Introduction

The marketing literature contains quite a few studies focused on contexts that seem to have limited generalizability to others. For instance, it is not straightforward to see how the screen allocation model for multi-screen movie theatres of Swami et al. (1999) could be applied to other empirical settings. In some cases, over time, a substantial literature stream has emerged focused on one such specific context. An example is the pharmaceutical marketing literature (Gönül et al., 2001; Mantrala et al., 1994; for other representative papers, see below). Video games, movies, platform markets, sharing economy, technology, sports, and mobile are other examples where research contributions tend to be highly specific to those contexts.

At the same time, informal scholarly debate lingers on whether such contributions should be published in major marketing journals because they lack generalizability or address a relatively small (i.e., specialized) audience. Prospective authors of context-specific studies often meet (initial) resistance in the adoption and diffusion of their research (e.g., peer review, impact, conference audiences) as reviewers, readers, and listeners may jump the fence on generalizability. Such feedbackFootnote 1 may (1) dismiss the specific context altogether (e.g., “Why should the readers of a broad-based journal like [journal] be concerned about [specific context]? Are the topics presented in the research agenda substantially more than an industry-specific version of (say) the MSI research priorities?“); or (2) ask to generalize what is not generalizable (e.g., “Could you argue that [specific context] is part of a broader array of … arrangements (this would increase the generalizability of your research). A broadened perspective would increase the potential impact of your research. I suggest that you position less on your context and more on …”. We believe that these types of statements reflect a tension between lack of generalizability and scientific merit that is often misunderstood. The current paper aims to clarify why and when context-specific studies with limited generalizability (at least, initially) have indeed scientific merit and to provide recommendations to authors and journal stewards on how to develop them (well).

To clarify what we mean by “context”, we refer to the setting in which a phenomenon is studied. While environmental factors sometimes can be usefully operationalized as a set of moderators in a study, contexts are difficult, non-sensical, or even impossible to operationalize because: (1) there is no variance across subjects in the extent to which they belong to the context (i.e., the context is equally present for all subjects of the study); and (2) contexts are holistic environments that cannot be usefully broken down into characteristics that can be introduced separately as environmental factors.

Not all contexts are “specific”. We consider a context specific when (1) the research findings obtained in studies in that context may be considered to have limited generalizability to other contexts; and (2) the limited generalizability of the findings is due to the uniqueness of the context. With uniqueness of context, we mean a context that is solitary in type or characteristics and not equal to any other. Such uniqueness—take as an example the context of healthcare—could arise from environmental factors (e.g., healthcare markets fall under a specific regulatory framework to protect human life), characteristics or roles of market stakeholders (e.g., the role of the doctor in healthcare markets), or marketing mix elements (e.g., the firm is not completely free in setting price and promotional support in many healthcare markets), or a combination thereof. We think of context as either unique or not unique.

We consider both conditions (uniqueness of context and limited generalizability of findings) to be necessary for a context-specific study. A context of a study can be unique (e.g., social influence among doctors in new drug diffusion as in Iyengar et al., 2011), which does not need to imply that the effects studied cannot be generalized to other contexts (e.g., social influence of key opinion leaders for the adoption of new products in other categories). Also, papers in a single context do not need to be context-specific, because the single context may not be a unique context. For example, the effect of the first review on subsequent product evaluation for vacuum cleaners (Park et al., 2021) is expected to exist also for other product categories.

In sum, we define a context-specific study as a study in a unique setting, yielding conclusions that can be considered to have limited generalizability to other settings. A firm’s industry is a common context of a study. When the effects studied limitedly generalize to firms in (all) other industries we may say that industry is an unambiguous example to operationalize context-specificity. In the rest of this paper, for the sake of clear argumentation, we will continue with specific industries as the operational example of context. Specific industries also have quite a legacy in the marketing literature (e.g., think of literature on movies, pharmaceuticals, or video games) as specific contexts in which studies may be set and provide a rich source for best practice examples. We explore other forms of context-specificity in the background and discussion sections of this paper.

We need to recognize that many marketing scholars see the generalizability of their (or, even more commonly, others’) findings as the gold standard. Often, scholars derive this gold standard from an admiration for the “hard” sciences such as physics or chemistry in which universal laws exist, such as the law of gravity. This gold standard is interpreted by many as an ideal-world construct, and often we are told to adopt it as closely as we can.

Nonetheless, we need to better appreciate that this ideal-world search for generalizability does not befit very well the marketing field, which is intrinsically context-bound. For instance, in teaching, physics professors teach the derivation and empirical verification of general laws. As marketing professors, we teach the analysis of context-specific business cases and the translation thereof to our context-specific managerial world to improve decision-making. Moreover, marketing scholarship often studies a dynamic context (i.e., the context undergoes frequent change). Researchers on physics can formulate laws that are supposed to have staying power. In marketing, new realities, concepts, constructs, and relationships are created and constantly evolve; thus, innovations or changes may arise in the (specific) context that are themselves again specific. Thus, we need to recognize that marketing scholarship is a “softer” science that generates knowledge essentially to help the practice of management in a context (e.g., Drucker & Zahra, 2003; Stremersch et al., 2021). This recognition leads to the adoption of the scientific method to intrinsically derive context-dependent propositions, hypotheses, or findings.

Marketing scholarship is not alone in that “practice of science” or “science of practice.“ Other practices of science develop a significant quantity of research that is valuable not because it is generally applicable, but because it is very specific in its reach. For instance, in the practice of medicine, certain therapies or diagnostics may only work with certain genetic predispositions or diseases (e.g., orphan drugs for rare diseases). Interestingly, in law studies, as a science of practice, it is customary to analyze case studies that are published in leading law journals. On the contrary, while marketing professors often also embrace case studies as a teaching method, leading marketing journals would not accept them as a scientific research method; maybe there is an opportunity to gear case studies more to the scientific method.

