article notes

68

https://doi.org/

California Management Review 2017, Vol. 59(2) 68 –91 © The Regents of the University of California 2017 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/0008125617697943 journals.sagepub.com/home/cmr

Managing Customer Relations

Seeding, Referral, and Recommendation: Creating Profitable Word-of-Mouth PrograMs Michael Haenlein1 and Barak Libai2

SUMMARY In recent years, word-of-mouth (WOM) marketing has been the subject of considerable interest among managers and academics alike. However, there is very little common knowledge on what drives the value of WOM programs and how they should be designed to optimize value. Firms therefore frequently rely on relatively simple metrics to measure the success of their WOM marketing efforts and mainly use rules of thumb when making crucial program design decisions. This article proposes a new method to measure WOM program value that is based on the impact of WOM on the firm’s customer equity. It then provides recommendations for the five main questions managers face when planning a WOM program: Who to target? When to launch the program? Where to launch it? Which incentives to offer? and How many participants to include?

KeYwoRdS: marketing, social media, customer relations word-of-mouth, word-of- mouth programs, customer relationship management, customer lifetime value, social influence

I n recent years, the rising importance of word-of-mouth (WOM) programs as a marketing tool has become ever more apparent. On one hand, this development is driven by progress in online and mobile technology. New tools nowadays enable customers to be highly connected to one another

while providing marketers with previously unavailable means to study the cus- tomer’s social influence process and to implement incentives. On the other hand, there is rising evidence of the essential role of social influence in consumer deci- sion making, combined with empirical indications of the decreasing effective- ness of mass media advertising in last decades.1 Studies by firms such as Nielsen

1ESCP Europe, Paris, France 2Arison School of Business, Interdisciplinary Center, Herzliya, Israel

Seeding, Referral, and Recommendation: Creating Profitable Word-of-Mouth Programs 69

consistently show that WOM from friends and family is the single most trusted source of information for consumers,2 and a recent industry report suggests that WOM drives $6 trillion of consumer spending per year and plays an important role in the sales of many brands.3 Interestingly, these benefits cannot be attrib- uted to social media interactions alone, as the effect of offline WOM on brand sales is still estimated to be double that of online WOM.

Such insights have shifted the perception of WOM across industries from a “black box” that cannot be really governed to a phenomenon that should be pro- actively managed and amplified via planned programs. Start-ups are encouraged to take advantage of WOM as a relatively cheap marketing tool that reaches a large audience and is vital to long-term growth.4 Large advertisers realize that WOM plays an important role not only in driving sales but also in amplifying existing advertising.5 Academics have highlighted the ability to enhance profits by managing WOM, particularly in the context of introducing new products.6 Firms have created means to increase their understanding in this new world by estab- lishing associations such as the Word of Mouth Marketing Association (WOMMA). It is therefore not surprising that most senior executives believe WOM programs are more effective than “traditional marketing” and that spending on such efforts is going to grow substantially in the coming years.7

Yet there is also an ever more present confusion regarding WOM programs. Only a minority of executives believe they can effectively measure the return-on- investment of WOM-related activities and most view this issue as a major obstacle impeding greater use of WOM marketing in their companies. A recent survey among Chief Marketing Officers (CMOs) further points to a “social media spend- impact disconnect” by providing evidence that only a small minority of marketing managers feel they can quantitatively show the impact of social media activities.8 Indeed, many consider determining the value of WOM programs to be practically a “riddle.” As a short-term solution, marketing performance in the field of WOM marketing is frequently measured by social media–related activities (e.g., the quantity of communications regarding a brand, such as the number of “likes” that a brand receives or the number of tweets or conversations in which a product is mentioned). Yet, whether and how these activities translate into real business value that aligns with executive-level business goals, such as an increase in mar- ket capitalization or shareholder value, is largely unknown.9

Three reasons can be named for this confusion. First, structured WOM pro- grams are a relatively new phenomenon. From an academic perspective, this makes them different from tools such as advertising or loyalty programs, for which marketers can build on longitudinal research that helps to design effective strate- gies.10 From a practical viewpoint, it is reflected in the tools available to marketers. WOM tracking software programs are just making their way gradually to the mar- ket, yet it will probably still take time until they will be widely used.

Second, there is an inherent difficulty in assessing the profit created by social interactions among consumers, since the value created by the information

CALIFORNIA MANAGEMENT REVIEW 59(2)70

flow among multiple customers is non-linear and hard to predict. (We will elabo- rate on this point in more detail below.)

Third, research on WOM programs and their effectiveness is relatively new and scattered across different disciplines, which prevents the big picture from coming into focus. Prior studies have essentially focused on specific types of WOM programs (e.g., seeding programs, referral reward programs, business reference programs, viral marketing programs, and recommendation programs), and there is a lack of structured attempts to adopt a more general perspective on such pro- grams and to assess the value that they can create.

Given these challenges, it is essential to present a value-based view that can help in planning profitable WOM programs. While the measurement of WOM effects may be more complex than some other marketing phenomena, their basic aim is similar: to increase the overall long-term profitability of the customer base. Thus, in order to understand the value created by WOM programs, managers should rely on tools and ideas that were first proposed in the context of customer relationship management, namely, the concepts of customer lifetime value (CLV) and customer equity. Over the past 20 years, it has been shown both conceptu- ally11 and empirically12 that there is a strong relationship between customer equity and market capitalization, which should be considered by managers when mak- ing marketing-related decisions.13 Building on this logic, we propose a framework that helps managers understand value creation in WOM programs, and we pro- vide guidance regarding the five main questions managers face when planning a WOM program: Who to target? When to launch the program? Where to launch it? Which incentives to offer? and How many participants to include? In doing so, we integrate recent research on the effectiveness of social interactions and customer profitability, and we explain how these findings can be incorporated into a coher- ent view.

Types of WOM Programs

Given the large number of WOM programs that have been studied in literature, it is first important to extract their common essence through a gen- eral definition. For the purpose of this article, we define a WOM program as a marketing initiative that aims to trigger a WOM process by targeting a cer- tain number of individuals and incentivizing them to spread WOM. We refer to these individuals as program participants. Note that although we use the term WOM, which implies verbal communication, our framework also includes other ways in which individuals can exert social influence on one another, such as social media interactions or observations based on functional or normative influence.14 Such a broader conceptualization of the term WOM is consistent with recent academic and industry writings in this respect. Nevertheless, one should note that the measurement approaches and expected effectiveness of social influence may largely differ between different types of social influence mechanisms.

