also the performance of those sequels also varied greatly. For example, Universal Soldier: Day of Reckoning grossed $5,460 at the box office whereas The Dark Knight grossed $533 mil lion. Within the sample 33.8% of the sequels represent first 46 sequels, 13.2% represent second sequels and 6.8% represent third sequels. Among the sequel films the James Bond franchise remains a clear outlie r. Including the franchise, sequel number has a mean of 2.51 and with 23 sequel films, is over 7 standard deviations beyond the mean level. Thus I have remove d this franchise in order to reduce the risk of the results being driven by extreme observations. Based on data coverage and the removal of extreme values yields a final sample of 499 sequel films. While the sample was primarily drawn from www.the - numbers.com and the IMDB other data about the films themselves was drawn from www.boxofficemojo.com , www.rottentomatoes.com and www.quigleypublishing.com . Dependent variable Domestic box office receipts. The focus of this study is primarily on how brands can be successfully managed. The performance of a film brand is largely measured by the films box office receipts. Box office receipts represent the total revenues collected f or a given film through theatre distribution channels. Past research has suggested that domestic box office re ceipts serves as an effective proxy for the value of the movie in other markets (Ainslie, Drèze, & Zufryden, 2005) . Further, box office receipts have been suggested to be a reliable indicator of overall revenue in the film industry (Ravid, 1999) . Finally, the use of market based measures of performance have been suggested as more useful for RBV research than measure s of appropriation (Crook et al., 2008) . T hus, box office receipts serves an appropriate measure of Domestic box office receipts were primarily gathered from www.the - numbers.com . For any films not reported on this website, data was collected from IMDB and www.boxofficemojo.com which both report the revenue associated with the entire domestic theatrical release . This variable is highly skewed and there fore is log transformed. 47 Independent variables Time e lapsed. Time elapsed represents the number of years that passed between the release of a sequel and the release of the film that preceded i t. When s equels follow each other more quickly , time e lapsed is low. Sequel number. Sequel number represents the temporal sequence in which the film appears within the franchise. For example, the first sequel is coded as 1, with each temporally subsequent sequel an increasing value. High levels of sequel num ber thus represent films later in the franchise whereas low levels of sequel number represent films early in the franchise. Previous actors. Previous actors represent maintaining an association between a brand and its central product components. Acting tal ent can play a critical role in how the characters are understood by the audience. By continuing to employ the previous acting talent , film studios maintain the image of key characters within the franchise however when acting talent changes between sequel films the image of the characters can be disrupted by differences in how the role is portrayed and understood by the audience . IMDB maintains a database of major actors in a given film with up to 5 major actors or actresses within a given film and was revi ewed by two researchers to ensure validity. As a measure, previous actor s represent t he percentage of central actors, meaning the key protagonist and antagonists in the film that are included from the previous film. Previous production company. Previous production company serves as proxy for brand concept consistency. The brand concept consistency in films is often a result of the look and feel of the film as experienced by an audience. These components are often arise from both the inputs, such as the ty pe of story and characters, the use of different types of talent and the visual and special effects. But these have to be carefully integrated to communicate a coherent brand 48 image. The production company represents the product development resources for cr eating the film through securing the talent, managing the shoot and handling post - production . Therefore production companies are largely responsible for the look and feel of the film. By maintaining a consistent production company film studios leverage the tacit knowledge held regarding the concept in subsequent films. The production companies were collected from IMDB and supplemented from Wikipedia. Previous prod uction company represents the percentage of production companies from the previous film that are involved in producing the sequel. Control variables Previous film performance. The performance of a sequel is likely to be partially driven by the performance of the previous film in the franchise. When the previous film has a higher level of performance the franchise is likely to be more well - known and recognized by the market. Therefore, higher performing franchise films are likely to benefit from market momen tum to enhance the performance of sequel films. Previous film performance represents the domestic box office receipts of the previous film in the franchise. The variable is highly skewed and therefore is logged. Production b udget. The production budget for a given sequel film can vary significantly based on the film studios commitment to the film. Higher production budgets enable the film studio to secure better talent both for acting and production crews. With higher budgets films often have stronger appeal based better known talent and production values . Data for the production budget is collected from the IMDB and supplemented with data from Wikipedia. Because this value is highly skewed the measure was logged to reduce the impact of extreme values. 49 Ticket price. Importantly, these sequels occurred over a large period of time when tick et prices and increased significantly. One approach would be to scale our dependent variable by this measure, however to avoid the risks associated with ratio dependent variables and to control for this potentially confounding variable, ticket price represents the average cost of a movie ticket during the year the sequel was released (Wiseman, 2009) . Book based franchise. Book based franchises differ from other franchises because they are based on characters from another creative characters and general plot lines for a book based franchise are likely to be different from a n on - book franchise. Further, because they are often part of a series, the number of sequels is likely to be significantly different for book based franchises than from those originating as a film. This variable is coded 1 for sequels based on books and 0 fo r sequels not based on books. Character based franchise. Character based franchises are franchises based on a certain protagonist within the story. Character based franchises such as Batman are likely to perform differently because of other brand based ele ments of the franchise and the character. Films were coded as character based when the title of the film includes a main character of the film. This variable is coded 1 for sequels based on character franchises and 0 for sequels not based on character fran chise. MPAA r ating change. The Motion Picture Association of America evaluates films according to the appropriateness of the content for certain audiences. A change in MPAA rating can significantly influence the audience which the film targets. Films which change their rating may or may not maintain their current customer base. This variable is coded with a dummy code with 1 representing a rating change and 0 representing a consistent rating. 50 Star power. Previous research has suggested that the performance a film can be driven to some degree by the appeal of the lead acting talent (K. K. Desai & Basuroy, 2005; Ravid, 1999) . Acting talent can vary wide ly in their popularity and audience recognition. By employing more recognized and popular stars , film studios may be able to enhance the performance of their films. To measure star power , I checked each lead actor or actress involved in the film to identify whether they were among the top 10 earning stars, as reported by Quigley Publishing in the year the film was released. Thus star power is a count variable which in creases with the number of top 10 actors or actresses filling major roles in the film. Unrelated sequel title. The title of a sequel film can vary based on how directly the film draws upon previous sequel products. For example, some franchises draw directly on previous sequel pr oducts as in the case of The Fast and the Furious franchise. Other films can generate entirely new titles that have little association with the previous film as in the case of A Fish Called Wanda had a follow up film entitled Fierce Creatures. To control f or this effect I include a dummy code which is 1 when the film has less than 2 of the same words as the previous film in the title. Critic reviews. The performance of a given film can be partially explained by the reception of the film by critics. Critics can often serve as expert consumers which evaluate the film based on its novelty and value to the market. Critic reviews are often related to domestic box office r eceipts because these reviews can serve as a meaningful signal regarding the value of the film (Lampel & Shamsie, 2000) . Critic reviews were collected from the lead critics average review as posted by www.rottentomatoes.com. 51 Table 2 Data Measures Construct Measure Level Calculation Source Dependent variable Domestic b ox office receipts Domestic box office receipts during the domestic