How Social-Cause Marketing Affects Consumer Perceptions

How should companies determine the best way to allocate marketing dollars between conventional promotional programs and affinity marketing programs? The former simply stress the benefits of buying a specific brand, while the latter prominently and publicly identify a company’s association with a particular sport, entertainment event, nonprofit organization or social cause. Experiments we have conducted suggest that the research method known as conjoint analysis could be a valuable market research tool to help companies predict which of several alternative affinity marketing affiliations would provide the best return on investment. Furthermore, based on both theory and our initial findings from a set of studies using conjoint analysis, many companies will obtain better returns through creating an affinity with a social cause than through affiliating with other, more clearly commercial ventures.

The Potential Returns From Affinity Marketing

According to an article in the IEG Sponsorship Report, spending on sponsorships in North America during 2005 was expected to reach $12.1 billion, with 69% going to sports, 10% to entertainment tours and attractions, 5% to the arts, 4% to festivals, fairs and annual events, 3% to associations and membership organizations and 9%, or about $1.15 billion, to social causes.1 Elsewhere, another article claimed that $9 billion was spent by corporations in 2001 on all types of social initiatives, including strategic philanthropy.2

Clearly, companies are making substantial investments to try to demonstrate an affinity with consumers interested in sports teams, entertainment events and social causes. While a company at one time might have sponsored or supported an activity simply because an executive wanted to help a favorite team or cause, companies today are increasingly treating investments in affinity marketing as important strategic moves. Affinity marketing programs are designed to achieve objectives such as improving overall corporate reputation, differentiating a brand, attracting the interest of targeted consumers, stimulating brand preference and loyalty, attracting loyal employees and, ultimately, increasing profits and stock prices.

Whether companies are achieving these objectives is only beginning to be understood. There is evidence that sports and entertainment sponsorships can be successful, but the available empirical research on this topic is limited.3 Considerably more research has been done on the effects of societal marketing programs, which emphasize a brand’s affinity with a social cause. (We define societal marketing programs as company initiatives involving the provision of money, resources and/or publicity to socially beneficial causes in a way that seeks to create an association in the minds of consumers between the cause and the company or one of its brands.)

The research on societal marketing programs has examined a range of initiatives, including cause-related marketing (in which every unit sold of a brand triggers a donation to a cause), green marketing (in which the environmental friendliness of a company or brand is stressed as a differentiating attribute), cause sponsorship (in which a brand is clearly identified as a cause supporter) and social advertising (in which a cause is promoted in a brand’s ads). Much of this research consists of case studies on companies that apparently benefited from pursuing societal marketing initiatives. Case histories of societal marketing initiatives at companies such as Timberland, Avon, Stonyfield Farm, ConAgra, Home Depot, Starbucks, American Express and Tom’s of Maine have been recounted frequently in textbooks and in the business press.4

The case histories suggest that these companies have been able to use societal marketing to help differentiate their brands from competitors in consumers’ minds, in turn leading to a range of desirable effects, including greater efficiency for other marketing efforts, an ability to charge higher prices, increased market share, greater brand loyalty and more favorable treatment from stakeholders such as regulators and investors. At the same time, societal marketing activities are supposed to have helped many of these firms achieve better scores on ratings of the most admired and socially responsible companies. This presumably has helped them become more efficient in recruiting, retaining and engaging talented employees, since people may have more positive feelings about working for a socially responsible company.5

Despite the results reported in many case studies, spending on societal marketing initiatives is dwarfed by spending on conventional marketing and other affinity marketing involving sports and entertainment. Many managers apparently have yet to be convinced that societal marketing programs offer the potential for higher returns on investment. The case studies seem not to have persuaded them, and neither have studies by consulting firms such as the 2002 Cone Corporate Citizenship Study, which reported that 84% of Americans agreed that they “would be likely to switch brands to one associated with a good cause, if price and quality are similar.”6 Later Cone studies showed similar results. Unfortunately, these types of opinion surveys only ask consumers abstract and hypothetical questions. As a result, consumers’ answers could be influenced by a desire to give responses that are perceived as socially desirable.

Managers are more likely to find persuasive evidence about the benefits of societal marketing if they examine the growing body of experimental studies reported in the academic literature. These involve controlled laboratory studies that ask consumers to evaluate brands that have an affiliation with a social cause versus those that have no such affiliation. The experimental results show that consumers will respond with more favorable ratings and a higher likelihood of choice to brands that have certain social-cause affiliations.7 Feelings of affinity or identification with the social cause often have been found as the drivers of these ratings and choices.8 Although these academic studies have not been conducted in real-world field settings where actual consumer behavior could be observed, the use of careful experimental controls provides some reassurance that the differences in ratings or choices were caused by the presence or absence of the social-cause affiliation.