Seeding, Referral, and Recommendation: Creating Profitable Word-of-Mouth Programs 71

Within this general definition, we differentiate between three archetypes of WOM programs (see Table 1). The first type is a seeding program. The aim of product seeding is to get a (typically new) product into the hands of some indi- viduals, in the hope that this early social influence will help to accelerate and expand the growth process. The seeding approach can include discounts, samples, and even free products given to the seeds. Another form of seeding is viral mar- keting,15 which seeks to encourage the spread, by electronic means, of a message that the firm would like to promote (such as a video ad).

The second archetype is a referral program, in which current customers are encouraged to contribute to customer acquisition by bringing new customers to the firm. This group includes referral reward programs in business-to-consumer (B2C) settings and business reference programs, the equivalent of referral rewards in the business-to-business (B2B) sphere. One can also include affiliate marketing programs into this group, which provide incentives to independent website own- ers, or affiliates, who recommend the firm via online links in order to gain rewards.

Table 1. Major Types of WOM Programs.

Program archetype Program Form Description

Seeding programs Product seeding Accelerate the overall adoption of a wider group by getting a (typically new) product into the hands of a small group of people (the “seeds”)

Viral marketing Encourage a seed of individuals to share and spread a marketing message through electronic channels

Referral programs Referral reward Incentivize existing customers (mainly in B2C settings) to make product recommendations by providing rewards that depend on turning a referral into a sale

Business reference Use references from client firms in a B2B setting when trying to influence specific potential customers favorably to become new customers

Affiliate marketing Pay a monetary incentive (based on sales or clicks) for referring a person to a certain site via online links

Recommendation programs

Narrowband recommendations

Encourage recommendations through the social network of the specific individual (e.g., Facebook)

Broadband recommendations

Encourage recommendations through dedicated (review) sites (e.g., TripAdvisor, Amazon)

Note: A WOM program is a marketing initiative that aims to trigger a WOM process by targeting a certain number of individuals and incentivizing them to spread WOM. WOM = word-of-mouth. B2C = business-to- consumer; B2B = business-to-business.

CALIFORNIA MANAGEMENT REVIEW 59(2)72

The third archetype of WOM program, which occurs primarily in online environments, is a recommendation program. We observe two types of efforts in this regard. The first is the case of “narrowband recommendations,” in which indi- viduals recommend products to their personal social networks. The second is the case of “broadband recommendations,” in which the recommendation is posted on a designated recommendation site, run either by the firm itself or by a third party such as TripAdvisor.

Value Creation in WOM Programs

Conventional Measures

Three measures have traditionally been used by managers to assess the value of WOM programs:

• Quantity of communications: A main objective of WOM programs is to create interactions in the marketplace and to foster engagement,16 so firms fre- quently use the quantity of these interactions as a measure of WOM program success. These interactions can occur either offline, such as conversations in the context of WOM agent programs, or online, as in the case of social media posts and “likes.” Measuring communication volume is used extensively among managers since the quantity of interactions regarding a brand or prod- uct is generally easy to access, especially in online environments.

• Changes in brand equity: A second approach consists of assessing brand-related measures that are attributed to the WOM campaign, such as brand awareness or brand co-creation.17 This approach is consistent with the managerial per- ception that a central goal of WOM programs should be the creation of brand equity.

• Incremental sales: Managers often aim to use sales that follow a WOM program campaign as a measure of the program’s value.18 The question remains, how- ever, to what extent such sales can actually be attributed to the specific WOM program. Incremental sales can only be analyzed effectively in cases in which managers have the ability to carry out a before/after analysis, which allows them to compare sales with and without the campaign. This is usually lim- ited to specific situations, for example, WOM programs that are implemented in the absence of other marketing activities of the firm, programs in which referred customers use specific coupon redemption codes, or cases in which buyers can be tracked online to ensure that their purchases can be directly attributed to the WOM program.

The main shortcoming of these conventional measures is that they are unlikely to fully capture the actual value created. The fundamental role of mar- keting in the firm is ultimately to enhance the (discounted) profit stream, which stems from its customer relationships, that is, the lifetime value of its custom- ers.19 The main objective in this context is to maximize customer equity, defined

Seeding, Referral, and Recommendation: Creating Profitable Word-of-Mouth Programs 73

as the sum of the lifetime value of current and future customers, which is related to market capitalization and shareholder value.20 Following this logic, program success should be measured by analyzing the impact of a WOM program on cus- tomer equity.

Amplification of Customer Equity through WOM Programs

To understand how WOM Programs can influence customer equity, we build on the logic and findings of the customer relationship management litera- ture. Within this literature, three fundamental elements are commonly consid- ered to be the main sources of customer equity: customer acquisition (getting new customers), customer development (increasing profits from existing customers), and customer retention (keeping existing customers). These elements are closely related to the brand objectives mentioned previously since acquisition, develop- ment, and retention can be seen as consequences of the firm’s customer-based brand equity.21

• Acquisition: The vast majority of studies analyzing WOM programs have focused, in one way or another, on WOM effects on customer acquisition. Yet, what is often neglected is the fact that two different types of effects should be distinguished in this context: expansion and acceleration. Expansion refers to acquiring customers who would not have been acquired otherwise, either because they would not have adopted at all or because they would have adopted a competing brand. Acceleration refers to earlier acquisition of customers who would otherwise have adopted at a later point in time. Accel- eration translates into monetary gains due to the discounting factor since cash streams have a higher discounted value the sooner they are realized. Looking only at (incremental) sales that follow a WOM program does not provide a clear distinction between sales that represent acceleration and sales that rep- resent expansion. This lack of clarity can significantly bias the estimation of the total value of a WOM program.22

• Development: WOM programs can increase the profit of current consum- ers through mechanisms such as cross-selling, up-selling, or increasing their overall margin. It has been shown that customers acquired through WOM tend to be more satisfied, to engage in more cross-buying, and to generate higher contribution margins, at least at the beginning of their relationship with the firm.23 In this context, it should be noted that the line between development and acquisition is frequently not easy to draw. Convincing an existing customer to adopt a new product (e.g., via cross-selling) can be con- sidered as acquisition in some cases and as development in others. It there- fore seems likely that the additional consumption following a WOM program can be considered as customer development at least in some of the cases.