Our diagnosis is that the generalizability-specificity dilemma often is misunderstood. Generalizability is excessively heralded as the ideal, and studies in specific contexts are too often denigrated, while both can be intrinsically valuable to the advancement of knowledge. The present paper aims to provide a more nuanced system of beliefs for marketing scholarship to adopt in favor of context-specificity. In short, we want marketing scholars to show more appreciation for the limitations we have in our field to provide generalizations and more admiration for rigorous context-specificity, i.e., propositions, hypotheses, or findings in a very specific context without too much hope for generalization outside context, at first.

There are several reasons to be (more) positive about the scholarly value of context-specific studies. Context-specific studies may bring, among others: (1) high internal validity (i.e., accuracy and precision in the context may be higher than across contexts); (2) high importance for stakeholders in the respective context (e.g., pharmaceutical marketing managers may be more interested in detailing elasticities for prescription drugs than salesforce effectiveness in general or in industrial markets); (3) innovation (i.e., creative new ideas may arise in specific contexts and have general appeal even if specific empirical results do not generalize).

The rest of the paper is structured as follows. First, we discuss the concept of context-specificity in marketing and elaborate on its sources. Then, we present the most important scholarly advantages of doing research on specific contexts. Next, we provide a helping hand when developing and publishing context-specific studies and how to do this well. Beyond suggesting practices to adopt, we also review two exemplars of context-specific literature streams (i.e., the movies and pharmaceutical literatures). We end by offering clear implications for scholars.

Generalizability and context-specificity

To fully appreciate context-specificity, a good starting point is to clarify what we mean by generalization and generalizability. We clarify that context-specificity entails a limited “generalizability to” other contexts outside of the context the study was specifically developed in. Next, we clarify that a context may derive its uniqueness from multiple sources, such as the environment, the market stakeholders, and the marketing mix. We exemplify these sources of uniqueness by studying the movies’ literature.

Generalization and generalizability

When we refer to the generalization of propositions, hypotheses, or findings, we refer to the extrapolation of a causal relationship over variations in people, organizations, settings, treatments, or outcomes (Cook et al., 2002). Such generalizations can offer laws or principles which may serve as a basis for prediction, decision, and action (Bartels, 1951). Generalizability refers to the extent to which propositions, hypotheses, or findings can be generalized based on logic, mathematics, or empirical documentation.

To understand the tension between generalizability and context-specificity, it is helpful to adopt the distinction introduced by Lynch (1999) between (1) generalizing “across” subpopulations of some larger population (i.e., “within the population”) and (2) generalizing “to” some defined population. While context-specific studies fail–by definition–at least to some degree the “generalizing to” test, they do not necessarily (or should we say, do not ideally) fail the “generalizing across” test.

For example, a study of Direct-To-Consumer-Advertising (DTCA) in pharmaceutical markets may not “generalize to” other industries because (1) DTCA may prompt patients to request a drug by brand name from their drug prescribers who as regulated gatekeepers may or may not prescribe the respective drugs (Stremersch et al., 2013); and (2) this process does not replicate in the same way in many other markets routinely studied in marketing (such as grocery retailing, automotive, or e-commerce platforms). At the same time, pharmaceutical marketing scholars may aim to “generalize across” drug categories and, therefore, observe careful sampling properties and examine category-specific moderators (e.g., drug characteristics as in Venkataraman & Stremersch, 2007).

There are several ways in which context-specific studies may fail to “generalize to,“ i.e., show limited generalizability. A first way is that “constructs may not travel.“ A unique context may lead to the identification or inclusion of constructs in a mental model that may not be useful to incorporate in other contexts. For instance, optimizing the number of screens in a movie theatre scheduling decision is a construct that does not travel well and is incomparable to other distribution channel constructs in different contexts.

A second way is that constructs may be measured differently (i.e., incomparably) from one specific context to the other. For example, the study of promotional effectiveness quite typically takes on context-specific measurement indicators or proxies, whether it be discounts and features in grocery retailing or physician detailing in pharmaceuticals, making these measures ultimately incomparable from one context to another.

A third way in which context-specific studies may fail to “generalize to” is that relationships between constructs take a different empirical realization from one context to another. For example, consider the effects of celebrity endorsements on brand equity or sales. While the constructs and the measures may be similar from one context to another, the empirical realization may differ. For example, top athlete endorsements for sports gear (e.g., Tiger Woods for Nike Golf balls) may have different commercial returns for sports gear manufacturers than celebrity endorsements (e.g., George Clooney for Nespresso) may have for consumer goods manufacturers (Van Everdingen et al., 2019).

In Fig. 1, we summarize such logic by exemplifying a specific context A and ways in which studies may fail to “generalize to” context B, in the variables (X, Y, Z to X’, Y’, and Z’), or the relationships (β and γ to β’ and γ’) that are unique to the specific context.

Fig. 1
figure 1

Study of effects in specific contexts. Note: the figure shows that contexts might differ due to different dependent, independent, and moderating variables (Y, X, and Z),  and relationships between them (\(\mathrm\beta\)). However, reality is far more complex: contexts might differ due to different mental models, structural/mathematical relationships among constructs, variable measurement, empirical realizations, etc

Full size image

Context-specificity

Figure 1 also contains the sources of context-specificity that may lie at the origin of such intrinsic lack of “generalizability to”. A context may derive its uniqueness from multiple sources, such as the environment, the market stakeholders, and the marketing mix, each of which may result into constructs, measurements, and relationships among constructs, built-in research mental models, that may be difficult to generalize to others context outside of the one where they were developed for. Next, we elaborate on each of these sources of context-specificity and exemplify them by reviewing the literature on movie marketing (Table 1 illustrates in more detail the sources of context-specificity in the marketing of movies).