What has been lacking from case histories and previous academic research has been a comparison of the beneficial effects of a social-cause affiliation with the effects of other types of affiliations. Most companies are faced with a choice of spending some of their marketing budget on either a social cause or a more commercial promotional venture. Comparing the performance of a societal marketing initiative to that of no initiative at all, which has been done in most of the previous experimental studies, does not provide very compelling evidence to managers about the value of societal marketing. Managers need to know whether they get a better payoff from spending marketing dollars affiliated with a social cause (such as sponsoring the Susan G. Komen Breast Cancer Foundation Race for the Cure Series) than from affiliating with a clearly commercial cause (such as sponsoring a rock concert tour). After all, consumers’ affinity for rock singer Bruce Springsteen could be greater and more influential in driving brand choices than their affinity for the cause of fighting breast cancer.

Understanding the Effect on Consumers

Exposure to any type of well-conceived promotional initiative for a brand leads, in theory, to more positive feelings and judgments about the brand in a consumer’s mind. We suspect that a promotional initiative emphasizing a brand’s affiliation with a team, event or social cause for which a consumer has a high degree of affinity is likely to have a more positive effect on that consumer’s judgments and feelings about the brand than similar initiatives that either emphasize no such affiliations or emphasize affiliations for which that consumer has little affinity. For example, an avid college basketball fan will probably respond more positively to messages about a sneaker brand that outfits the fan’s favorite college team than he or she would to messages about a sneaker brand that did not mention any affiliations or that highlighted its sponsorship of something the fan did not care about, such as a marathon.

We hypothesize that the primary reason why demonstrating a high degree of affinity can enhance the effectiveness of a promotional initiative is that it increases the likelihood that consumers will treat the initiative itself — or, more generally, the brand’s “style of marketing” — as an important and positively weighted attribute of the brand. However, when a promotional initiative does not mention an affiliation or when the affiliation is one that consumers view as run-of-the-mill or deserving of skepticism, consumers may weight a brand’s style of marketing as an unimportant or possibly even negative attribute. Substantial segments of consumers may see sports or entertainment affiliations as too “commercial” and opportunistic; these consumers would not weight the style of marketing as positively with these types of affiliations as they would if the affiliation was with a cause like cancer research, environmental protection or disaster relief. Moreover, a social-cause affiliation could also have a “halo effect” on how a brand is seen on other attributes, such as trustworthiness or quality.

We suspect that, in general, affinity marketing programs that have a high degree of fit with a brand — that is, those initiatives in which the logic behind the brand’s affiliation can be easily recognized by most consumers — typically produce a more positive effect on consumer brand judgments and feelings than initiatives with weaker fit. Hence, a bicycle company sponsoring the Tour de France might produce more positive responses than a bicycle company sponsoring a motor sport rally. Previous research documenting the benefits of “fit” in launching brand extensions is consistent with this idea.9

The interactive effects between degree of affinity and degree of fit, however, are less obvious. In particular, it is unclear whether a promotional initiative that highlights affiliation with a social cause will be more or less effective when there is a high versus low degree of fit between the brand and the cause. Although a few recent studies suggest that a social cause with high brand fit (for example, a milk company providing milk to poor children) will improve consumers’ judgments and feelings about the brand more than a social cause with low brand fit (for example, a milk company promoting responsible alcohol consumption),10 it is possible that the opposite could occur. When a brand promotes a high-fit social cause, it may be more likely to be viewed by some consumers as opportunistic and seeking commercial gain. It is possible that consumers’ skepticism about such an effort could lead to their reacting negatively to the high-fit societal marketing initiative. The lack of surprise or uniqueness in high-fit societal marketing initiatives might dampen effectiveness among other consumers, while aligning with a low-fit social cause actually might help to differentiate a brand as being more sincere in its generosity, thus leading consumers to react more strongly and positively to the promotional initiative.11 Furthermore, some causes are already supported by so many companies that supporting them does not lend a clear distinction to a brand.

Comparing Affinity Marketing Options

To gain empirical insight into how consumers might respond to different affinity marketing initiatives, we used the research technique known as conjoint analysis. Conjoint analysis has become a staple in the marketing research field for helping companies predict the most effective configuration of attributes, such as warranty length, price or product size, for new products or services. To the best of our knowledge, however, conjoint analysis has not been used for pretesting societal marketing initiatives or other affinity marketing programs.