• Retention: Historically, few studies have explored the effect of WOM on cus- tomer retention. Still it has been shown that social influence can have a strong effect on defection decisions comparable in strength with the ones observed in cases of customer acquisition.24 Furthermore, customers acquired

CALIFORNIA MANAGEMENT REVIEW 59(2)74

through WOM programs have a higher retention rate than clients enter- ing the firm through other channels and the same applies to customers who actively participate in certain WOM programs post-acquisition, such as brand communities.25

The Value Created by WOM Programs

To exemplify how WOM programs can create value through customer acquisition, development, and retention, we use the example of a WOM pro- gram that influences the behavior of an individual program participant (partici- pant A). Figure 1 illustrates how such a WOM program can influence customer equity. We now discuss each element in Figure 1 in more detail:

• WOM program: The WOM program can be any of the ones listed above, that is, seeding, referral, or recommendation programs.

• Participant A and transmitter B: Participant A may be a customer of the firm or a non-customer who creates value by affecting others, even without making purchases herself. Participant A can create value by starting a social influence that creates or changes the lifetime value of other customers. This can either be done directly, through a WOM effect on people in A’s social network, or indirectly, by affecting another individual, transmitter B, who in turn affects the lifetime value of a third individual, customer C.

• Customer C: In both cases, participant A’s behavior will change the number of customers acquired by the firm and/or their CLV. This can be achieved by either impacting the time of acquisition, or through customer development, or by increasing the retention probability of customer C.

Figure 1. The chain of value from a WOM program to customer equity.

Note: WOM = word-of-mouth; CLV = customer lifetime value.

Seeding, Referral, and Recommendation: Creating Profitable Word-of-Mouth Programs 75

• Social value: The aggregated value of these social effects—that is, the change in customer equity that occurs as a result of participant A’s effects on the acqui- sition, development, or retention of other customers—is referred to as the social value of the program. Social value will be impacted by the effects on the CLV of each individual customer such as customer C and by the number of such customers affected.

• Direct value of participant A: While such social influence is probably the most common form of value creation, it is important to highlight that WOM pro- grams can create value even without social effects. This occurs when the CLV of participant A herself is enhanced by her participation in the program. Such enhancement may occur for two reasons. First, participation in the program may lead participant A to adopt the product earlier, which increases the net present value of her cash flows. Second, participation in the program might increase participant A’s attitudinal and behavioral loyalty. A recent study has, for example, found that defection rates of recommenders participating in a referral reward program fell from 19% to 7% within a year while their aver- age monthly revenue grew by 11%.26 We refer to the effect of a WOM pro- gram on the lifetime value of the participants themselves as the direct value of the program. Many approaches to evaluating WOM programs neglect to take into account such direct facts although doing so can lead to a (substantial) underestimation of the true value creation potential.

• Organic customer equity: Determining the value created by a WOM program is complicated by the fact that WOM can drive profitability regardless of any intervention by the firm. It is thus important to include only the incremental effect of the WOM program on customer equity. This makes the question how customers would have behaved in the absence of the program essential to the assessment of WOM program value since the absolute change in cus- tomer equity created by a WOM program needs to be benchmarked against the case in which the program does not exist. On the individual level, this means that we need to distinguish between activated WOM, which is directly triggered by the program (e.g., via an incentive), and non-activated WOM, which is not. On the social system level, this categorization is com- parable with a distinction between amplified WOM, which occurs in the presence of a WOM program, versus organic WOM, which occurs in the social system naturally.

• WOM program value: The WOM program social value is equivalent to the dif- ference between the amplified and organic customer equity. Calculating the incremental value generated by a WOM program is somewhat complicated by the fact that amplified WOM can start with activated WOM followed by non- activated WOM. Looking at Figure 1, for example, the firm can provide an incentive to participant A, who then spreads (activated) WOM to transmit- ter B. In response to this initial impulse, transmitter B may spread (non-acti- vated) WOM to customer C. Thus, it is necessary to understand the dynamics of both activated and non-activated WOM within the social network in order to be able to fully assess the value of amplified WOM. Adding the direct value

CALIFORNIA MANAGEMENT REVIEW 59(2)76

component and subtracting the WOM program cost from the WOM program social value result in an estimate of the WOM program value.

A Simple Measurement Approach

The methodology outlined above and illustrated in Figure 1 might, on first glance, seem complex—which could discourage firms from taking structured steps to implement it in order to assess the value of WOM programs. Therefore, we next present a straightforward four-step approach to WOM program value measurement that can help firms to assess whether a WOM program might be a good option and what its expected value could be. Based on the outcome of this simple approach, firms can subsequently build more detailed research mecha- nisms that will enable them to better explore the market. This approach will of course need to be adapted to any specific situation and might not fit all firms in all industries equally well.

For our illustration, we use the example of a WOM program targeted at customer acquisition, which is where most interest typically lies in the context of WOM. Nevertheless, our approach can easily be adapted to examine social value created by WOM programs aimed at customer development and customer reten- tion. In terms of data requirements, our focus is on determining the value created for a focal brand, and thus the information required should ideally stem from customers of this brand. Yet, if firms believe that the basic WOM dynamics (i.e., how often people talk and how they are influenced by WOM) are similar among the various brands within the same product category, it would be possible to replace this brand-specific information by category-specific data.

Step 1. Establish the importance of WOM for customer acquisition in the target market. To begin with, it is necessary to understand the relative importance of internal influence (i.e., WOM) versus external influence (e.g., advertising, public rela- tions) for customer acquisition in the target market. The relative importance of these two factors likely depends on industry and geography and may differ a lot from one target market to another. For example, research conducted in the mid-2000s surveyed consumers about their main information sources regarding firms in 23 categories and a variety of countries and found large discrepancies in information sources between categories. While the overall average percentage of consumers affected by WOM was 31%, the numbers reported varied between 9% (supermarkets in France) and 65% (coffee shops in the United Kingdom).27

Third-party information, such as syndicated WOM reports28 and methods of online monitoring,29 can be of help here, although a survey among customers acquired may be the most straightforward way of obtaining such information. When relying on surveys, firms should examine alternative ways of phrasing the WOM effect question (e.g., “How likely is it that you would have purchased this product/service without WOM?”)30 before deciding on the exact wording to be used, and the same applies to the type of measurement scale used.31 Given the complex “journeys” customers go through before making a decision, which

Seeding, Referral, and Recommendation: Creating Profitable Word-of-Mouth Programs 77

involve a multitude of channels and social media outlets,32 WOM influence can appear in various forms at different steps. Experimenting with different questions in this regard, and using interviews to validate the questionnaires, can help to achieve a reliable assessment on this important matter.