theatrical release Sequel www.the - numbers.com www.imdb.com www.boxofficemojo.com Independent variables Time elapsed Number of years since the prior film released Sequel www.imdb.com Sequel number Order in which the film is positioned within the franchise based on the release date Sequel www.imdb.com Previous actors The percentage of central actors from the previous film Sequel www.imdb.com Previous production company The percentage of production companies from the previous film Sequel www.imdb.com www.the - numbers.com www.wikipedia.org Control variables Previous film performance Box office receipts of previous film Sequel www.the - numbers.com www.imdb.com www.boxofficemojo.com Production Budget Published production budget Sequel www.imdb.com www.wikipedia.com Ticket price Average ticket price spent for movies during the year the film is released Year www.boxofficemojo.com 52 Table 3 Data Measures (continued) Book based Dummy code for franchises which are based on a previously published book Sequel www.the - numbers.com www.wikipedia.org Character based Dummy code for films based on characters Sequel Code based on title reference to a character MPAA r ating change Dummy code 1 for a change in rating 0 for consistent rating Sequel www.imdb.com Previous film performance Box office receipts of previous film Sequel www.the - numbers.com www.imdb.com www.boxofficemojo.com Star power Number of top 5 grossing stars in a given film Sequel www.quigleypublishing.com Unrelated sequel title Dummy code for whether the focal sequel title has words from the previous film Sequel Coded based on title comparisons Critic reviews Rating from 0 - 1 based on the reviews of major critics Sequel www.rottentomatoes.com 53 CHAPTER 6 RESULTS The descriptive statistics and inter - correlations of the variables in this study are found in Tab le 3 . The descriptive statistics are largely in line with expectation. Domestic box office receipts has a very large variance with the maximum value of $533 million for The Dark Kn ight to a minimum value of $5,460 for Universal Soldier: Day of Reckoning which was distributed in only 3 theatres. Production b udget also has a large standard deviation. The raw values are reported in the descriptive statistics however the log was taken of both d omestic box office receipts and p roduction b udget in the analysis. The highest budget film was at $ 300 million with the lowest budget sequel film at $ 600,000 million for The Return of Texas Chainsaw Meassacre . There were several (6) sequel films released during the same calendar ye ar as their predecessor and 3 films were among the longest time periods between sequel films including Herbie Fully Loaded, Planet of the Apes and Jungle Book 2 at 25, 28 and 36 years respectively. The longest running franchises include Star Trek and Frida y the 13 th at 11 and The Pink Panther at 9 sequel films. In general the descriptive statistics are in line with what would be expected from the dataset. There are some strong correlations among the variables in the dataset. For example, the correlation bet ween previous film p erformance and domestic box office r eceipts is .79. This correlation is high, but to be expected, as films from the same franchise are likely to perform similarly. Furthermore both production budget and critic reviews are strongly correlated with domestic box office r eceipts at .77 and .43 respectively. These correlations are to be expected as high budget films are likely to generate greater market buzz and offer special effects and 54 production resources likely to gener ate high attendance. Also, while critic r eviews are related to performance the correlation is somewhat weaker as expert reviews are likely to evaluate different criteria when compared to average consumers. The other correlations are largely as expected. Fi nally, the focal independent variables indicate some level of variance with standard deviations close to their respective means, suggesting that over - dispersion does not appear to be an issue for the focal independent variables. In addition, the focal inde pendent variables are somewhat related to domestic box o ffice r eceipts. Specifically, time el apsed has a small negative relationship (r= - .04) sequel n umber a s mall positive correlation (r=.05 ) with d omestic box office receipts while both previous actors an d previous production c ompany have a moderate positive relationships with d omestic box office receipts (r=.36 & r=.14 ). The unit of analysis within this study is the film sequel. While the use of d omestic box office receipts as a dependent variable can be considered a continuous dependent variable and thus analyzed using ordinary least squares regression (OLS) , this approach poses a number of restrictions when compared to generalized linear modeling using maximum likelihood estimation (MLE) . In particular , because of the difference in assumptions, MLE uses more information in the data to estimate parameters and standa rd errors when compared to OLS. The differences in the amount of information used by either OLS or MLE is based in the assumptions of either te chnique. OLS fit s parameter estimates by minimizing the difference between the estimate and the distance to surrounding observations otherwise known as the mean square error. OLS does this by imposing normality assumption on the population from which the d ependent and independent variables are drawn. Therefore, OLS uses individual observations in order to minimize the error across the sample range. In contrast, MLE requires a distributional assumption of the error s while not imposing assumptions on the population variables . In doing 55 so, MLE maximizes the probability of obtaining the sample data observed. Based on these assumption MLE estimates the probability of obtaining a sample size of N, for a given population b ased on the probability of N observations. Thus MLE more fully considers the information in the sample when estimating coefficients and standard errors. MLE also assumes independence of observations when estimating standard errors . Because sequel film obse rvations are nested within franchises, which are likely to perform similarly, I correct for this by using clustered standard errors by franchise (Wooldridge, 2003) . I use an identify link function because the dependent variable is logged and a Gaussian fa mily for the standard errors because the error are assumed to be normal. The hypotheses in this analysis include a number of higher order coefficients such as interactions and quadratic terms which can increase the risk of multicollinearity. In order to re duce the risk of multicollinearity I mean center all of the variables in the analysis by subtracting the mean value from each variable. In doing so I remove non - essential multicollinearity making the results more stable. The results of the regression analy sis are reported in Table 2 with d omestic box office receipts as the dependent variable. Model 1 includes the control variables. Previous film performance should be positively related to domestic box office receipts as sequels of previously very successful film are also likely to be more successful than sequels of less successful films. The coefficient s for both previous film performance and production budget are positive and statistically significant (p<.001). C riti c r eviews is also positively related to domestic box office receipts and is statistically significant (p<.001). Ticket p rice represents the average ticket price for the year of the focal sequel to measure the difference in performance due to increasing tic ket prices. This variable is negative but not st atistically significant (p= .612 ). Book b ased is a 56 dummy code representing whether a given film was based on a previously published book and is negative but not s tatistically significant (p=.212 ). Character b ased represents whether the focal film is based on a central character such as Superman or Harry Potter these films are expected to perform at a higher level because of the general recognition of the focal character related to the franchise. The coefficien t for character b ased is positive but not s tatistically significant (p=.936 ). One might expect MPAA rating c hange to be negatively related to d omestic box office receipts because changing the target audience may reduce their familiarity with the franchise bra n d. The coefficient for MPAA rating c hange is negative and s tatistically significant (p=.019 ). Star p ower should be positively associated with domestic box office receipts as films with more well - recognized stars are likely to benefit from that positive association. The coefficient for star p ower is positive and s tatistically significant (p=.030 ). Finally, unrelated sequel t itle is a dummy code representing whether the title of the sequel could not be readily associated with the previous film. There were very few instances in the database and thus the coefficient while positive is not s tatistically significant (p=.877 ). The AIC and BIC model fit statistics for Model 1 are 1383.073 and 1425.199 respectively. 57 T able 4 Correlations and Descriptive Statistics Correlations Mean S.D. Min Max 1 2 3 4 5 6 7 8 9 10 11 12 13 1 Domestic box o ffice receipts (000) 82,4 00 93,4 00 5 . 