This technique asks consumers to review a set of profiles, each consisting of a combination of various hypothetical attributes of a particular product or service; each consumer ranks the different profiles according to his or her preferences. The profiles are varied systematically so that a range of attribute combinations are considered, with some profiles containing high levels of certain attributes, and some containing low levels of those attributes. Based on how a consumer shows preference for the profiles, statistical techniques can be used to determine which attributes the consumer weights most strongly positively, most strongly negatively and in between.12

Importantly, conjoint analysis can help marketers estimate the effects of specific promotional programs using a small subset of target consumers prior to making a financial commitment to a particular program. Unlike direct survey questions, where a desire to give socially desirable responses may lead to posturing by consumers, conjoint analysis has consumers reveal their preferences for certain attributes indirectly, by making choices and tradeoffs among realistic options.

To illustrate how conjoint analysis might be used for these purposes, consider the following hypothetical problem that might be faced by a brand manager for a U.S. beer brand. Assume that the beer’s brand manager is considering four options for affinity marketing initiatives in the next budget period, and the options vary in terms of the type of sponsorship and the degree of fit with the brand: a high-fit social cause (a designated driver program), a low-fit social cause (a program to help children read), a high-fit sports/entertainment association (a stock car racing team) or a low-fit sports/entertainment association (a network Sunday night movie).

Instead of just asking consumers to rate these four options, the brand manager could obtain more reliable predictions of consumer behavior by setting up an experimental study using conjoint analysis. In a study we recently conducted, a sample of 135 beer consumers were randomly assigned to one of four parts of the experiment, with each part of the experiment testing reactions to one of the four potential affinity marketing programs described above. Random assignment was used to ensure that the participants in each section of the experiment were comparable. (See “The Impact of Various Affinity Marketing Programs.”)

In each of the four sections of the experiment, there were eight profiles describing various possible attributes of the beer brand. Half the profiles described the brand as sponsoring the cause or event being studied in that section of the experiment, and half stated that the brand was host of an amusement park. In that way, reactions to two variations of a “style of marketing” attribute (that is, sponsor or host) were obtained. In addition, each profile had one of two variations of four other attributes: half packaged in bottles and half in cans; half at a higher price per six-pack and half at a lower price per six-pack; half with 100 calories per serving and half with 150 calories per serving; and half fortified with 1.5 grams of protein per serving and half with 3 grams of protein per serving. The resulting brand profiles were put on eight cards that were set up so that each variation of an attribute appeared at least twice with each of the variations of the other four attributes.

Each study participant ranked the profiles according to preference. The ranking process is a way for consumers to reveal the importance to them of different attributes. Thus, if the four best-ranked profiles for a consumer all were packaged in bottles, it would be clear that bottles are a very important, positively weighted attribute in evaluating beers for that consumer. Indeed, by running a regression analysis using data about the consumer’s rankings and the brand attributes, it is possible to see how much influence each attribute has on the way that consumer determines rankings.

In particular, the regression analysis reveals how much a consumer’s ranking of a profile would change if the beer brand sponsored the event or cause being evaluated, rather than just hosting an amusement park. The coefficient that measures this can be considered the “weight” or the “part-worth” that the participant gave to the beer brand’s style of marketing. Weights or part-worths for the brand’s other attributes, such as price, were also calculated. Assuming equality in everything else across the four sections of the experiment, the study was able to isolate how much the type of affinity marketing program and its fit with the brand affected the weights consumers gave to the beer’s marketing approach, as well as to its other attributes.

Participants found it very easy to do these ranking exercises and provided provocative information. For example, on average, the two groups randomly assigned to the social-cause partners revealed significantly more positive part-worths for marketing style than the two groups randomly assigned to the sports/entertainment partners. The numbers can be interpreted as meaning that having the beer brand sponsor a designated driver program or a program to help children read would improve how consumers rank a profile by an average of about 10% of a ranking point on the eight-point ranking scale over how they would rank a profile where the brand only hosted an amusement park. Although this may not seem like a substantial impact, another way of looking at this is to compare how much the price would have to change to get 10% of a ranking point improvement. Based on the part-worths assigned to price by the participants in this study, achieving a 10% ranking point improvement would require roughly a 50 cent cut in price. This does not mean that the sponsorship would allow the beer company to charge 50 cents per six-pack more than if it only hosted an amusement park, or that consumers are willing to pay such a premium for a beer that sponsors social causes. It only means that the societal marketing affiliation would have an impact similar to a price cut of that amount.

The study of this beer brand also found that, overall, the high-fit sponsorships did not improve ranking scores significantly better than the low-fit ones. Although on average the social-cause partner with the low fit (that is, the program to help children read) improved ranking scores more than the social-cause partner with a high fit (that is, the designated driver program), the difference was not statistically significant. Similarly, the reduction in ranking scores caused by the sports/entertainment partner with the low fit (that is, the Sunday night movie on network TV) compared to the reduction produced by the sports/entertainment partner with a high fit (that is, stock car racing) was not statistically significant. These results confirm our previously stated uncertainty about the interactive effects of affinity program and fit.