Very low and very high levels of WOM importance speak against the use of a WOM program. For some products (e.g., many low involvement supermarket goods), WOM may not play a sufficiently large role to warrant a WOM program and any WOM spread through the program is likely to have little effect on profit- ability. On the contrary, if WOM is very important, there may be limited potential for amplification and hence social value creation.33 A WOM program is therefore likely to be particularly effective at medium levels of WOM importance.

Step 2. Estimate the extent of organic conversations about the brand from current cus- tomers. The second step is to determine the degree of WOM by current custom- ers in the organic state. The same tools as before can be used in this context such as online monitoring, WOM reports, and customer surveys in which cur- rent customers are asked about the extent of talking about the brand in a recent period (e.g., “How often did you talk about this product/service within the past month?”).34 The appropriate time unit used is important and needs to balance that customers better recall recent events, while allowing for a sufficiently large period to capture incidences of social interactions.

Step 3. Assess organic WOM effect. Using information from steps 1 and 2, firms can easily create a “back-of-the-envelope” measure for organic WOM effect. Multiplying the total number of customers (e.g., 30 million) by the average amount of organic WOM spread per customer from step 2 (e.g., 0.8 conver- sations per 1,000 customers) results in the total number of WOM conversa- tions by current customers in the marketplace (e.g., 24,000). Combining this information with the number of customers estimated to have been largely affected in their acquisition through WOM (information obtained in Step 1; for example, 8,000 customers) allows firms to derive an assessment of organic WOM intensity (e.g., three conversations from a current customer to get one new customer).

Step 4. Assess expected WOM program social value. The last step consists of assessing which impact the WOM program is likely to have on the total number of conver- sations (e.g., a 10% increase from 24,000 to 26,400). Combining this assessment with the organic WOM conversion rate gives the amount of incremental cus- tomer acquisitions to be expected (e.g., 2,400/3 = 800). A value assessment can subsequently be obtained by multiplying the number of customer acquisitions by the average CLV per customer. One should, however, be careful not to overesti- mate the effect of the WOM program, as it would need to be adjusted by a decay factor that accounts for potential overlapping social networks.

This four-step approach can of course only give a very rough estimate of the social value to be expected from a WOM program. Nevertheless, many firms

CALIFORNIA MANAGEMENT REVIEW 59(2)78

can apply this method with reasonable effort, since the input data required are realistic to obtain. In addition, this approach better follows the social value cre- ation process compared with the more conventional measures mentioned previously.

WOM Program Design Decisions

After having analyzed how WOM programs can create value for firms, we now look into how to design profitable WOM programs that lead to maxi- mum value creation (see Figure 2). In this context, we focus on the five main questions managers face when planning a WOM program: Who to target? When to launch the program? Where to launch it? Which incentives to offer? and How many participants to include?

Who to Target: Opinion Leaders?

The question of who to target—and specifically whether participants who have disproportional influence on others (usually referred to as opinion leaders, influencers, or hubs) deserve particular attention—has created a lively debate among researchers. On one hand, there is considerable evidence in both market- ing and computer science that supports the essential role of opinion leaders in the spread of market information.35 On the other hand, there have been claims that marketers are wasting their money when they attempt to identify and influ- ence opinion leaders since the cascades of influence they create may not be that large. Instead of focusing on senders and their potential influence, it has been suggested that marketers should consider the nature of the receivers and to introduce programs in markets with populations that are highly susceptible to social influence.36 There has also been criticism about the ability of firms to iden- tify individual “mavens” who influence others in multiple areas, which creates the need to separately identify opinion leaders in specific categories.37

To complicate this discussion further, it can be argued that even in the absence of a WOM program, highly connected hubs may adopt a product earlier anyhow since they are subject to multiple social influences. This further reduces the effect of targeting hubs on customer equity, and the incremental value of such programs may therefore be smaller than what one might expect.38 Yet, con- trary to this logic there are indications that influencers may not necessarily be early adopters organically, but instead prefer to keep the status quo due to a desire of not being affected by others with lower status.39 If this is true, the impact of a WOM program targeting influencers could be even stronger than it might seem at first glance.

These different arguments show that there are numerous factors that need to be taken into account when assessing the value of approaching opinion leaders. A main reason for the ambiguity on this issue is the fact that most prior studies did not investigate the impact of opinion leader targeting on customer equity, but instead on more conventional measures of WOM program success. In recent years, the subject has been examined by academics in a more holistic manner. The

Seeding, Referral, and Recommendation: Creating Profitable Word-of-Mouth Programs 79

results of these studies are consistent in confirming the significant superiority of opinion leaders when taking a customer equity lens.40 This is well reflected in a rising emphasis on targeting influencers in various industries. For example, look at the fashion and beauty sector where 57% of marketers use influencers as part of their marketing strategy, with an additional 21% looking to introduce this type of activity in the near future.41 Furthermore, 26% spend at least 30% of the mar- keting budget on influencer marketing and a majority of marketers indicated expectations to increase this spending.

One additional point to note is the importance of distinguishing between two groups that are usually summarized under the label opinion leaders or influencers. The first group are mega-influencers that can often be found in online environments. This group includes well-known experts with many followers, popular blog writers, and celebrities. Unsurprisingly, aligning these individuals to the brand’s cause can be fruitful, although this effort may be expensive and not relevant for many firms. The good news is that opinion leaders can also fall into a second group called micro-influencers—everyday people that affect the (much smaller) social circle around them and are still of much interest to firms. The importance of this segment has been well demonstrated in a recent study

Figure 2. Recommendations for designing effective WOM programs.

Note: WOM = word-of-mouth; CLV = customer lifetime value.

CALIFORNIA MANAGEMENT REVIEW 59(2)80

conducted by the Keller Fay group, a prominent WOM and social influence marketing research firm, in collaboration with the Wharton School. This study found not only that micro-influencers have over 20 times more conversations than average consumers but also that 80% of people are very likely to follow their recommendations. Marketers therefore do not need to turn to celebrities to enhance their WOM programs. This leads us to the following conjecture:

Conjecture 1: For both mega-influencers and micro-influencers, WOM programs generate significantly higher value when they target opinion leaders compared with random customers.