46 533, 000 1 2 Previous f ilm performance (000) 97,5 00 93,800 7 .6 533, 000 0.79 1 3 Production b udget (000) 49,300 55,400 6 00 300, 000 0.77 0.71 1 4 Ticket p rice 5.47 1.76 1.53 8.02 0.39 0.34 0.5 0 1 5 Book b ased 0.2 0 0.4 0 0 1 0.09 0.12 0.11 - 0.05 1 6 Character b ased 0.21 0.41 0 1 0.15 0.1 0 0.18 - 0.05 - 0.03 1 7 MPAA rating c hange 0.19 0.39 0 1 - 0.14 - 0.13 - 0.06 - 0.16 0.06 0.07 1 8 Star p ower 0.36 0.64 0 4 0.25 0.23 0.26 0.02 0.06 0.12 0 .00 1 9 Unrelated s equ el t itle 0.01 0.1 0 0 1 - 0.06 - 0.05 - 0.05 - 0.08 0.1 0 - 0.05 0.15 0.07 1 10 Critic r eview s 0.41 0.28 0 1 0.43 0.26 0.3 0 0.1 0 0.11 0.21 - 0.03 0.09 0.06 1 11 Time elapsed 4.54 4.5 0 0 36 - 0.04 - 0.14 0.04 0.1 0 0.04 0.01 0.16 0.08 0.24 0.1 0 1 12 Sequel n umber 2.11 1.74 1 11 0.05 - 0.04 0.07 0.18 - 0.08 0.09 0 .00 - 0.13 - 0.06 0.07 0.05 1 13 Previous actors 0.37 0.31 0 1 0.36 0.38 0.25 0.09 0.02 0.02 - 0.06 0.18 - 0.01 0.22 - 0.22 - 0.13 1 14 Previous production c ompany 0.73 0.28 0 1 0.14 0.19 0.04 - 0.05 - 0.01 - 0.05 - 0.07 0.09 - 0.05 0.09 - 0.26 - 0.11 0.23 58 Table 5 Maximum Likelihood Regression Coefficients and Robust Standard Errors Model 1 Model 2 Model 3 Domestic box office receipts Previous film performance 0.455*** 0.435*** 0.457*** (0.105) (0.116) (0.111) Production budget 0.445*** 0.447*** 0.425*** (0.088) (0.089) (0.088) Ticket price - 0.02 - 0.02 - 0.014 (0.040) (0.040) (0.039) Book based - 0.145 - 0.122 - 0.114 (0.116) (0.115) (0.109) Character based 0.006 0.012 - 0.011 (0.071) (0.072) (0.072) MPAA rating change - 0.234** - 0.224** - 0.184* (0.100) (0.096) (0.092) Star power 0.144* 0.155* 0.190** (0.066) (0.067) (0.070) Unrelated sequel title 0.124 0.199 0.314 (0.803) (0.802) (0.663) Critic reviews 1.251*** 1.209*** 1.167*** (0.140) (0.149) (0.150) Time elapsed - 0.004 - 0.061** (0.014) (0.025) Sequel number 0.037+ 0.028 (0.027) (0.026) Previous actors 0.089 0.072 (0.149) (0.142) Previous production company 0.250+ 0.176 (0.167) (0.153) Sequel number * Time elapsed 0.009 (0.007) Time elapsed * Time elapsed 0.005** (0.002) Previous actors * Time elapsed - 0.045 (0.040) Previous actors * Sequel number - 0.113* (0.060) Previous production company * Time elapsed 0.023 (0.046) Previous production company * Sequel number - 0.068 (0.077) Constant 17.028*** 17.037*** 16.934*** (0.092) (0.091) (0.098) N 499 499 499 AIC 1383.073 1385.86 1372.457 BIC 1425.199 1444.837 1456.709 59 To test the hypothesis I first test the main effects in Model 2 and then test the interactions and curvilinear effects in Model 3. I begin with the main effects for Hypothesis 1, 3, 5 and 7 in Model 2. Because each of the hypotheses are directional, meanin g that I not only hypothesize an effect but also whether the effect is positive or negative, I conduct and report the results of a one tail statistical test. First, Hypothesis 1 argues that longer time between sequel films is negatively related to domestic box office receipts . To test this hypothesis, I include the t ime e lapsed variable in Model 2. While the coefficient is negative it is not statistically significant (B= - 0.004 , p=. 391 ). Thus Hypothesis 1 is not supported. Hypothesis 3 argues that mature franchises are likely to have high er performing sequels and thus sequel n umber will be positively related to d omestic box office receipts . To t est Hypothesis 3 I include the sequel n umber i n Model 2. The coefficient for sequel n umber is positive and margin ally statis tically significant (B=.037 , p=.0 81 ) thus partially supporting Hypothesis 3. To further illustrate these results I plot the linear prediction of s equel n umber across the range of the data along with the standard errors in Figure 1. According to the figure the estimates are more precise at low levels of sequel n umber w hen compared to high levels of sequel n umber. This is to be expected as the mean for sequel n umber is 2.15 and thus the higher number of observations at this level lead to more prec ise estimates . Finally, to further quantify the results I calculated the estimates of the linear combination of Model 1 at the mean level of s equel number as well as two standard deviations above (6 ) and one below because a negative value in sequel number is beyond the data range (1) the mean. Based on this prediction domestic box office is predicted to be $6 .0 million higher for the 6 th sequel when compared to the first . In sum, these results suggest partial support for Hypothesis 3 and suggests that the i ncreasing level of s equel n umber is materially related to d omestic box office 60 receipts . Figure 1 : Predicted values of log domestic box o ffice according to different levels of s equel n umb er and 90% confidence intervals Hypothesis 5 argues that previous actors will be positively associated with d omestic box office receipts . To te st Hypothesis 5 I also include previous actors in Model 2. The coefficient is positive but not s tatistically significant (B=.089, p=.275 ) which does not support Hypothesis 5. Previous Production Company is predicted to be positively related to d omestic box office receipts in Hypothesis 7. To test Hypothesis 7 I include p revious p roduction c ompany in Model 2. The results support Hypothesis 7 as the coefficient is positive and marginally statisti cally significant (B=.250 p=.068 ). To illustrate these results I plot these 61 results in Figu re 2 across multiple levels of previous production c ompany across the range of the data. Based on the figure the standard errors are narrower around .7 and .8 which is the sample mean when compared to the less precise estimates at very low levels of p revious p roduction c ompany. Finally to quantify the strength of the relationship I computed the prediction of d ome stic b ox o ffice receipts at three levels of p revious p roduction c ompany the mean and standard deviation above and below. These calculated results suggest that two standard deviations below the me an (.3 mean of .7) results in $3.6 million difference in dome stic box office receipts and one standard deviation above the mean ( 1) results in approximately $3.9 million increase relative to the mean in d omestic box office receipts . 62 Figure 2 : Predicted values of l og d omestic box office receipts according to different levels of p revious p roduction c ompany and 90% confidence intervals In summary, Model 2 provides marginal support for two of the four hypotheses tested. In particular there is marginal support f or the positive association of sequel n um ber and previous production c ompany with d omestic box office receipts and no support for the negative hyp othesized relationship between time e lapsed and d omestic box office receipts and no support for the positive relationship between p revious ac tors and d omestic box office receipts . The AIC and BIC model fit statistics for Model 2 are 1385.86 and 1444.837 respectively. AIC and BIC fit statistics assess the explained variance relative to the complexity of the model. When the fit statistics of the focal model are lower when compared 63 to the control model then it is suggested to have greater fit, meaning that the increased complexity of the model due to additional coefficients is merited because of the additional variance explained. Because the AIC an d BIC fit statistics for Model 2 are higher relative to Model 1 it suggests that while there are statistically significant coefficients in Model 2 the model does not explain incremental variance to merit the additional complexity. Model 3 tests Hypothesis 2, 4, 6 (a) and (b), and 8 (a) and (b). Due to the repeated use of both s equel n umber and t ime e lapsed in Model 3 there is some risk of multi - collinearity. Multi - collinearity can distort results, making them less generalizable and stable. Typically multi - collinearity is measured using variance inflation factor (VIF) statistics. VIF statistics are calculated based on tolerance , which is the amount of variance of independent variables unexplained by other independent variables (Hair, 1998) . Calculating VIF s tatistics requires calculating the coefficient of determination or R 2 , which cannot be calculated using MLE. Therefore to assess multi - collinearity in Model 3 I ran the model using OLS and found the highes t VIF statistic to be 3.60 for time e lapsed which i s well below the commonly accepted threshold of 10 (Hair, 1998) . Hypothesis 2 argues that the relationship between time elapsed and domestic box office receipts is curvilinear such that it begins negative and turns positive. To test this hypothesis I include th e squared term for time e lapsed in Model 3 as well as the main effect for time e lapsed. In order to interpret the coefficient one must consider both the main and quadra tic terms. The main effect for time e lapsed is negative and the squared term is positive and stati stically significant (B=.005 , p= .0 05 ). These results support Hypothesis 2 and Figure 3 illustrates this relationship. Based on the figure the curvilinear relationship is present and further that the estimates of d omestic box office receipts are more precise at lower levels of 64 time e lapsed when compared to higher levels. In order to quantify these results I compared the estimates at less than a year time e lapsed compared to 11 years which is the point at which d omestic box office receipts is the lowest. The difference in these predicted values is $ 22 million, suggesting that time e lapsed is not only statistically significant but that the difference d omestic box office receipts . Figure 3 : Predicted values for t ime e lapsed on d omestic box office receipts with 90% confidence intervals shaded Hypothesis 4 argues that the negative relationship between time e lapsed and d omestic box office receipts is moderated by s equel n umber such that that relationship is weaker when s equel n umber is high than when it is low. To test this hypothesis I include both of the main 65 effects and their product. In Model 3 the product coefficient is positive and nearly mar ginally significant ( B= . 