Table 1 Sources of context-specificity in the marketing of movies
Full size table

Environment

A first set of sources for context-specificity may lie in the environment a context is in, defined by its legal principles, technology, politics, sociology, and economics. Consider the movie industry and the specific environmental factors that it is subject to that distinguish it from other sectors. The movie industry is subject to specific legal principles, not so much affected by the government, as in the life sciences industry, but by the legal practices the different players follow. For example, studios and distributors usually engage in initial short contracts that are then negotiated every week, depending on the movie performance (e.g., Eliashberg, 2006; Sawhney & Eliashberg, 1996).

Also, technology has a specific influence in this industry. Digital technologies are impacting both production and distribution, and new digital channels are raising the importance of non-theatrical windows (Eliashberg, 2006). For example, streaming services such as Netflix and HBO are on the rise, with new and very specific business models. In addition, these channels are also establishing new contractual agreements with studios. Hence, legal and technological factors are fundamentally affecting the way the industry operates in a specific matter.

Movies also have a specific political and sociological influence. For instance, movies produced in Hollywood in the U.S. dominate the global theatrical market (Eliashberg, 2006) and affect political views across the globe. Moreover, movies are cultural products that are sometimes difficult to transfer to other countries (Gao et al., 2020).

Finally, the economics of movies are very different from other products. They usually have high production costs, high marketing costs, and high failure rates (Eliashberg et al., 2000; Eliashberg, 2006). Movies are also products of high economic importance for exports, but this is only for a few countries and irrelevant for others.

Stakeholders

A second source of specificity lies in the roles of market stakeholders, routinely analyzed according to the following 4 Cs: companies, customers, collaborators, and competitors. In the movie industry, the leading companies in the value chain are producers, distributors, and exhibitors. The business model of each of these players is unique and, as mentioned before, is heavily impacted by the high costs of production and marketing together with a high failure rate. That is why much of the early literature in the industry has been concentrated on (early) forecasting of box-office revenue (Eliashberg et al., 2000; Neelamegham & Chintagunta, 1999; Sawhney & Eliashberg, 1996).

Customers of movies also behave quite differently as compared to customers of other products. They, for example, do not act instantaneously on the motivating information they receive about new movies (Sawhney & Eliashberg, 1996). Additionally, customers spend a disproportionate amount of time-consuming entertainment (Eliashberg et al., 2000; Eliashberg, 2006; Weinberg, 2006), being perhaps the most influential product category for younger audiences, and the one that has a higher impact on their viewpoints. Because movies are an experiential product, they require moviegoers to rely on movie-related information to judge quality. Hence, the level of consumers’ expectations determines the success of the opening week (Chakravarty et al., 2010). Finally, because of all these factors, customers are increasingly connected and share positive and negative word-of-mouth through social media and specific recommendation sites such as yahoo movies or IMDB (Dellarocas et al., 2007; Godes & Mayzlin, 2004, 2009; Lehrer & Xie, 2021; Liu, 2006).

This industry also gives rise to some specific collaborators. For example, in the movies industry, critics and critics aggregators are important and play a specific role unlike in any other industry (Austin, 1984; Eliashberg & Shugan, 1997; Ham et al., 2021). Finally, competitors are tough to define since a history of M&As among studios, distributors, and theaters have blurred some of the old rivalries, and now most players work with all others or even co-finance new projects.

Marketing mix

The features of movies are also specific when compared to other products, even cultural ones. For example, with experience products they share being highly subjective and emotional (Basuroy et al., 2003; Eliashberg et al., 2000; Nelson, 1970), but movies have specific features, such as genre, sequel information, star power, distributor, or MPAA rating. Movies have also pioneered brand franchises to reduce risk, which explains the popularity of sequels. However, this practice has been challenging to scale. Still today, most movies have to establish a brand from scratch, building on the reputation of their star power (Eliashberg 2006; Sood & Drèze, 2006).

The movies industry also employs unique channels. For example, the signaling effects of advertising spending impact both screen allocation and customer expectations. Hence, modeling advertising and distribution is especially difficult because of endogeneity issues (Basuroy et al., 2006; Elberse & Eliashberg, 2003; Shugan, 2004). A distinct channel topic is that of the existence of sequential windows: local and global theaters, streaming, network TV, pay TV, video games, merchandising (Eliashberg, 2006).

As for pricing, the standard practice of theaters is to have extremely low variability (Eliashberg, 2006; Sawhney & Eliashberg, 1996), while for streaming services, there is not a unique consumer price for specific movies and TV series.

As for promotion, the general use of advertising is not specific, but what is specific is the short time window to capture customer attention (Sawhney & Eliashberg, 1996). And while analogs exist in theater (Tony awards) and Television (Emmy awards), the Oscars are a promotional vehicle that is certainly unique for this industry and thus has raised attention among scholarly research (Eliashberg & Shugan, 1997).

Prevalence of context-specific studies in marketing

Now that we provided theoretical grounding and explored specific examples, a next foundational question is whether context-specific studies are at all relevant to marketing as a field. And whether we can validate such relevance somehow beyond our own observations, which may be potentially qualified as anecdotal. We believe there are two ways to offer at least some validation for the relevance of context-specificity to our field.

A first way to validate this view is by literature study. Have others before us raised some of the issues above, and if so, how? Several authors before us have subscribed to the particular relevance of context-specific studies in arguing that universal “generalization to” is often not possible in marketing. Leone & Schultz (1980) state that “when we report, for example, that advertising has a positive influence on sales, we do not imply that this is true in every circumstance. It simply means that there is corroboration for this proposition from a number of sources for particular types of goods.“ (p. 12). Zinkhan & Hirschheim (1992) advocate that it is unreasonable to strive for fundamental truths in marketing because human behavior, which is “mutable, unpredictable, and reactive” (p. 80), is central to the field. Sheth & Sisodia (1999) support the notion that marketing is by definition a context-driven discipline and, importantly, the contextual elements in which marketing operates evolve continuously and rapidly (e.g., economy, societal norms, demographic characteristics, public policy, globalization, or new communication technologies): “As marketing academics, we need to question and challenge well-accepted lawlike generalizations in marketing” (p. 84).