Who to Target: Revenue Leaders?

One of the main limitations of targeting opinion leaders is the need to have information on the social network in general and on the social influence of the targeted individuals in particular. In the absence of such information, an alternative is to target participants on the basis of their revenue or expected CLV. High-value customers are likely to be connected to similar others, which makes them an attractive target not necessarily because they influence many other cus- tomers but because they influence the right customers.42 This phenomenon can be attributed to the well-observed phenomenon of assortative mixing, that is, the tendency of members of a network to attach to others who are similar in some way. In addition, such customers may exert a stronger-than-average influence. Heavy users may be more brand loyal and thus more willing to talk about the brands, which may lead others to perceive them as experts and to be more likely to be persuaded by the WOM that they distribute.43

Recent research has confirmed this intuition by showing that targeting rev- enue leaders is particularly attractive when introducing new products in indus- tries with high heterogeneity of CLV within the population and high assortativity (i.e., the correlation between the value of a consumer and that of one of his or her friends). This applies, for example, to sectors such as mobile phones, restaurants, and fashion items, which all have been found to show substantial values of assor- tativity. The comparison of different seeding strategies in the launch of new soft drinks, for example, shows that the best target group to choose in such a setting is people who do not know the product yet but have high value for the brand in general.44 This is aligned with other studies that show that targeting revenue lead- ers generates higher value than targeting random customers and sometimes even higher than targeting opinion leaders.45

Given these examples, the option to target revenue leaders should there- fore be attractive to firms. This is especially the case since managers usually have access to the data used in lifetime value modeling and can make use of established statistical techniques to help identify customers with high expected lifetime value. Yet, one point of caution should be taken. A study on seeding in the context of a mature restaurant chain found that brand loyal customers (which are likely to be revenue leaders) may not be the best targets to seed to. This might be the case since they have already affected their friends previously or since those friends

Seeding, Referral, and Recommendation: Creating Profitable Word-of-Mouth Programs 81

might be loyal customers themselves.46 Building on this finding, recent research that has examined various seeding campaigns in Europe suggests that the matu- rity of the market may play a dominant role in the social value of heavy users. Specifically, brand loyal customers may be better candidates for seeding in the context of an introduction of an additional new product, and less so when rein- forcing an existing one.47 Overall, we therefore come to the following assertion:

Conjecture 2: In new product markets, WOM programs generate signifi- cantly higher value when they target revenue leaders compared with ran- dom customers.

When to Launch the Program?

Once the decision on who to target has been taken, the next question is when to target those people. There are two main reasons why we expect a rela- tionship between the timing of a WOM program and its value. First, any poten- tial ripple effect created by individual program participants is likely to be larger when the potential market to adopt is larger. This happens to be the case in early phases of the product life cycle. If only a few people have adopted the product, it is easier for a program participant to influence many others who in turn influ- ence even more people. Second, early on, when there are fewer customers who can talk about the product, the contribution of any additional customer toward accelerating the product’s takeoff is likely to be larger. Increasing the number of people who can start a conversation from 1 to 2 is much more impactful than an increase from 101 to 102, since each incremental customer represents a larger percentage of the total base of adopters the fewer customers have already adopted.

Research in this area has validated this intuition in the context of custom- ers’ decision to “disadopt” (i.e., stop using) a new product.48 The negative conse- quences of disadoption of an innovator (i.e., early in the product life cycle), in the case of online banking, have been found to be more than twice the loss due to the disadoption of the average adopter. In fact, for earlier adopters who disadopt, the loss of social value can be considerably higher than the loss of CLV itself. Applying this logic to WOM programs (this time the value of a newly acquired customer rather than a lost customer), we can expect that a WOM program should have a larger effect on customer equity the earlier in the product life cycle it is launched.

By consequence, many firms use WOM programs today to enhance the launch of new products (in parallel to promotions), and industry reports suggest that the use of influencer programs is very important or critical in this context.49 Consider, for example, Philips Male Grooming and the launch of their new Aqua Touch razors in India. The WOM program, whose objective was to drive aware- ness about the skin problems that occur due to the use of razors, started with a series of videos featuring a razor named “Bladey,” which confessed to its crimes against proper care of skin and asked for forgiveness. These videos were uploaded on YouTube and then shared on Facebook, Twitter, and a webpage maintained by Philips. Consumers were subsequently asked to forgive Bladey by tweeting or

CALIFORNIA MANAGEMENT REVIEW 59(2)82

posting with the hashtag #I forgive on Twitter and Facebook. This step was inte- grated with offline promotions in which booths were installed in various shopping malls in which razors could be buried. In total, the “Bladey Confessions” channel on YouTube received roughly 750,000 views, and the Facebook community increased from 40,000 to 140,000 users during the campaign, representing a six- fold increase in engagement. The combination of highlighted personalized con- tent and a humorous interactive online fictitious character received several industry awards for Philips and the responsible agency Isobar. Overall this leads to the following conjecture:

Conjecture 3: WOM programs generate significantly higher value when they are launched earlier following the product launch.

Where to Launch the Program: Online or Offline?

WOM programs can be oriented either toward online or offline media, which raises the issue of where they should be launched with priority. In recent years, there has been a growing focus on online platforms, and in particular on social media,50 as a channel through which WOM is transferred. This is caused by the scope of these platforms and the speed at which information spreads through them. Numerous studies have shown that online platforms can have substantial effects on product success. However, despite the fact that online WOM programs seem more fashionable these days, it should not be forgotten that traditional offline WOM still plays a substantial role in customer decision making and may be potentially more influential than online WOM.51 Therefore, basing WOM program efforts exclusively on online media may miss much of the influence process. Both, online as well as offline components, should play a role in most WOM programs, although their relative importance might differ from case to case. This also implies that the effect of online and offline activities may be combined. A firm may, for example, conduct online seeding of a product but can expect that some or even much of the actual social influence will be offline.