009, p=.12 ) . Figure 4 illustrates that once again the precision of the estimates is stronger a lower levels of time e lapsed. In addition, the figure considers both the interaction and curvilinear relation for t ime e lapsed on perfor mance. It appears that the most meaningful distinction between high and low levels of s equel n umber occurs between 5 and 20 years of t ime e lapsed where higher levels of s equel n umber are likely to have higher performance when compared to lower s equel n umbe r. Also , there does not appear to be a strong distinction between high and low levels of s equel n umber at very low and very high levels of t ime e lapsed. These results however should be interpreted with caution because the standard errors remain large relative to the effects and statistical significance remains quite low . Notwithstanding these limitations the material effect of these predictors rema ins relatively large. For example, I calculated the predicted d omestic box office receipts at both high and low levels of s equel n umber (1 and 6) at 11 years of t ime e lapsed. This difference represents an incremental $1 6.5 million in d omestic box office re ceipts . 66 Figure 4 : Predicted values of domestic box office receipts based on varying levels of time elapsed and at low and high levels of sequel number shaded areas represent 90 % confidence intervals Hypothesis 6a and 6b refer to moderating relationships between previous actors, time elapsed and sequel number . To test this hypothesis I include both main effects and their product in Model 3. The coefficient for the product of previous actors and time el apsed is negative and not statistically significant (B= - . 045, p<.134 ). Similar to Hypothesis 6a, Hypothesis 6b argues that the positive effect of previous actors is moderated by sequel n umber such that the relationship becomes less positive as s equel n umb er increases. To test Hypothesis 6b I include the product of previous actors and 67 sequel n umber along with their main effects. The coefficient for the product term is negative and st atistically significant (B= - .113 , p =.03 ). This relationship is illustrated in Figure 6 which suggests that similar to Hypothesis 6a the relationship between previous actors and d omestic box office receipts is positive for low levels of s equel n umber but that this relationship becomes negative when s equel n umber is higher. Further by graphing the standard errors this graph suggests that the contingency of s equel n umber is most distinctive at very low levels of previous actors when compared to high levels. Collectively the figures and results in Model 3 offer support for Hypotheses 6a and 6b. In order to quantify these results I calculate the difference in d omestic box office receipts at low levels of previous actors (0) and both low and high levels of s equel n umber (1 and 4). These results suggest that low le vels of p revious actors is associat ed with a $7.6 million increase in d omestic box office receipts when s equel n umber is high rather than low. 68 Figure 5 : Predicted values of deomestic box office receipts based on previous actor and sequel number at both high and low levels shaded areas represent 90% confidence intervals In Model 3 I also test Hypotheses 8a and 8b. Hypothesis 8a argu es that the positive effect of previous production c ompany is moderated by t ime e lap sed such that the relationship becomes less positive when t ime e lapsed increases. To test this Hypothesis I include both of the main effects for p revious p roduction c ompany and t ime e lapsed as well as their product. The product coefficient is positive but not statistical ly significant (B= 0.023, p=.307 ) thus not supporting Hypothesis 8a. Similarly, Hypothesis 8b argues that the positive effect of previous production company is moderated by s equel n umber such that the relatio nship becomes less 69 positive as se quel n umber increases. Testing this hypothesis also requires both main effects and their product be included in Model 5. The product coefficient is negative and not s tatistically significant ( - 0.068, p=0.189 ) thus not providing support for Hypothesis 8b. Finally, to assess model fit I compare the AIC and BIC fit statistics. The fit statistics for Model 3 are 1372.457 and 1456.709 for AIC and BIC respectively. Interestingly, while the AIC suggests that Model 3 is a better fit when compared to the control mo del, the BIC statistic suggests otherwise. Divergence in these measures can occur because of the to AIC but more heavily penalizes complexity in a model when c ompared to AIC. BIC fit statistics are more helpful when seeking to avoid a Type I error in contrast to AIC fit statistics which are more helpful in avoiding Type II errors. Therefore while there may be some support for the hypotheses these results must be interpreted with caution due to model fit. In summa ry while there is sup port for the Hypotheses 2, and 6b along with marginal support for Hypothesis 3 and 5 the null results for the other hypotheses weaken model fit. Suggesting that notwithstanding some s upport for the theoretical framework there remains some uncertainty regarding the size and consistency of these relationships. 70 CHAPTER 7 DISCUSSION Brands can serve as a key strategic resource as firms compete for competitive advantage (Aaker, 1991; Keller, 200 8; M. Peteraf, 1993; Wernerfelt, 1984) . In industries such as books, movies and video games firms face significant challenges exploiting the value of their brands because consumers are unlikely to purchase the same brand more than once. The focus of thi s study is to better understand how firms manage their brands for competitive advantage in these types of industries. In particular , I develop the notion of brand re - creation as a brand management strategy which focuses on launching new products associated with a prior successful brand after some interval during which the brand has largely been absent from the market. In this way, brand re - creation differs significantly from existing models of brand management which largely rely on either repeated purchases or consistent brand presence. Thus brand re - creation considers how firms exploit their brand intermittently with new and innovative products. Investigating brand re - creation as a brand management strategy expands our understanding of how firms manage and exploit market based resources (R. K. Srivastava, 2001) . To date , there has been little research about how firms manage their brands from a broader strategic perspective for competitive advantage. Early RBV research suggested that market based assets suc h as brands could serve as strategic resources (M. Peteraf, 1993; Wernerfelt, 1984) which could be used to exploit market opportunities (Penrose, 1959) . Implicit in this line of thinking was that brands can be created and, once established, become a resource that can be leveraged per petually. There was little thought given to how brands might need to be re - created from time to time with intervals when the brand is largely absent 71 from the market. Thus by investigating brand re - creation I draw upon early RBV work to answer repeated call s for research bridging strategic management and marketing domains (J. Barney et al., 2001; R. Priem, Butler, & Li, 2013; R. L. Priem, Li, & Carr, 2012; R. K. Srivastava, 2001) . This research has also built on the importance of consistency in effective brand management strategies (Keller, 1999; Sethuraman, Tellis, & Briesch, 2011) . In particular, consistency can be broken down into product level similarity and brand concept consi stency (Park et al., 1991) . In the film industry, product level similarity is best achieved by u sing the same lead actors across sequels, and brand concept consistency can be facilitated through the use of the same production company across sequels. By maintaining both of these forms of consistency film studios can signal the quality of their products and fa cilitate more efficient consumption for their audiences. Theoretical implications. There are a number of theoretical implications arising from this work regarding how market based assets , such as brands , can be managed for SCA. These implications include f ormalizing arguments for brands servi ng as a strategic resource, distinguishing brand management strategies according to industry characteristics and link ing the value of brands to their effective deployment. In order to formalize the arguments of brands s erving as strategic resources I draw together assertions of the RBV (M. Peteraf, 1993; Wernerfelt, 1984) with the findings and rationale in the marketing literature regarding brand management (Aaker & Keller, 1990; Berry, 1988) . To date, the development of theory around brands as a resource has been limited by a lack of cross functional research (R. K. Srivastava, 2001) . Thus RBV research lacked specificity re garding the mechanisms by which market based assets like brands evoke 72 value in transactions. Furthermore marketing research has been criticized for a lack of strategic perspective regarding how firms strategically manage downstream for SCA (Dawar, 2013) . In this work , I bring together marketing and RBV research to formalize the arguments explaining how brands m eet VRIN based characteristics. To the RBV literature , this connection provides explanations by which brands serve as value creating assets to enhance value for consumers. Finally , these arguments illustrate the cognitive constraints of consumers which make brands rare and difficult to imitate thus addressing the criticism that the RBV lacks specificity and transaction level mechanisms for SCA (R. Priem et al., 2013; R. L. Priem & Butler, 2001a) . Furthermore, to the marketing literature it provides a framework by which brands