A second way is to empirically gauge how prevalent context-specificity may be in marketing by counting how many papers in marketing study only one industry. We do not suggest by any means that single-industry papers need to be context-specific. However, by default, one could at least question whether the results of a single-industry study can be generalized to other industries making the reflection about context specificity a relevant one for the marketing field.

We took a convenience sample of all papers using field data published in four leading marketing journals in 2019 and 2020 (Journal of the Academy of Marketing Science, Journal of Marketing, Journal of Marketing Research, and Marketing Science). Of the 257 papers in our sample, we found that 64% of the papers use data from one single industry and 7% from two industries (see Table 2).Footnote 2 Thus, single-industry studies are very common in marketing.

Table 2 Prevalence of single industry papers in leading marketing journals: 2 out of 3 studies in marketing present evidence from only one industry
Full size table

The economic activities most routinely examined in these single-industry papers seem to be very diverse and with a long tail (see Table 3). Among the most studied industries, some are “traditional”, such as retail, FMCG, banking, and automotive (with 18, 18, 14, and 14 papers, respectively), while others are more recent (such as platforms and e-commerce with 10 and 8 papers, respectively). In the long tail, there are all kinds of traditional and newer industries with one or two papers (e.g., airlines, construction, industrial equipment, user-generated content, hybrid vehicles, e-books, or craft breweries, to cite a few).

Table 3 Frequency of single-industry studies according to which industry studied
Full size table

Of course, these descriptive statistics do not mean that these single-industry studies do not have the potential to generalize beyond their specific industry context, nor have we presented any evidence that they do not. While one could consider coding a collection of papers on context-specificity, this is not trivial for multiple reasons. First, context-specificity cannot easily be derived from the paper’s claims itself as authors or editors may have pushed generalizability beyond the bounds the paper’s study allows (scholars and editors may vary in their motives and discipline to correctly bound studies). Second, context-specificity may be time-dependent. While scanner data was once unique to the grocery retail setting, it ultimately developed beyond its original context and the insights derived on consumer choice became quite generalizable. Third, context-specificity may be in the eye of the beholder making coding subjective. What is specific to a context for one may be a generalizable insight for others. This variance across observers is, in fact, a main motivation for the present paper to increase sophistication on context-specificity.

Overall, informed by the above conceptualization, examples, prior literature validation, and empirical occurrence of single-industry studies in marketing, one may wonder: If context-specific studies have so much prior art in marketing, there must be some real advantages they present to the field, which we review next.

The beauty and the beast: Scholarly advantages from context-specific studies for marketing

While generalizability seems the gold standard, it would be simplistic to “…assume that more broadly applicable results are always more desirable” (Cook et al., 2002, p. 19). Instead, we should always consider the dark side of generalizability: generalizability may artificially inflate perceived relevance, while, in fact, it may be a beast that destroys it. Or, to frame it positively, we should also consider the beauty of context-specific studies, which lie in multiple aspects. First, context-specific studies may access new and reliable data with high internal validity. Second, findings in context-specific studies may be important for a meaningful audience, as well as conceptually impactful in the future. Third, context-specific studies may fuel innovation, as new phenomena or relationships are often first discovered in unique circumstances, at the time often first considered a context-specific departure from generalized principles. Next, we elaborate on these beautiful features of context-specific studies, each in turn.

The beauty of context-specific studies: Internal validity

A first reason why context-specific studies are beautiful is that they can offer high internal validity as a return on sacrificing external validity. We are not the first to point out the tension between internal and external validity. It is well known that increasing external validity (i.e., increasing the generalizability “to” and “across”) may come at the expense of internal validity (i.e., accuracy and precision inside the specific study context) (e.g., Lynch, 1982, 1983). Context-specific studies may achieve higher levels of internal validity of cause-and-effect relationships if they offer superior measurement opportunities.

For instance, a lot of the literature on inter-firm alliances is set in the biopharma industry (e.g., see Wuyts et al., 2004). In this industry, there is a vibrant intermediary industry to monitor the existence of such alliances that has a strong incentive to provide highly reliable data purchased by venture capitalists or pharmaceutical firms in the context of investment or M&A decisions. Similarly, highly reliable intermediaries such as IQVIA (previously IMS Health) operate panels of physicians, under the supervision of the American Medical Association, in which they monitor sales visits received and scripts scribed, enabling salesforce effectiveness studies, specifically set in this industry. These types of datasets are available for different geographic and therapeutical markets enabling the replicability of the findings. It is hard to find or build databases that achieve an equally high measurement reliability across industries.

Context-specific studies may also gain accuracy from controlling possibly confounding processes on the focal relationship under study. For example, in many industries, other variables such as pricing, discounting, or advertising will interfere with the relationship between sales visits and sales. Therefore, we would need to control for these other processes when estimating the effectiveness of sales visits in generating sales. Still, we may not have that data, or the interference of these processes may be very complex to model. At times, what context-specific studies enable is canceling out some of these interferences. For instance, in the highly regulated context of pharmaceutical markets, doctors often do not consider prices (that are paid by a third party) in their prescription decisions, and Direct-to-Consumer advertising is not allowed in many markets (exceptions are U.S., New Zealand, and to some extent Canada). Thus, in pharmaceutical markets, we do not need to control for such interfering processes as they do not interfere.