An important factor to consider in this context is the tendency of individu- als to discuss different product types in certain media. For example, in an online environment, where people interact with large audiences with whom they often only have weak relationships,52 the issue of social status enhancement plays a key role. Generally, consumers prefer to talk online versus offline over premium brands and over products and brands that are more “interesting” and enable the person to enhance his or her social status.53 The question is therefore less whether one medium should be preferred in principle and more which medium better fits the specific product in question.

Fashion items and high-end cosmetics, for instance, fall into the category of products that people love to discuss online. On Facebook, luxury brands attract more than 4 times the fans and 20 times the “likes” of average consumer brands. The French cosmetics brand Guerlain, for example, who invented the first com- mercial lipstick in 1884, recently successfully used Instagram, a photo- and video- sharing site, to rejuvenate the brand image of its Terracotta bronzing powder.

Seeding, Referral, and Recommendation: Creating Profitable Word-of-Mouth Programs 83

Terracotta is one of the star products of Guerlain, introduced over 30 years ago, of which a product is sold every 20 seconds worldwide. Over the duration of four weeks, Guerlain created a photo campaign designed to showcase the link between Terracotta and Paris, which consisted of six pictures that showed landmarks such as the Eiffel Tower and Sacré-Cœur reflected in the mirror of a Terracotta product. In fewer than three weeks, the campaign reached almost two million people (of the five million active Instagram members in France) and generated over 185,000 likes and thousands of comments. In response, ad recall increased by 23 points and campaign awareness by 15 points (3.8 times Nielsen average). The imagery par- ticularly resonated with 13- to 17-year-old girls, who made up 29% of the audi- ence. This discussion leads to the following conjecture:

Conjecture 4: WOM programs generate significantly higher value when they have a stronger online than offline component if the underlying product allows enhancing social status.

Where to Launch the Program: Concentration or Spreading?

Regarding the offline component that most WOM programs should have, a second issue relates to the spread of influence and the question of whether the WOM program should be concentrated in a limited number of geographic areas or spread widely. On one hand, concentrating the program in one area may lead to increasing returns on additional users, due to threshold effects in adoption. This is especially the case since geographical location has been shown to have a strong impact on social influence, even for online products.54 On the other hand, a WOM program that targets a specific area is more likely to encounter over- lap among social influences, compared with a program that targets participants who are distributed across diverse regions. This trade-off makes the answer to the question whether concentration or spreading is preferable, not trivial. The overall picture that emerges from research in this area (e.g., research conducted on the spread of services such as Netgrocer.com) is that spreading is superior to clustering.55 Such spreading should not be too thin, however, since a “critical mass” of users in each area is necessary to ignite the process.

An additional point that needs to be considered in this context is the con- centration of customer profitability in certain geographic areas. Given the ten- dency of individuals to cluster near people with similar socio-economic characteristics, one can expect an uneven geographical dispersion of CLV in the market. This is not new for marketing managers, who for years have taken account of such factors—for example, when making decisions on where to locate new retail outlets. Targeting high-profitability areas is a common approach in marketing practice to acquire high-value customers. Similar to the logic of target- ing revenue leaders, such clustering of profitability should also have an effect on WOM program value. The fact that a given participant is in the vicinity of indi- viduals with potentially high CLV may increase her impact on customer equity and therefore WOM program value. The above discussion leads to the following conjecture:

CALIFORNIA MANAGEMENT REVIEW 59(2)84

Conjecture 5: WOM programs generate significantly higher value when they are spread geographically instead of concentrated; however, a strong dispersion in CLV among geographical areas may mitigate this effect.

Which Incentive Structure to Create?

Should firms offer an incentive to motivate participation in WOM pro- grams and, if yes, which incentives work best? For “mega-influencers” who make social influence their profession and expect to be monetarily incentivized, the issue may be straightforward. Yet the answer is more complex in cases where people affect their closer social circle, as introducing incentives into otherwise non-incentivized relationships can be a highly sensitive process. Incentives can affect the program participant’s willingness to spread WOM as well as the influ- enced individual’s tendency to act on it as both the sender and the receiver try to assess the possible motivations of the other side when deciding whether to dis- tribute or act on WOM, respectively.

To maximize the benefits of providing incentives, while mitigating receiv- ers’ potential concerns regarding senders’ sincerity, conventional wisdom is to consider rewarding both parties or to use in-kind rewards rather than monetary ones.56 Which strategy should be preferred depends on relationship norms and the strength of the relationship between the sender and the receiver.57 If incen- tives are paid, doing so should be disclosed and there is evidence that such disclo- sure may actually benefit the success of a program, since it supports the credibility of the message and the tendency of receivers to further discuss the message with others.58 Firms can learn, in this respect, from programs that use incentives to motivate employees to hire others with the potential to become successful employ- ees themselves—a process that is common in many industries.

To get some inspiration of how an incentivized WOM program can be designed, look at the ride-sourcing company Uber. In December 2011, Uber launched in Paris as its first non-U.S. city and today the service is available in over 50 countries. This impressive international expansion is partly driven by two smart referral programs: one focused on riders and one on drivers. For riders, Uber gives credits (which represent free rides) to both the referred and referring customer. Drivers, on the contrary, can earn up to $500 in cash for brining other drivers to Uber. The exact amount depends on the experience of the driver (the more experienced the new driver, the higher the reward) and his or her previous affiliation. For example, convincing a driver to switch from Lyft (a main competi- tor of Uber) to Uber results in higher rewards than bringing a virgin driver to the firm. There are also cross-over referrals since drivers can hand out cash credits to new customers who have not used Uber before. This allows drivers to print per- sonalized business cards, which, from a customer perspective, represent coupon codes for free rides.

An interesting question is whether paying incentives can lead to opportu- nistic behavior on the side of WOM program participants. If spreading WOM is something to earn money with, participants may prefer targeting receivers who

Seeding, Referral, and Recommendation: Creating Profitable Word-of-Mouth Programs 85

are easy to access (hence maximizing their incentives) versus receivers who are financially attractive for the firm. Recent research has examined this issue and shown that this is unlikely to be a problem by providing evidence that referred customers are actually worth more than non-referred ones.59 Combining all of this, we provide the following conjecture:

Conjecture 6: WOM programs generate significantly higher value when they make informed use of incentive structures, despite the sensitivity of intervention in the social influence process of WOM program participants.

How Many Participants to Include?