can lead to SCA. By connecting these two disparate streams of research I provides a way for future researchers to investigate how market based assets can be managed for SCA. Another theoretical implication relates to associating brand management strategies to industry characteristics. The brand management literatu re has explicated a number of approaches for managing brands to enhance their value (Berry, 1988; Kapferer, 2008; Park et al., 1991) . N otwithstanding these various approaches there has been little high level theoretical work connecting these brand management strategies to the larger industry contex t. Contextual factors have been considered but currently they are not linked to demand based differences in industry (see Fischer et al., 2010 for a notable exception) . Theoretically, this work brings demand based characteristics of different industries to the forefron t by emphasizing the consumer purchase cycle as well as the differences in consumer purchase patterns in determining how brand can be appropriately managed (Dawar , 2013; Pringle et al., 73 1982) . Thus effectively leveraging brand can take many forms depending on the demand based characteristics of the industry. Finally, this research illustrates the interwoven nature of resource characteristics with resource deplo yment strategy. The RBV literature often asserts that resource value is innate, however by focusing on market based assets I bring light to how the value of brands as a resource can be heavily influenced by their appropriate deployment. In particular, I su ggest that the value of brands is contingent on their effective timing and configuration. In this case the characteristics of the resource , specifically their value , is tightly connected to deployment strategy. Strategic implications. This study also emph asizes a number of strategic implications associated with market based assets and their deployment. The RBV has been criticized with (Kraaijenbrink et al., 2010) . This critique argues that firms are always going to be searching for both higher order and higher value resources. To address this critique it is important for researchers to consider the interplay of resources and their management (Kraaijenbrink et al., 2010; Sirmon et al., 2007) . While the RBV asserts that in order to be strategic, resources need to be rare, this rareness relates to the availability of other firms possessing a sim ilar resource. To date there have been fewer studies indicating the potential value of a firm suspending customer access to a given resource in order to enhance its value. In this study, I use such a perspective suggesting that the value of market based assets is partially realized in their scarcity. Strategically there have been a number of firms utilizing this strategy in order to enhance their performance but to date there has been little effort to quantify and specify the usefulness of the strategy. Walt Disney has utilized a scarcity strategy by k eeping their movie 74 the opportunity for consumers to purchase the title except during certain re - releases of the film (Felten, 2011) . In this work , these strategies take on more nuance. For example, the results suggest that it may be important to quickly reinforce an emerging fra nchise by quickly releasing sequel films whereas a more mature franchise may be more appropriately exploited through sequ els mediated by significant time . Furthermore the results suggest a curvilinear main effect of time elapsed with domestic box office re ceipts suggesting that effectively leveraging nostalgia may require more time between sequel films. Thus , this work begins to unpack the nuances of managing the appropriate level of exposure and scarcity for market based assets in this industry. The RBV ha s also faced challenges regarding how to conceptualize and integrate the issue of dynamics. In this study, I suggest that market based assets are unique because the value of market based assets can be highly transient. Because the value of brands is largel y based in the minds of consumers , this value can vary according to changes in brand exposure, reinforcement and satiation. Illustrating these dynamics is a primary focus of this study. Furthermore, by utilizing marketing research I suggest that brands can resources and their management but also how the management of those resources may vary according to their age and timing. In so doing, I conside r the strategic implications of how their effective management may need to change as market based assets age and mature. In addition to considering the role of age and resource management this work also considers how firms can go about making their resour ces more firm specific. Previous brand management research emphasized the importance of consistency in maintaining the brand image (Keller, 1999; Sood & Drèze, 2006) . This approach suggests that firms enhance brand 75 value by maintaining consistent mess aging and content. In doing so however , can be come image. In these cases, the firm becomes dependent on resources that it does not own. For example in the film indus try, acting talent that is contracted to appear in every sequel or production companies that are used to make all of the sequels. This dependence on external resources can allow these contributors to appropriate a considerable share of the revenues that a studio generates from each of the franchise films (Coff, 1999) . For example, in the film industry the Marvel franchise has largely been built and supported by the key main actors, which derive increasing salaries with each sequel (Fritz, 2015) . Thus the success of the franchise becomes increasingly dependent on th e focal acting talent. In this work, I argue that the effectiveness of this consistency can diminish with increased time between subsequent offerings and with increasing numbers of sequels. With some marginal support t hese results indicate a path by which firms can separate value of the franchise brand from the supporting components making the franchise less dependent on their consistent use to enhance the value of the franchise. Thus through appropriate management film studios can effectively separate the value of a given brand from its consistent product components. Consequently, although the studios are able to develop and exploit branding through franchises, the value of this brand to the firm is somewhat limited because of their inability to change lead actors or production companies . It is therefore interesting to find that studios may be able to move away from using the same acting talent by offering more films within the franchise while maintaining performance . This development can allow a firm to continue to benefit from a brand as a fully owned resource without having to depend on external resources to maintain its value for consumers. 76 Managerial implications. This work offers a number of findings helpful to brand managers. First this study invites managers to consider the demand based characteristics of their industry. Essentially, by considering the purchase patterns of their consumers, brand managers and top managers can better allocate resources in order to maintain and exploit their brand. For example, in industries with high customer heterogeneity and repeated purchasing cycles managers should invest in product development to develop brand extensions. Furthermore when managers identify changing consumer preferences , as key demand based characteristics, it may be important for managers to invest in brand revitalization to keep the brand fresh and meaningful to consumers. Finally, in industries with single purchase cycles, changing preferences and short pro duct life - cycles it is important for managers to invest in brand re - creation strategies that enable them to build brand franchises across subsequent offerings. For managers engaging in brand re - creation this study suggests that timing can play an important role in establishing and exploiting a franchise. Brand re - creation is inherently a temporal phenomenon and therefore managers face important decisions regarding how to properly time their sequel products. According to the results of this study it is parti cularly valuable to quickly release sequel films; particularly when the franchise is relatively young. Doing so will help to establish the franchise brand in the market by capitalizing on fresher memories. Furthermore, these results suggest that for mature franchises it can be useful to spread out sequel films in order to avoid satiation and to capitalize on nostalgia. These temporal results also suggest tha t there may be significant opportunity in purchasing older franchises which may evoke nostalgia from audiences with more time elapsed since the previous film. 77 Using brand re - creation and demand - based industry characteristics as a lens I identify product level similarity, concept level consistency as key characteristics predicting the effectiveness of brand re - creation strategies. Specifically , for managers , these results suggest that consistency in key supporting personnel can be critical for the effectiveness of the franchise brand. In the film industry I identify acting talent and production companie s as the key components necessary to maintain product level similarity and brand concept consistency. By maintaining these components managers ensure a connection to their audience and the key tacit knowledge necessary to ensure the franchise brand is cohe rent across sequel films. Therefore , for managers , these results emphasize the importance of consistency in sequel film performance. This study also outlines the opportunity for managers to make the franchise brand less dependent on