The beauty of context-specific studies: Importance

A second reason why context-specific studies are beautiful is that they may be important for the stakeholders in the respective context. Studies show a higher importance as they bring a larger magnitude of change among a larger number of higher-status stakeholders (Stremersch, 2020). Of course, context-specific studies are, by their very nature, limited in the number of stakeholders (i.e., only those belonging to the context of the study). At the same time, context-specific insights can gain in importance to such stakeholders the more specific they get. Thus, a finding that is specific to a given context might be more valuable than more generalizable knowledge. For example, pharma marketers may be more interested in detailing elasticities for prescription drugs than salesforce effectiveness in industrial markets.

The specificity of the insights of a scholar may also elevate a scholar’s standing more easily in the context of choice than if the scholar is a generalist. When companies are searching for key expert witnesses or key advisors to boardrooms, some deeper experience of the specific context in which they operate is typically desired. The high standing of Anita Elberse in the entertainment industry, Jan-Benedict Steenkamp in FMCG, or Andy Zoltners in the pharmaceutical industry is not uncorrelated to their context-specific work in these respective areas.

The reason is that we develop knowledge not only for academic reasons but also for the practice of marketing management. And managers care more about how effects apply to their specific context instead of cross-context averages. Generalizations may artificially inflate the perceived wider reach of a study but decrease the depth of the impact it has on the (intended) larger audience. What is the value for firms of knowing the average advertising ROI across many industries and firms is x%? Which firm wants to be the “average” firm? And which firm believes that the decision it takes will have “average” effects? It seems reasonable that at least some managers may appreciate specificity over generalizability. Hence, while it is true that findings may gain importance as they reach a wider audience (i.e., be more general), it is also true that there is another way: to gain importance by being more interesting for a narrower base big enough to matter. In sum, studies in a specific context, while they reduce the size of the potential audience, have the potential to be more impactful for the specific audience that matches their context.

The beauty of context-specific studies: Creativity and innovation

A third reason why context-specific studies are beautiful is that they enable creativity and innovation. An excessive emphasis on generalizability might stifle innovation: new or not previously observed constructs and phenomena are bound to first arise in a bounded context, especially insights of the more radical type. Innovation literature teaches us that radical innovations typically find a first market base in small, very unique applications; before they may expand into more mass usage, of which it is unclear if they ever will (Moore & McKenna, 1999). Think of the typical Christensen example of the transistor being first applied in the small specific market of hearing aids, only finding mass appeal after a long incubation time in the TV market where it replaced vacuum tubes. However, many radical innovations also exist that remained small in their application context and never “generalized” (e.g., supersonic passenger jets).

Consider studies in marketing on video game consoles (as in Binken & Stremersch, 2009; Landsman & Stremersch, 2011; Shankar & Bayus, 2003), which were for a long time believed to be context-specific, with their unique platform economics setting. In such a setting, platform owners (e.g., console manufacturers) have a coordination problem with complement providers (e.g., game developers and publishers), and the joint outcome of their supply will affect consumer demand for both platforms and games. Over time, platform economics became more omnipresent, with the economy getting increasingly interconnected. And, after video game consoles came online gaming, mobile games, mobile app platforms, and the sharing economy. This societal evolution enabled the relevance of the original work in games to areas that were inconceivable at first for scholars working on games. In sum, context-specific research has the potential to spark new ideas. Next, we proceed to propose some practical recommendations on how to develop and write studies with a high level of context specificity.

Cosmetics or inner beauty? How to develop and write a context-specific study

To fully excavate the inner beauty of context-specific studies, we now detail several guidelines authors may wish to follow to develop and write a context-specific study. In the development of a context-specific study, authors may want to immerse themselves deeply in the context and seek data sources specialized in this unique context. In the writing of a context-specific study, authors may want to clarify why the study in that specific context is important, despite its lack of general appeal, and may wish to define well the bounds of the particular context. Next, we discuss each of these four elements.

Developing a context-specific study: Immersion

A good starting point for scholars interested in context-specific studies is the immersion in a specific context of choice that is sufficiently important (more on this below). For good research ideas to emerge, a deep immersion in the specific context is essential to develop a thorough enough understanding of the context studied and generate meaningful insights relevant to the stakeholders in that context (Stremersch, 2020). One can think of several examples of useful sources into which one could immerse oneself (keeping in mind that in this paper, we take the industry as the example of choice on context).

A first source are industry-specific conferences. Industries with unique features with unique research questions often have their own specific conferences (e.g., think of the World Mobile Congress in Barcelona). Similarly, many industries also have their own industry-specific magazines. For instance, the automotive industry has many industry-specific outlets (e.g., Automotive News). Regular reading of such magazines enables one to more deeply grasp the institutional characteristics of an industry (e.g., the New Product Development Cycle regularities of the car industry) or the priority of certain unique decisions (e.g., F1 entry or exit decisions for car or engine makers). Such institutional knowledge of industries provides scholars with a deeper understanding of the importance of potential research topics and whether particular study implications the researcher envisions are real or are merely a product of the researcher’s most lively imagination arising at the top of the ivory tower.

These two sources (industry-specific conferences and magazines) also lay an excellent foundation for a third source, namely conversations with managers in the specific context. Prior research has already emphasized the value of consulting in the study of important academic research questions (Roberts et al., 2014; Stremersch, 2020). Often, managers are more informed about the nuances of their industries, and they have insights that are novel or counterintuitive for a more general audience. Specific characteristics of the context are often the reason why a general pattern does not work equally well in that industry compared to others. Explaining and testing empirically those specific insights might be a great origin of a study in a specific context.

Developing a context-specific study: Data

Next, in developing a context-specific study, data sourcing may be enabled by industry-specific data opportunities. The immersion subsection above may also enable contacts with prime data providers in the respective industry. Roberts et al. (2014) give a central position to marketing intermediaries in their marketing science value chain. An important group concerns data intermediaries, which are excellent sources of context-specific data. Think of IQVIA for pharmaceutical market data, Polk and JD Power for automotive market data, NPD for video game data, or GfK for retail data.