Determining the optimal size of a WOM program, which corresponds to answering the question how many program participants to target in a given popu- lation, is not trivial. While numerous studies have created algorithms to identify the best number of participants, the computational complexity of the problem makes it hard to reach a consistent solution. It is therefore not possible to give a one-size-fits-all “participation percentage” that works for all or even most WOM programs. While industry rules of thumb for (seeding) programs have been men- tioned to be around 1% of the potential market,60 academic research has used sizes as small as 0.2%61 to as large as 7% to 9%.62 What we know is that the optimal program size first and foremost depends on the structure of the social network. In a network in which people are densely connected (which leads to a significant overlap in circles of friendship), optimal seed size will be smaller than in networks in which this is not the case.

Identifying the extent of such overlap is, however, far from easy. Prior studies have, for example, looked into the degree of overlap between followers of different brands on Twitter. Such analyses are relatively easy to conduct and pro- vide interesting insights in terms of similarities between different brands (e.g., a quarter of Louis Vuitton fans also follow Burberry). Nevertheless, they can only serve as a very rough indication of the overlap that should be expected in the friendship circles of two WOM program participants. Prior research of a WOM program for a wine brand in Australia using Facebook friendship networks has shown that such overlap can be very substantial, leading to an overestimation of WOM program reach of nearly 60%.63

This makes it likely that the general assumption that a WOM program of twice the size also generates twice the success is unlikely to hold in most real-life settings. Instead, the existence of a “saturation effect” is one of the few findings regarding program size that has consistently emerged in most studies. The larger a program, the more likely it is that the social networks of individual participants overlap, which limits the incremental benefit that each additional participant can generate. There are therefore decreasing marginal returns of increasing partici- pant size. Research has suggested that this problem is particularly relevant to pro- grams targeting opinion leaders. In one study, the estimated decline in the contribution of an additional individual decreased by 43% for random customers

CALIFORNIA MANAGEMENT REVIEW 59(2)86

but 70% for opinion leader seeding when looking at a change in seeding percent- age from 0.5% to 3%.64 The overlap problem may be thus especially critical for opinion leaders. This discussion leads to the following conjecture:

Conjecture 7: The incremental effect of an additional WOM program participant on WOM program value declines with increasing WOM pro- gram size, and more so for programs targeting opinion leaders.

Further Insights

There are many other aspects of WOM programs that can be further con- sidered. Indeed, any factor that affects the social influence on individuals can be translated into insights on the effects of WOM programs. There are three exemplary areas—competition, target market, and WOM valence—that manag- ers should take account of in this context. Our general framework is sufficiently flexible to allow for the assessment of new strategic choices and new forms of social influence that may emerge as new technologies enter the market. The dif- ferential customer equity measure will still be the means by which the value of WOM programs should be assessed.

Competition

The question of competition is only rarely considered in the discussion of WOM programs, although it clearly plays a role in their success. Recent analysis in this area suggests that under competition, WOM programs generally create more value via market expansion (getting customers from the competition) than via acceleration (making a future customer adopt early), and that the interplay among the two can have a significant impact on the value created by the pro- gram.65 Understanding the source of value creation in such environments is thus vital for proper valuation.

An additional issue in this context is the brand-category relationship. The classic view of new product growth in marketing has been that the social influ- ence acts on the category level or cross-brand so that adopters of a certain brand can also affect the adoptions of other brands through WOM. This might, however, not hold true in all settings and there are cases where the social influence occurs only within brand.66 Understanding the within- or cross-brand effect in a certain market has notable implications for the planning of WOM programs and for the decision whether research should be conducted on the brand or the category level. At the minimum, firms should understand how they affect, and are affected by, competitors via the WOM programs.

Target Market (B2B vs. B2C)

The vast majority of analysis on WOM programs focuses on a B2C con- text, partly because this setting offers a more straightforward ability to identify and affect individuals in both offline and online environments. However, in a

Seeding, Referral, and Recommendation: Creating Profitable Word-of-Mouth Programs 87

B2B setting, a firm that initiates a WOM program may be able to take a more active role in the program’s progression. For example, the firm might select cer- tain customers as referrals, use the WOM program for quality signaling and stra- tegic pricing, and generate profit in the form of business reference value. This enhanced control over the WOM process can enable the firm to derive higher levels of value from the WOM program.67

WOM Valence

Naturally, WOM programs are formed to create positive WOM in the mar- ket. Yet, despite the consistent findings that positive WOM is more ubiquitous in most markets,68 there is empirical evidence that managers are much more concerned with suppressing negative WOM than they are with promoting posi- tive WOM.69 An interesting direction for managers would therefore be to use WOM programs to mitigate the undesirable consequences of negative WOM. For example, it has been suggested that in some markets, the existence of opinion leaders who oppose a certain innovation (“resistance leaders”) may significantly harm the growth of a new product, yet activation at the right time and place of other more positive adopters may mitigate this harm.70 Examining this from the customer equity framework implies that the organic WOM assumed should take into account the negative effects, while the amplified one is the one that includes the attempts to mitigate it. As before, the CLV of a customer should be taken into account, in particular as the distribution of CLV in the population might affect the effect of negative WOM on profits.71

Conclusion

It took marketers dozens of years to build a body of knowledge and meth- ods of assessment for established tools such as advertising and sales promotions. Our knowledge on WOM programs is much younger, and the rate of change of technology—and consequently of the tools used to design and implement WOM programs—is very high.

In the last decades, we witness a fundamental change in the marketing function. Technological changes—from databases to online and mobile technolo- gies—enable marketers to manage individual customers on a large-scale basis, creating measures that enable managers to tie marketing actions to the bottom line. In recent years, this revolution has been broadened by the inclusion of the importance of social influence direction. Marketers identify how customer profit- ability stems not only from their own lifetime value but also from their social value, that is, their effect on the lifetime value of other customers. As customers become more connected through social media and mobile tools, the management of the social part of their profitability becomes a pressing marketing priority.

We highlighted two main issues here. First is the need for measurement. As customer social influence management becomes an integral part of firms’ market- ing mix, marketers will be required to justify their investment in WOM programs,

CALIFORNIA MANAGEMENT REVIEW 59(2)88

as they do for any other tool. To this end, they will need to move from the lan- guage of conversations and impressions to that of lifetime value and customer equity, taking into account the benchmark social value created for their brands in the absence of a program. This will demand cross-function integration within the firm, where managers dealing with WOM and social media will need to move to become part of the customer management functions of the firm, cooperating with and learning from other customer management tools such as loyalty programs.