In addition to examining various affinity marketing partners for the beer brand, we conducted other studies looking at potential affinity partners for other well-known U.S. and Mexican brands of beer, juice and milk, using both U.S. and Mexican study participants. (See “About the Research.”) In studies of three different beer brands, we found a similar pattern, with social-cause partners such as designated driver programs, programs providing milk to poor children, recycling programs and cornea research programs viewed by consumers as making the style of marketing of the brand a more positively weighted feature. Also, for the beer brands, the social-cause affiliations produced more positive weights than the various hypothetical sports/entertainment affiliations, with the exception of an affiliation with the hypothetical organization World Cup Soccer, which produced positive weights comparable to those of the social-cause affiliations. (A World Cup Soccer affiliation also had a similar effect for a well-known Mexican juice brand, indicating that demonstrating affinity with this activity can be quite effective.) The results comparing affinity partners for the milk brands were less conclusive, suggesting that societal marketing may have very different effects in different industries.

Evaluating Societal Marketing Initiatives

We have made the case that affinity marketing initiatives, especially societal marketing initiatives, have the potential to improve consumers’ attitudes about a brand in a number of different ways. How much a given initiative will help or hurt a given brand will, of course, depend on the characteristics of its target markets. Hence, brand managers need to address questions such as the following as they evaluate what degree of commitment to make to affinity marketing:

• Are there a sufficient number of consumers in the brand’s desired target market who have a strong affinity for the sport, event or social cause under consideration for affiliation?
• Will consumers from desired target markets find it credible that this brand is affiliated with this sport, event or social cause, or will they view such support with suspicion?
• Does the brand differentiate itself from its competitors in the eyes of desired target markets through supporting this affiliate, or does the brand look like a copycat?
• How does the affiliate stack up versus other potential beneficiaries of the brand’s promotional initiatives, in terms of affecting target consumers’ view of the brand’s style of marketing and its image and performance attributes?

Negative answers to any of the first three questions suggest that support of the affiliate in question should not be an integral part of a brand’s promotional strategy. Without a critical mass of consumers possessing an affinity for the affiliate, without credibility in promoting the interests of the affiliate or without gaining differentiation from the affiliation, the brand would be better off using its promotional dollars in other ways, such as simply promoting a brand’s best and differentiating benefits.

Conventional survey research with samples of target consumers could help to find answers to the first three questions above. But more sophisticated research, including studies using conjoint analysis, will be needed to address the fourth question. Our experimental results have persuaded us that it would be very feasible for real-world brand managers to test affinity partner ideas using a conjoint analysis. In designing conjoint analysis studies, however, managers should pay particular attention to several issues.

Have potential customers from highly valued target markets participate in the studies.

The affinity preferences of the general public, while interesting, are much less important to determine than the affinity preferences of consumers who are prime prospects for becoming long-term, heavy-using customers of the brand.

Use random assignment to determine the affinity partners that each study participant will evaluate.

This is the best way to ensure that the groups are similar and that differences in the responses to the partners are caused by partner differences rather than differences across the evaluating customer groups.

Make sure that other attributes evaluated in the profiles do not vary in different parts of the study.

Keeping the nature and levels of attributes (such as prices or nutrients) the same in multiple studies done over time also is advisable. If these characteristics are not kept constant, they could affect how participants do ranking. For example, if one study had a small price gap in its profiles while another had a large price gap, participants would weight price more heavily in the latter study, potentially leading them to weight the brand’s style of marketing less heavily.

Find realistic promotional options to test that are, at a minimum, low on fit, not very poor on fit.

As mentioned earlier, low-fit initiatives may actually be responded to more positively by consumers under certain conditions. But initiatives with very poor fit are unlikely to do anything but hurt a brand and could distort results in conjoint studies, just as having wide gaps in price levels could.

Select a neutral alternative to the affinity partner to use as a comparison in the brand profiles.

In our first study,“host of an amusement park” was used as a comparison to having a social-cause or sports or entertainment partner. Having an alternative that is viewed too positively could make it hard to detect the effects of variation in the style of marketing attribute.

In determining whether to pursue affinity marketing in general, or societal marketing in particular, managers should recognize that every brand is different. While, for example, a low-fit societal marketing initiative might work best for one brand, it might do very little for — or even harm — another brand’s image. We recommend doing careful experimental research, using conjoint analysis to refine and test ideas for affinity marketing initiatives against one another and against other kinds of marketing initiatives. A cautious, research-based approach seems appropriate, given our limited knowledge about the true effects of these programs. However, the results of our initial studies suggest that companies may be underestimating the potential bottom-line benefits of societal marketing initiatives. Societal marketing may give a brand just the edge it needs to win the hearts and minds of the inundated, skeptical consumers populating today’s cluttered, supercompetitive marketplaces.