Also, primary data gathering can benefit from a deep liaison in a specific context. Consider, for example, the work of Li et al. (2020) on how salespeople working in teams adjust effort as the abilities of their coworkers change. They investigate this question using a field experiment in retail booths at a major department store in China. While it is not completely clear that the findings of this study are replicable with other types of salespeople, industries, or countries, this paper is the only empirical study that has analyzed rigorously how different payment schemas may affect the effort of members of a sales team with heterogeneous skills.

Writing a context-specific study: Justify the importance

We have argued above that the specific is not equal to the unimportant. Still, it is vital for scholars as they seek publication of their context-specific work to justify such importance in their paper as much as possible. Some specific contexts are important per se since they are interesting for an audience big enough to matter to support the publication of a study in a journal. An indicator of the size of this audience is the size of the field related to the specific context. For example, Chung (2013) investigates the monetary effects of operating a winning athletics program for an academic institution (e.g., universities). The paper devotes a meaningful part of the introduction to how much a sport win is worth and the size of the intercollegiate athletics sector. Authors can support the size of an audience for studies in a specific context by multiple metrics, such as (1) contribution to GDP; (2) total revenues of all firms in the context; (3) total ad spend or marketing budgets; (4) impact on critical aspects of human life (e.g., healthcare or education); (5) share of spare time spent (e.g., video games); etc.

Even if the potential audience in a specific context is rather small, a study can pass the bar for publication of leading journals if the studied phenomenon promises to impact senior stakeholders in that context significantly. One could even see the two as compensating one another. The smaller the audience, the bigger and more profound the impact one desires of a study. A good example is the study of the influence of game supply on video game platform success. For a manager in the video game platform market, little is of more importance than game supply and what influences it, which is also the reason why firms in this industry spend many millions on deals with game providers (Binken & Stremersch, 2009). Thus, this phenomenon is typically boardroom material in such firms as it presents crucial decisions for such firms. Also, new methods very specific to a context may impact senior stakeholders if they are highly applicable to practical problems. A good example are studies that propose decision support systems to solve specific tasks such as those developed for the movie industry (Eliashberg et al., 2000; Eliashberg et al., 2001). For the healthcare industry, sometimes the relevance is related to the impact on policy-making as some of the findings may have direct impact on legislation and the well-being of society at large (another example is work on food consumption and nutrition).

Writing a context-specific study: Clarify the bounds of the context

Authors of context-specific studies need to clearly delineate the institutional details of the unique context and bound the context to guide readers on the extent to which findings can be generalized. This requires authors to clarify the bounds of the context in which the study is set and explain if similar contexts exist to which the findings in the context in which the study is set might generalize (which is not always clear a priori).

Authors can clarify the bounds of the context in several convincing ways. A first way is to contrast the specific context with another that has seen a lot of research attention before, to exemplify differences. For instance, long before the service-logic gained dominance (Vargo & Lusch, 2004), it was customary to contrast the specific context of services with the context of goods. A second way to clarify the bounds of context is to distinguish foundational concepts. For instance, John et al. (1999) called upon knowledge intensity as foundational to technology-intensive markets, and distinguished it from complexity and speed of change, which they termed “corollaries” of knowledge intensity. Later, Stremersch and Van Dyck (2009) distinguished know-how from know-why as foundational concepts to separately bound high-tech industries from life science industries within technology-intensive industries. A third way to clarify bounds is to list examples within and outside the boundaries of the context to clarify the scope of the chosen domain. For instance, Stremersch and Van Dyck (2009) (1) position life sciences in the bigger healthcare value chain; (2) discern boundary industries, such as cosmeceuticals, nutraceuticals, and medical devices and equipment; and (3) list specific applications of each.

Once foundational concepts of a specific context are well clarified, one can also easily understand the transferability of the findings from the study context to other contexts that share such foundational concepts. For instance, Guitart and Stremersch (2021) empirically study advertising, online search, and sales in the car market and clarify that the effects they find may be bound to high-involvement products. The latter identification of foundational concept allows transferability of findings to other markets, such as laptops, TVs, or speaker systems. A balanced (limited) generalization will need to acknowledge that even when authors propose the potential extension of their findings to other settings, this proposal is, above all, an avenue for further research.

Developing and writing context-specific studies: The case of life science marketing

Another way to exemplify best practices in developing and writing context-specific studies is to review one particular context-specific research area as a case study. We have chosen as research area “life sciences marketing,“ also sometimes framed as “pharmaceutical marketing” (which is a little narrower) or “healthcare marketing” (which is much broader). This literature originates predominantly in the ‘80s and ‘90s (e.g., Hahn et al., 1994; Mantrala et al., 1994; Parsons & Van den Abeele, 1981). The real takeoff of the life sciences marketing field probably took place in the period 2000–2015, which saw an increasing number of scholars enter this field of research on topics, such as: (1) life science therapy creation (e.g., Prabhu et al., 2005; Sorescu et al., 2003); (2) therapy launch, diffusion and sales growth (e.g., Shankar, 1997; Stremersch & Lemmens, 2009; Van den Bulte & Lilien, 2001); and (3) therapy promotion (Manchanda et al., 2004; Narayanan et al., 2004; Venkataraman & Stremersch, 2007).

How did authors in the life sciences marketing context expose the most critical elements described above? First, which are the arguments they used to explain its uniqueness? Second, how have they justified the importance of the life science marketing area? Third, which are some of the specific sources that have provided high-precision data?