supporting components. While consistency is important early in the franchise and when sequel films follow in quick succession the importance of consistency declines as the franchise matures and sequels are separated by more time. In this case , managers can change these component s in order to weaken the bargaining influence of contributing members without lowering the overall performance of the franchise. In summary, the results of this study outline the nuances of brand re - creation strategies helping to both explain and predict t he effectiveness of different managerial decisions in the brand re - creation process. Limitations Although broadly there appears to be some support for the theoretical framework there were a number of hypothesis without support specifically the main effect of years lapsed and the interactions with production company. First, some may argue Hypothesis 1 stating that the main effect of time el apsed on d omestic box office receipts being negative would conflict 78 with the curvilinear relationship in Hypothesis 2, w hich states that the effect would start negative and then turn positive. It is empirically possible to simultaneously have both a linear and curvilinear effect depending on the shape of the curvilinear effect. (Weisberg, 2013) . For example, if the general trend of the data is toward a either positive or negative direction then there can be both a linear and curvilinear effect. W hile there is support for t he curvilinear relationship of years l apsed on d omestic box office receipts there is no support for a linear effect. This finding can be due to a number of reasons both theoretical and empirical. First, theoretically it appears that there is some support f or a negative relationship for years l apsed particularly up to about 11 years. However at this point the relationship turns positive. Furthermore from a sampling standpoint , as discussed below , there is some risk of survivor bias because the sequel films a re likely to be produced after a long time period are more likely to be associated with high performing franchise brands. Therefore it is difficult to separate the effect of brands from time because of the selection effect of studio m anagers systematically tends toward more popular franchises for sequel films being produced after longer time periods. Finally, there is also data issue of fewer films being produced after long time periods. As illustrated in Figure 3 the tighter st andard errors at low levels o f years l apsed is partially the result of more data at this level then at higher levels where the standard errors are substantially larger. Another null finding was for the main effect of previous actors on d omestic box office receipts . This null finding i s particularly surpr ising as the interaction with sequel n umber is significant. While there does not appear to be a strong theoretical reason that this hypothesis was not supported , there are a number of empirical challenges that inhibit testing this hypot hesis. The measurement of this variable particularly imposes challenges in accurately 79 measuring its relationship. Currently this measurement is drawn from up to 5 individuals per sequel film based on the prominence of acting talent as reported by IMDB. How ever in many films there may be roles which are not anticipated to be played in sequel films such as the role of the antagonist. Because my measure includes up to five individuals there may be some error in this measurement approach. In this case the incre ased error may have limited the chances of efficiently measuring this relationship because a udiences may only expect recurring characters to be played by the same stars when compare d to lead roles that are not expected to perform in future sequel films. Fo r example, audiences may enjoy the introduction of other actors to play new characters such a love interest or the main villain . While there is some support for a n interaction of sequel number and previous actors there was only very marginal support for ti me lapsed and previous actors and no suppo rt for these interactions with previous production c ompany. The lack of support for the interactions with production company may be due to either theoretical factors or measurement error. First, theoreti cally, I ar gue that the use of previous p roducti on c ompany enables the film studio to capitalize on tacit knowledge and the network of resources necessary for maintaining the brand concept of the franchise. This tacit knowledge however may also be held by directors, acting talent and other important resources associated with the franchise. Thus without these other consistent factors it may be difficult for a production company to maintain the brand concept. Similarly this consistency in brand concept may require a sim ilar or larger production budget when compared to the previous film. Therefore there may be a number of factors necessary to maintain consistency in the brand concept. Finally, for measurement error, this variable included up to eight production companies associated with the focal film. Often franchises , as they mature , add on production 80 companies while maintaining the focal production company. Therefore the lack of results may be due to this measurement error where the focal production company remained stable across sequel films but the measurement of consistency went down with the additi on of each production company across the franchise. Thus future research may look at keeping any previous production company rather than percentage of previous production company. While the results of this study offer insight into brand re - creation strate gies the results are not without limitations. In particular, the results of this study are subject to limitations due to survivor bias. Survivor bias occurs in backward looking studies when the sample chosen includes only those observations which survived to be measured at the time of the study (Smith, 2014) . This bias is problematic because the sampling approach does not give an equal chance to all population observations to be included and therefore is no t random. Survivor bias was Wald was tasked with recommending where to add armor to military aircraft based on data from returning damaged airplanes. Whi le the data included significant detail regarding the number and locations of rounds, the database only included observations of planes that returned rather than those which fell (Mangel & Samaniego, 1984) . While Wald went to significant lengths to compensate for the biased sample the strict assumptions required to derive parameters in the analysis were also limiting. Survivor bias is likely to exist within the sequel film dataset because film franchises are often run by profit driven agents and therefore the choice of producing a sequel film is not random. An ideal sample for this st udy would include an equal chance for both successful and unsucc essful film release. The sample in this work is less likely to include unsuccessful sequel films because of the selection criteria mangers use to produce a sequel film. The decision to 81 produce another sequel film within a franchise is generally influenced by the potential of the sequel film to generate economic profit. When managers do not foresee economic profit from a sequel film they are unlikely to produce and release the film. Therefore, w hile the sample is intended to include all released sequel films, it does not include those franchises which may have been intended to generate sequel films, such as John Carter , but because of poor box office performance did not. With regard to the resul ts in this study , survivor bias possess es a risk for some of the results. Notwithstanding a control variable for previous film performance there remains some risk of survivor bias in the results. For example, Hypothesis 3 argues that sequel number is posit ively related to domestic box office receipts . Because it is less likely for low performing franchises to produce a high number of sequel films, a positive relationship may be due to a survivor bias rather than the impact of the franchise brand on the mark et and subsequent performance. One potential approach for dealing with this issue would be to utilize the backward looking selection data as an advantage. To do so, I created a dummy variable for prequel which was coded as 1 for if the focal film was foll owed with an additional sequel within the dataset and 0 for no subsequent sequel. This variable is intended to control for the survivorship bias by evaluating whether the film led to a subsequent film. In these results the main effects for sequel number wa s not supported, thus the positive effect of sequel number may be due to survivor bias rather than brand impact. Other results such as the interaction between sequel number and years lapsed however are supported. Data availability is a key obstacle to deal ing with survivor bias. Ideally future studies on brand re - creation may better deal with survivor bias with access to film studio planning 82 data. In particular, to measure the intention of each film to become a long running franchise. This approach will be limited in sample size but may better deal with the risk of survivor bias. Conclusion In closing, while research in the RBV has emphasized how resources can be differentially accumulated and maintained for SCA, this view pushes the RBV toward consumers and asks questions regarding how the impact of resources can be contingent on demand factors such as the memories and satiation of consumers with a given brand. In doing so, this research suggests that the effectiveness of resource management and deployment c an be enhanced when these strategies include demand related factors. More broadly still, this research suggests that consideration of demand related factors cannot be relegated only to the marketing discipline; rather , such factors are within the purview of strategic management research , particularly when they have material implications for how firms effectively manage and deploy their resources. 