Defining the uniqueness of the industry

Scholars have used several arguments to support the uniqueness of the life sciences marketing context. From a technical perspective, the life science industry has a very uniquely shaped new product development funnel in which only 1 out of 5,000–10,000 new inventions makes it to market (Ding & Eliashberg, 2002). Legally, market entry and product promotion are very strictly regulated by governments. Such regulations are unique to pharmaceutical markets, and the regulations vary across countries (Stremersch & Lemmens, 2009; Verniers et al., 2011). Examples include that certain marketing activities very commonly used in other industries are forbidden for pharmaceuticals in many markets (e.g., Direct-to-Consumer Advertising is only allowed in few countries (see Stremersch et al., 2013), or specific marketing activities may be capped to a certain level (e.g., physician detailing in some European markets)).

Life science firms also work in a unique channel context of payer-provider-patient. The payer, typically an insurance company or a government agency, pays the treatment prescribed and/or administered by a healthcare provider (e.g., doctor) to a specific patient (in marketing language the consumer). Sometimes a fourth player comes in, such as a friend, spouse, family member of the patient, as patients routinely seek social feedback. That channel structure gives rise to unique perspectives. For example, the prescriber may have special decision rules that regular consumers do not have (e.g., “first do no harm”), which may be even inflated by liability claim risks, influencing her decisions (as in Camacho et al., 2011). Or, therapy decisions that the doctor makes may not be accurately followed on by patients endangering health outcomes (Camacho et al., 2014).

The life sciences industry is also unique in some features that have allowed scholars greater environmental control in their studies. For instance, Narayanan & Manchanda (2009) aim to calibrate the heterogeneous learning on new product quality across consumers to then ground marketing allocation across consumers and over time thereon. They consider the pharmaceutical industry and the learning on new drug therapy of physicians as a uniquely suited context because: (1) true uncertainty exists about new drug therapy (also after clinical studies and approval); (2) firms spend large amounts on marketing; and (3) allocate such spending on the individual physician level. Another example is Kolsarici and Vakratsas (2010), who study category-level (i.e., only information about the disease can be communicated) and brand-level (i.e., only information about the brand can be communicated, void of any therapeutic-related information) advertising for a new pharmaceutical drug and argue that they can accurately assess the relative effectiveness of both types because regulations in the pharmaceutical market require both types to be mutually exclusive.

The importance of the life sciences marketing context

Scholars have used several arguments to support the importance of the life sciences marketing context. Prior literature routinely used the magnitude of (1) the industry to motivate its study; for instance, Stremersch and Van Dyck (2009) cite that the life sciences industry in the U.S. represented $271 Billion of sales in 2007; and (2) marketing spending as a way to substantiate the importance of the context, such as $4.3 Billion ad spend (Stremersch et al., 2013) or $18 Billion total marketing spending in 2005 (Montoya et al., 2010).

Further, scholars have emphasized the societal value of many topics that can be studied in the life sciences industry. For instance, studies on product compliance in drug markets, have typically emphasized the loss of life that results from poor compliance to drug therapy, estimated at 125,000 premature deaths in the U.S. each year (Wosinska, 2005).

Another reason why life sciences markets have been cited as important is through their impact on public or government costs. For instance, pharmaceutical drug pricing and generic substitution both have a large impact on public spending.

The specific sources in the development of life sciences marketing studies

While life sciences researchers have access to cross-industry data suppliers such as Kantar or Nielsen (e.g., data on DTCA spending at the individual drug level, at the campaign level) similar to more generalizable studies, they also use several unique sources that offer data opportunities not available in other markets.

  • IQVIA (previously IMS Health) provides data from panels of doctors on their prescription behavior (and the detail visits they receive, among others) (Kappe & Stremersch, 2016), as well as market-level data on monthly sales and promotional spending (broken down by different types of promotion) (Stremersch & Lemmens, 2009).

  • Co-development alliance and pipeline data: Clarivate Analytics Recap provides detailed information on alliances between biotech and pharma firms (Wuyts et al., 2004). Informa Pharmaintelligence provides data on pharmaceutical R&D pipelines. The FDA provides clinical trial information (Sood et al., 2014) and drug characteristics (Wuyts et al., 2004).

  • Individual company proprietary data: Several scholars have worked with dedicated data sets from one specific life sciences company (Iyengar et al., 2011, 2015).

  • International organizations such as the OECD and WHO: The OECD has data on regulations and health infrastructure on its member states (Stremersch & Lemmens, 2009). The WHO is an important source for its ATC classification, which divides drugs into therapeutic classes and thereby bounds the competitive landscape.

  • Governments or policymakers: Governments and public stakeholders monitor different aspects of the life sciences market. For example, Guo et al. (2021) use information reported to the Centers for Medicare and Medicaid Services (CMS)—part of the Department of Health of the government of the U.S.—to study the effect of information disclosure on industry payments to physicians.

Discussion

This paper aims to enrich the appreciation of marketing scholarship for context-specific studies. To do so, we have provided the main advantages of research in specific contexts and several considerations on how to improve the contribution of these studies. In this final section, we give practical recommendations on how to write context-specific studies. Additionally, we also discuss, beyond publication, other practical implications on how scientific work in specific contexts may also positively impact the career of a scholar. Next, we define how our claims can be generalized in two different directions (1) other ways of defining context-specificity beyond industry, and (2) other fields where the tension between context-specificity and generalizability is a relevant issue. We conclude with limitations and some directions for further research.

Practical guide to authors and reviewers: How to improve the contribution of context-specific studies

In Table 4, we present a checklist of the three key stages that authors developing studies in specific contexts should be especially aware of. This tool might also be useful for editors and reviewers to check whether the paper under evaluation delivers in the elements that are critical for a context-specific study. We believe the value of this checklist is that it helps identify further margin of improvement in each of the development stages of the paper.

Table 4 Checklist for research in specific contexts
Full size table

Stage 1: Define and elaborate on the uniqueness of the context

The first stage is to define and elaborate on the uniqueness of the context clearly. As pointed out above, this is done with the proper explanation of how the sources of specificity of the context create consequences of such specificity that are studied. Authors should strike a balance between providing too much detail–penalizing the paper on excessive length and potential loss of readership–and not providing enough explanation of the institutional and empirical setting–running the risk of missing critical parts of the phenomenon under study, not well defining bounds, or missing the intended purpose.

Stage 2: Justify the importance of the context

The second stage is to justify the importance of the context. A first way in which this can be achieved is to show that the context of the study is important in terms of economic size, societal implications, and the like. A second way is to justify why the context-specific study will resonate strongly among critical stakeholders of the specific context. Ideally, authors achieve both breadth and depth. On the one hand, they can show that the topic is relevant enough for somebody who is generally and genuinely interested in marketing. On the other hand, they can put the right emphasis on the aspects that might have a higher impact among the context-specific audience of the paper.

Stage 3: Integrate All in a Well-Written Paper

Authors also need to integrate it all into a well-written paper. First, do the authors show good understanding and immersion in the context? Unique contexts often have unique features that need to be theoretically described and well understood by the authors (e.g., proper description of context, beyond-textbook-examples, credible managerial implications). Second, does the study exploit the full advantages of the data showcasing that the authors are deeply immersed, rather than seemingly stumbled over a wonderful data set? Third, do the authors clearly describe boundary conditions and show the generalizability of the findings within the same context (“generalize across”)? Do they carefully elaborate on transferability to other contexts and limit generalizability (“(limits to) generalize to”)? A well-developed section of further research both within and outside the context ideally completes the research effort.

Context-specific research and the development of an academic career

Research on context-specific domains also has advantages for the development of an academic career. First, there is a growing concern within the field that the impact on practice is declining (e.g., Reibstein et al., 2009), and prior research recommends business school professors to engage more in practice-relevant work to improve the health of today’s business schools (Stremersch et al., 2021). Context-specific research is uniquely positioned to impact practice and fuel the careers of academic scholars that conduct this research.

Second, a deeper understanding of a specific context also aids in teaching, especially at more advanced levels. A core curriculum marketing course in an MBA program tends to look for more generalizable topics. However, when designing an MBA elective or an open-enrollment-focused program, the content tends to be context-specific. Context is even more critical when designing an in-company or custom program, where a professor needs to carefully choose topics and course materials of interest to a company, often accounting for its specific context.

Third, scholars also play an impactful role when disseminating knowledge: writing books, reports, and news articles addressed to a non-academic audience. Specialized reports on a specific context (industry, methodology, geography, population, etc.) might have a narrower audience but, potentially, a more engaged one. Thought leaders in specific domains can be invited as expert speakers to non-academic meetings or conferences. Furthermore, scholars with strong industry impact can lead a research center aimed at developing applied knowledge in a specific context.

Finally, some business professors directly advise companies as consultants or board members. This practice should be more actively encouraged by business schools if they wish to stay or become more relevant (Stremersch et al., 2021). Typically, these companies seek advice from well-rounded scholars but also with specific expert knowledge on certain industries.

Generalizing context-specific studies beyond industry and discipline

In this paper, we have used industry as an example of a specific context and applied it to the field of marketing. There are several operationalizations of context beyond industry one can easily think of: (1) specific organization types (e.g., start-ups vs. established companies, local vs. multinationals); (2) specific environmental circumstances (e.g., recession (Srinivasan et al., 2005)) or pandemics); (3) specific countries (where one can debate whether they are a holistic context or whether they can be broken down into a set of country characteristics).

Moreover, the trade-offs between generalizability and context-specificity are not exclusive to the field of marketing. This tension is equally relevant to other management fields such as operations management, competitive strategy, organizational behavior, or other social science fields, such as economics, sociology, or psychology. For example, in operations management, we see context-specific studies on travel, retail, healthcare, or the public sector.

Limitations and further research

In this study, we have conceptualized the advantages context-specific studies may offer and have exemplified how to do them well. This conceptualization offers many opportunities to empirically verify or falsify some of the claims we make and some of the proposals we advance. For example, could a measurement instrument to assess the magnitude of the contribution of context-specific studies be developed, advancing the checklist in Table 4 from a guideline to a practical quantification tool for authors and journals? An immediate application of such a measurement instrument would be to correlate scores of past articles with their impact on the field. It could also allow us to measure the frequency of occurrence of context-specific studies in marketing, which we have only proxied for.

Relatedly, it would be useful for future research to investigate boundary-setting in marketing research. It seems marketing journals publish a significant proportion of single-firm studies, often fueled by data availability or field experimentation. It would be meaningful to develop a practical guide on how to assess whether a study defines well the generalizability of its findings or, conversely, does it in a way that could be considered an overstatement or overgeneralization. However, at present, we lack tools and concepts to more strongly assess the context-specificity of findings and consistently bound study findings to the context the study is set in. Consequently, authors may easily overclaim the generalizability of their findings if not sufficiently disciplined by the journals. To us, marketing as a field may greatly benefit from a stronger bounding of context and a stricter discipline in generalizing.

At the same time, one may question–very much in line with our logic above–whether we should not rediscover case studies as a research method and find ways to do them well. Case-based research is a well-established research paradigm to develop theory (Eisenhardt, 1989), but it would be interesting to reflect on their value as a theory-testing tool. Many of the field studies or experiments today are actually quantitative case studies that report the results of marketing interventions as we find in other fields such as in medicine, rather than empirical generalizations. However, since we shy away from the word case study, these field studies and experiments by their very nature have the intrinsic drive to overclaim. Thus, should we as a discipline not just learn to be humbler and call studies what they actually are, whether it be context-specific or even cases, recognizing that these are beautiful in their own ways?