83 REFERENCES 84 REFEREN CES Aaker, D. A. (1990). Brand Extensions: The Good, the Bad, and the Ugly. Retrieved from http://sloanreview.mit.edu/article/brand - extensions - the - good - the - bad - and - the - ugly/ Aaker, D. A. (1991). Managing Brand Equity Aaker, D. A., & Keller, K. L. (1990). Consumer Evaluations of Brand Extensions. Journal of Marketing , 54 , 27 41. Acedo, F. J., Barroso, C., & Galan, J. L. (2006). The Resource - Based Theory: Dissemina tion and Main Trends. Strategic Management Journal , 27 , 621 636. Adner, R. (2002). When are technologies disruptive? a demand - based view of the emergence of competition. Strategic Management Journal , 23 , 667 688. Adner, R., & Snow, D. (2010). Old technolog y responses to new technology threats: demand heterogeneity and technology retreats. Industrial & Corporate Change , 19 , 1655 1675. Adner, R., & Zemsky, P. (2006). A demand - based perspective on sustainable competitive advantage. Strategic Management Journal , 27 , 215 239. Ainslie, A., Drèze, X., & Zufryden, F. (2005). Modeling Movie Life Cycles and Market Share. Marketing Science , 24 , 508 517. Alba, J. W., & Hutchinson, J. W. (1987). Dimensions of Consumer Expertise. Journal of Consumer Research , 13 , 411 454. Anand, J., & Delios, A. (2002). Absolute and relative resources as determinants of international acquisitions. Strategic Management Journal , 23 , 119 134. Anderson, R. E. (1973). Consumer Dissatisfaction: The Effect of Disconfirmed Expectancy on Perceived Product Performance. Journal of Marketing Research , 10 , 38 44. Arend, R. J. (2006). Tests of the resource - based view: do the empirics have any clothes? Strategic Organization , 4 , 409 421. Armstrong, C. E., & Shimizu, K. (2007). A Review of Approaches to Em pirical Research on the Resource - Journal of Management , 33 , 959 986. Barney, J. (1991). Firm Resources and Sustained Competitive Advantage. Journal of Management , 17 , 99. Barney, J. B. (2001). Is the Resource - Perspective for Strategic Management Research? Yes. Academy of Management Review , 26 , 41 56. 85 Barney, J., Wright, M., & Ketchen, D. J. (2001). The resource - based view of the firm: Ten years after 1991. Journal of Management , 27 , 625 641. Basuroy, S., & Chat terjee, S. (2008). Fast and frequent: Investigating box office revenues of motion picture sequels. Journal of Business Research , 61 . Becerra, M. (2008). A Resource - Based Analysis of the Conditions for the Emergence of Profits. Journal of Management , 34 , 11 10 1126. Berry, N. C. (1988). Revitalizing brands. Journal of Consumer Marketing , 5 , 15 20. Bivainiene, L. (2010). Brand Life Cycle: Theoretical Discourses. Economics & Management , 408 414. Boush, D. M., & Loken, B. (1991). A Process - Tracing Study of Brand Extension Evaluation. Journal of Marketing Research , 28 , 16 28. Bowman, C., & Ambrosini, V. (2000). Value creation versus value capture: towards a coherent definition of value in strategy. British Journal of Management , 11 , 1 15. Brown, S., Kozinets, R. V ., & Sherry, J. F. (2003). Teaching old brands new tricks: Retro branding and the revival of brand meaning. Journal of Marketing , 67 , 19 33. Campbell, M. C., & Keller, K. L. (2003). Brand Familiarity and Advertising Repetition Effects. Journal of Consumer Research , 30 , 292 304. Capron, L., & Hulland, J. (1999). Redeployment of Brands, Sales Forces, and General Marketing Management Expertise Following Horizontal Acquisitions: A Resource - Based View. Journal of Marketing , 63 , 41 54. Carpenter, G. S., & Nakamot o, K. (1989). Consumer Preference Formation and Pioneering Advantage. Journal of Marketing Research , 26 , 285 298. Chang, J. W. (2002). Will a Family Brand Image be Diluted by an Unfavorable Brand Extension? A Brand Trial - Based Approach. Advances in Consumer Research , 29 , 299 304. Chang, S. - J., & Singh, H. (2000). Corporate and industry effects on business unit competitive position. Strategic Management Journal , 21 , 739 752. Claycamp, H. J., & Liddy, L. E. (1969). Prediction of New Product Performance: An Analytical Approach. Journal of Marketing Research , 6 , 414 420. Resource - Based View and Stakeholder Bargaining P ower. Organization Science , 10 , 119 133. 86 Coff, R. W. (2002). Human Capital, Shared Expertise, and the Likelihood of Impasse in Corporate Acquisitions. Journal of Management , 28 , 107 128. Collins - Dodd, C., & Louviere, J. J. (1999). Brand equity and retailer acceptance of brand extensions. Journal of Retailing and Consumer Services , 6 , 1 13. Coombs, C. H., & Avrunin, G. S. (1977). Single - Peaked Functions and the Theory of Preference. Psychological Review , 84 . Retrieved from http://search.proquest.com.proxy1.c l.msu.edu/docview/1290809054?accountid=12598 Cooper, R. G., & Kleinschmidt, E. J. (1987). New products: What separates winners from losers? Journal of Product Innovation Management , 4 , 169 184. Crook, T. R., Ketchen Jr., D. J., Combs, J. G., & Todd, S. Y. (2008). Strategic resources and performance: a meta - analysis. Strategic Management Journal , 29 , 1141 1154. Danneels, E. (2011). Trying to become a different type of company: dynamic capability at Smith Corona. Strategic Management Journal , 32 , 1 31. Dawar, N. (2013). Tilt: Shifting Your Strategy from Products to Customers . Harvard Business Review Press. Desai, K. K., & Basuroy, S. (2005). Interactive influence of genre familiarity, star power, and motion pictures. Psychology and Marketing , 22 , 203 223. Desai, M. A., Loeb, G. J., & Veblen, M. (2002). Strategy and Sources of Motion Picture Finance, The. Retrieved from http://www.hbs.edu/faculty/Pages/item.aspx?num=29426 Dierickx, I., & Cool, K. (1989 ). Asset stock accumulation and sustainability of competitive advantage. Management Science , 35 , 1504 1511. Eliashberg, J., Elberse, A., & Leenders, M. A. A. M. (2006). The Motion Picture Industry: Critical Issues in Practice, Current Research, and New Res earch Directions. Marketing Science , 25 , 638 661. Erdem, T., & Swait, J. (2004). Brand Credibility, Brand Consideration, and Choice. Journal of Consumer Research , 31 , 191 198. Farquhar, P. H. (1989). Managing Brand Equity. Marketing Research , 1 , 24 33. Fel Wall Street Journal . Retrieved from http://www.wsj.com/articles/SB10001424052748704904604576333344180431886 Fischer, M., Völckner, F., & Sattler, H. (2010). How Importan t Are Brands? A Cross - Category, Cross - Country Study. Journal of Marketing Research (JMR) , 47 , 823 839. 87 Foss, N. J., & Knudsen, T. (2003). The Resource - Based Tangle: Towards a Sustainable Explanation of Competitive Advantage. Managerial and Decision Economi cs , 24 , 291 307. Wall Street Journal . Retrieved from http://www.wsj.com/articles/avengers - age - of - ultron - kicks - off - the - summer - 1429801731 Garnsey, E. (1998). A Theory of the Early Gr owth of the Firm. Industrial and Corporate Change , 7 , 523 556. Gilmore, J. Marie, & Wansink, B. (1999, March). New Uses That Revitalize Old Brands. Journal of Advertising Research , 39 , 90. Grant, R. M. (1996). Prospering in Dynamically - Competitive Environm ents: Organizational Capability as Knowledge Integration. Organization Science , 7 , 375 387. Hagerty, J. R. (2011). Zippo Preps for a Post - Smoker World. Wall Street Journal - Eastern Edition , 257 , B1 B2. Hair, J. F. (1998). Multivariate Data Analysis . Prent ice - Hall International. Hall, R. (1992). The strategic analysis of intangible resources. Strategic Management Journal , 13 , 135 144. Helfat, C. E., & Peteraf, M. A. (2003). The dynamic resource - based view: capability lifecycles. Strategic Management Journal , 24 , 997 1010. Ho - Dac, N. N., Carson, S. J., & Moore, W. L. (2013). The Effects of Positive and Negative Online Customer Reviews: Do Brand Strength and Category Maturity Matter? Journal of Marketing , 77 , 37 53. Holbrook, M. B., & Hirschman, E. C. (1982). The Experiential Aspects of Consumption: Consumer Fantasies, Feelings, and Fun. Journal of Consumer Research , 9 , 132 140. Holcomb, T. R., Holmes Jr., R. M., & Connelly, B. L. (2009). Making the most of what you have: managerial ability as a source of resource value creation. Strategic Management Journal , 30 , 457 485. Hoopes, D. G., & Madsen, T. L. (2008). A capability - based view of competitive heterogeneity. In dustrial and Corporate Change , 17 , 393 426. Hoskisson, R. E., Hitt, M. A., Wan, W. P., & Yiu, D. (1999). Theory and research in strategic management: Swings of a pendulum. Journal of Management , 25 , 417 456. Innis, D. E., & Unnava, H. R. (1991). The Useful ness of Product Warranties For Reputable and New Brands. In R. H. Holman & M. R. Solomon (Eds.), Advances in Consumer Research (Vol. 18, pp. 317 322). Provo, UT: Association for Consumer Research. 88 Johnson, E. J., & Russo, J. E. (1984). Product Familiarity and Learning New Information. Journal of Consumer Research , 11 , 542 550. Kapferer, J. - N. (2008). The New Strategic Brand Management: Creating and Sustaining Brand Equity Long Term . Kogan Page Publishers. Keller, K. L. (1999). Managing brands for the long r un: Brand reinforcement and revitalization strategies. California Management Review , 41 , 102 124. Keller, K. L. (2008). Strategic Brand Management: Building, Measuring, and Managing Brand Equity . Pearson/Prentice Hall. Keller, K. L., & Aaker, D. A. (1992). The Effects of Sequential Introduction of Brand Extensions. Journal of Marketing Research (JMR) , 29 , 35 60. Kor, Y. Y., & Mahoney, J. T. (2005). How dynamics, management, and governance of resource deployments influence firm - level performance. Strategic M anagement Journal , 26 , 489 496. Kraaijenbrink, J., Spender, J. - C., & Groen, A. J. (2010). The Resource - Based View: A Review and Assessment of Its Critiques. Journal of Management , 36 , 349 372. Landro, L. (1989, June 6). Formula Films: Sequels and Stars Hel p Top Movie Studios Avoid Major Risks --- Some Critics --- Wall Street Journal, Eastern Edition , p. 1. New York, N.Y., United States. Lockett, A., Thompson, S., & Morgenstern, U. (2009). The development of the resource - based view of the firm: A critical appraisal. International Journal of Management Reviews , 11 , 9 28. http://www.mediapost. com/publications/article/244307/lovin - mcdonalds - back - to - health.html Low, G. S., & Fullerton, R. A. (1994). Brands, Brand Management, and the Brand Manager System: A Critical - Historical Evaluation. Journal of Marketing Research , 31 , 173 190. Luan, Y. J., & Sudhir, K. (2010). Forecasting Marketing - Mix Responsiveness for New Products. Journal of Marketing Research (JMR) , 47 , 444 457. Lynch, J. G., Jr., & Srull, T. K. (1982). Memory and Attentional Factors in Consumer Choice: Concepts and Research Methods. Jour nal of Consumer Research , 9 , 18 37. Makadok, R., & Coff, R. (2002). The Theory of Value and the Value of Theory: Breaking New Ground versus Reinventing the Wheel. The Academy of Management Review , 27 , 10 13. 89 Mangel, M., & Samaniego, F. J. (1984). Abraham W Journal of the American Statistical Association , 79 , 259 267. McAlister, L. (1982). A Dynamic Attribute Satiation Model of Variety - Seeking Behavior. Journal of Consumer Research , 9 , 141 150. McAlister, L., & Pessemier, E. (1982). Variety Seeking Behavior: An Interdisciplinary Review. Journal of Consumer Research , 9 , 311 322. McGahan, A. M., & Porter, M. E. (1997). How much does industry matter, really. Strategic Management Journal , 15 30. Milewicz, J., & Herbig, P. (199 4). Evaluating the brand extension decision using a model of reputation building. The Journal of Product and Brand Management , 3 , 39. Morgan, N. A., & Rego, L. L. (2009). Brand Portfolio Strategy and Firm Performance. Journal of Marketing , 73 , 59 74. Nag, R., Hambrick, D. C., & Chen, M. - J. (2007). What is strategic management, really? Inductive derivation of a consensus definition of the field. Strategic Management Journal , 28 , 935 955. Nelson, P. (1970). Information and Consumer Behavior. Journal of Political Economy , 78 , 311 329. Newbert, S. L. (2007). Empirical Research on the Resource - Based View of the Firm: An Assessment and Suggestions for Future Research. Strategic Management Journal , 28 , 121 146. Park, C. W., Milberg, S., & Lawson, R. (1991). E valuation of Brand Extensions: The Role of Product Feature Similarity and Brand Concept Consistency. Journal of Consumer Research , 18 , 185 193. Peltoniemi, M. (2015). Cultural Industries: Product Market Characteristics, Management Challenges and Industry Dynamics. International Journal of Management Reviews , 17 , 41 68. Penrose, E. T. (1959). The Theory of the Growth of the Firm . Wiley & Sons. Pete raf, M. (1993). The cornerstones of competitive advantage: a resource based view. Strategic Management Journal , 14 , 179 191. Peteraf, M. A., & Barney, J. B. (2003). Unraveling The Resource - Based Tangle. Managerial & Decision Economics , 24 , 309 323. Ployhar t, R. E., Nyberg, A. J., Reilly, G., & Maltarich, M. A. (2014). Human Capital Is Dead; Long Live Human Capital Resources! Journal of Management , 40 , 371 398. 90 Pollock, T. G., Rindova, V. P., & Maggitti, P. G. (2008). Market Watch: Information and Availabili ty Cascades Among the Media and Investors in the U.S. IPO Market. Academy of Management Journal , 51 , 335 358. Priem, R., Butler, J., & Li, S. (2013). Toward Reimagining Strategy Research: Retrospection and Prospection on the 2011 AMR Decade Award Article. Academy of Management Review , amr.2013.0097. Priem, R. L. (2007). A consumer perspective on value creation. Academy of Management Review , 32 , 219 235. Priem, R. L., & Butler, J. E. (2001a). Is the Resource - Strategic Ma nagement Research? The Academy of Management Review , 26 , 22 40. Priem, R. L., & Butler, J. E. (2001b). Tautology in the Resource - Based View and the Implications of Externally Determined Resource Value: Further Comments. Academy of Management Review , 26 , 57 66. Priem, R. L., Li, S., & Carr, J. C. (2012). Insights and New Directions from Demand - Side Approaches to Technology Innovation, Entrepreneurship, and Strategic Management Research. Journal of Management , 38 , 346 374. Pringle, L. G., Wilson, R. D., & Bro dy, E. I. (1982). News: A Decision - Oriented Model for New Product Analysis and Forecasting. Marketing Science , 1 , 1 29. Ratchford, B. T. (2001). The Economics of Consumer Knowledge. Journal of Consumer Research , 27 , 397 411. Ravid, S. A. (1999). Informatio n, Blockbusters, and Stars: A Study of the Film Industry. The Journal of Business , 72 , 463 492. Redden, J. P. (2008). Reducing Satiation: The Role of Categorization Level. Journal of Consumer Research , 34 , 624 634. Reddy, S. K., Holak, S. L., & Bhat, S. (1 994). To Extend or Not to Extend: Success Determinants of Line Extensions. Journal of Marketing Research , 31 , 243 262. Roberts, J. H., & Lattin, J. M. (1991). Development and Testing of a Model of Consideration Set Composition. Journal of Marketing Researc h , 28 , 429 440. Schmalensee, R. (1985). Do Markets Differ Much? The American Economic Review , 75 , 341 351. Schmidt, J., & Keil, T. (2013). What Makes a Resource Valuable? Identifying the Drivers of Firm - Idiosyncratic Resource Value. Academy of Management R eview , 38 , 206 228. 91 Sethuraman, R., Tellis, G. J., & Briesch, R. A. (2011). How Well Does Advertising Work? Generalizations from Meta - Analysis of Brand Advertising Elasticities. Journal of Marketing Research , 48 , 457 471. Shamsie, J. (2003). The Context of Dominance: An Industry - Driven Framework for Exploiting Reputation. Strategic Management Journal , 24 , 199 215. Shepard, R. N. (1967). Recognition Memory for Words, Sentences, and Pictures. Journal of Verbal Learning and Ve rbal Behavior , 6 . Retrieved from http://search.proquest.com.proxy2.cl.msu.edu/docview/1297350611/citation/415B2DE 262294B75PQ/1?accountid=12598 Short, J. C., Ketchen, D. J., Palmer, T. B., & Hult, G. T. M. (2007). Firm, strategic group, and industry influen ces on performance. Strategic Management Journal , 28 , 147 167. Simon, H. (1979). Dynamics of Price Elasticity and Brand Life Cycles: An Empirical Study. Journal of Marketing Research , 16 , 439 452. Sirmon, D. G., & Hitt, M. A. (2003). Managing Resources: Li nking Unique Resources, Management, and Wealth Creation in Family Firms. Entrepreneurship: Theory & Practice , 27 , 339 358. Sirmon, D. G., Hitt, M. A., & Ireland, R. D. (2007). Managing firm resources in dynamic environments to create value: Looking inside the black box. Academy of Management Review , 32 , 273 292. Smith, G. (2014). Standard Deviations: Flawed Assumptions, Tortured Data, and Other Ways to Lie with Statistics . The Overlook Press. Sood, S., & Drèze, X. (2006). Brand Extensions of Experiential Go ods: Movie Sequel Evaluations. Journal of Consumer Research , 33 , 352 360. Spiggle, S., Nguyen, H. T., & Caravella, M. (2012). More than fit: Brand extension authenticity. Journal of Marketing Research (JMR) , 49 , 967 983. Srivastava, R. K. (2001). The resource - based view and marketing: The role of market - based assets in gaining competitive advantage. Journal of Management , 27 , 777 802. Srivastava, R. K., Shervani, T. A., & Fahey, L. (1998). Market - Based Assets and Sharehold er Value: A Framework for Analysis. Journal of Marketing , 62 , 2 18. Stigler, G. J., & Becker, G. S. (1977). De Gustibus Non Est Disputandum. The American Economic Review , 67 , 76 90. Teece, D. J., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and str ategic management. Strategic Management Journal , 18 , 509 533. 92 Thomas, S., & Kohli, C. (2009). A brand is forever! A framework for revitalizing declining and dead brands. Business Horizons , 52 , 377 386. Tripsas, M. (2008). Customer preference discontinuitie s: a trigger for radical technological change. Managerial and Decision Economics , 29 , 79 97. Vany, A. D., & Walls, W. D. (1999). Uncertainty in the Movie Industry: Does Star Power Reduce the Terror of the Box Office? Journal of Cultural Economics , 23 , 285 318. Vogel, H. L. (2010). Entertainment Industry Economics: A Guide for Financial Analysis . Cambridge University Press. Völckner, F., & Sattler, H. (2006). Drivers of Brand Extension Success. Journal of Marketing , 70 , 18 34. Weisberg, S. (2013). Applied Li near Regression . John Wiley & Sons. Wernerfelt, B. (1984). A Resource - based View of the Firm: Summary. Strategic Management Journal (pre - 1986) , 5 , 171. Wernerfelt, B. (1985). Brand loyalty and user skills. Journal of Economic Behavior & Organization , 6 , 38 1 385. Wernerfelt, B. (1988). Umbrella Branding as a Signal of New Product Quality: An Example of Signalling by Posting a Bond. The RAND Journal of Economics , 19 , 458 466. Wiseman, R. (2009). On the use and misuse of ratios in strategic management research . Research Methodology in Strategy and Management , 5 , 75. Wooldridge, J. M. (2003). Cluster - Sample Methods in Applied Econometrics. The American Economic Review , 93 , 133 138. Wyer, R. S., & Srull, T. K. (1986). Human cognition in its social context. Psycho logical Review , 93 , 322 359.