The second issue is the need to follow the emerging knowledge in this area. In order to develop marketing strategies for WOM programs, firms should understand the fundamental findings and drivers that are related to the main planning parameters of the programs: who to target, when to launch the pro- gram, where to launch it, which incentive structure to offer, and how many participants to include. Even more than other parts of the marketing function, this will require firms to follow and learn from academic research. Given the complexities and the non-linear effects of WOM, attempts to create generaliza- tions on how profit emerges from social influence are far from trivial and may become less based on managerial intuition. Yet managers, consultants, and research organizations should continue to monitor the emerging research stream on WOM and WOM marketing, examine the applicability of the findings to their specific case, and see how they can further use informed decision making to enhance customer equity.

Author Biographies

Michael Haenlein is a Professor of Marketing at the Paris campus of ESCP Europe, specialized in the fields of word-of-mouth, customer relationship management, social influence, and social media (email: [email protected]).

Barak Libai is a Professor of Marketing at the Arison School of Business, Interdisciplinary Center (IDC) Israel, and a recent co-author of Innovation Equity (The University of Chicago Press) (email: [email protected]).

Notes 1. Raj Sethuraman, Gerard J. Tellis, and Richard A. Briesch, “How Well Does Advertising Work?

Generalizations from Meta-analysis of Brand Advertising Elasticities,” Journal of Marketing Research, 48/3 (June 2011): 457-471.

2. Nielsen, “Word-of-Mouth Recommendations Remain the Most Credible,” Nielsen.com, July 10 2015, http://www.nielsen.com/id/en/press-room/2015/WORD-OF-MOUTH- RECOMMENDATIONS-REMAIN-THE-MOST-CREDIBLE.html.

3. Word of Mouth Marketing Association, Return on Word of Mouth, (Chicago, Word of Mouth Marketing Association, 2014).

4. Gabriel Weinberg and Justin Mares, Traction: How Any Startup Can Achieve Explosive Customer Growth (New York, NY: Portfolio, 2015).

5. Word of Mouth Marketing Association, op. cit. 6. Barak Libai, Eitan Muller, and Renana Peres, “Decomposing the Value of Word-of-Mouth

Seeding Programs: Acceleration versus Expansion,” Journal of Marketing Research, 50/2 (April 2013): 161-176.

7. “American Marketing Association 2013 Fact Book,” Marketing Insights, 25/4 (Winter 2013): 24-28.

Seeding, Referral, and Recommendation: Creating Profitable Word-of-Mouth Programs 89

8. The CMO Survey, “The Social Media Spend-Impact Disconnect,” 2016, https://cmosurvey. org/blog/the-social-media-spend-impact-disconnect/.

9. Walter Carl and Neil Beam, “Solving the ROI Riddle: Perspectives from Marketers on Measuring Word of Mouth Marketing,” Word of Mouth Marketing Association, 2012.

10. Barry Berman, “Developing an Effective Customer Loyalty Program,” California Management Review, 49/1 (Fall 2006): 123-148.

11. Rajendra K. Srivastava, Tasadduq A. Shervani, and Liam Fahey, “Market-Based Assets and Shareholder Value: A Framework for Analysis,” Journal of Marketing, 62/1 (January 1998): 2-18.

12. V. Kumar and Denish Shah, “Expanding the Role of Marketing: From Customer Equity to Market Capitalization,” Journal of Marketing, 73/6 (November 2009): 119-136.

13. Robert C. Blattberg and John Deighton, “Manage Marketing by the Customer Equity Test,” Harvard Business Review, 74/4 (July/August 1996): 136-144.

14. Ans Kolk, Hsin-Hsuan Meg Lee, and Willemijn Van Dolen, “A Fat Debate on Big Food? Unraveling Blogosphere Reactions,” California Management Review, 55/1 (Fall 2012): 47-73.

15. Andreas M. Kaplan and Michael Haenlein, “Two Hearts in Three-Quarter Time: How to Waltz the Social Media/Viral Marketing Dance,” Business Horizons, 54/3 (May/June 2011): 253-263.

16. Constance Elise Porter, Naveen Donthu, William H. MacElroy, and Donna Wydra, “How to Foster and Sustain Engagement in Virtual Communities,” California Management Review, 53/4 (Summer 2011): 80-110.

17. Nicholas Ind, Oriol Iglesias, and Majken Schultz, “Building Brands Together: Emergence and Outcomes of Co-creation,” California Management Review, 55/3 (Spring 2013): 5-26.

18. Robert V. Kozinets, Kristine de Valck, Andrea C. Wojnicki, and Sarah J. S. Wilner, “Networked Narratives: Understanding Word-of-Mouth Marketing in Online Communities,” Journal of Marketing, 74/2 (March 2010): 71-89.

19. Valarie A. Zeithaml, Roland T. Rust, and Katherine N. Lemon, “The Customer Pyramid: Creating and Serving Profitable Customers,” California Management Review, 43/4 (Summer 2001): 118-142.

20. Kumar and Shah, op. cit. 21. Florian Stahl, Mark Heitmann, Donald R. Lehmann, and Scott A. Neslin, “The Impact of

Brand Equity on Customer Acquisition, Retention, and Profit Margin,” Journal of Marketing, 76/4 (July 2012): 44-63.

22. Libai et al., op. cit. 23. Florian Wangenheim and Thomas Bayon, “Satisfaction, Loyalty and Word of Mouth within

the Customer Base of a Utility Provider: Differences between Stayers, Switchers and Referral Switchers,” Journal of Consumer Behaviour, 3/3 (March 2004): 211-220; Torsten Dierkes, Martin Bichler, and Ramayya Krishnan, “Estimating the Effect of Word of Mouth on Churn and Cross-Buying in the Mobile Phone Market with Markov Logic Networks,” Decision Support Systems, 51/3 (June 2011): 361-371; Philipp Schmitt, Bernd Skiera, and Christophe Van den Bulte, “Referral Programs and Customer Value,” Journal of Marketing, 75/1 (January 2011): 46-59.

24. Michael Haenlein, “Social Interactions in Customer Churn Decisions: The Impact of Relationship Directionality,” International Journal of Research in Marketing, 30/3 (September 2013): 236-248.

25. Julian Villanueva, Shijin Yoo, and Dominique M. Hanssens, “The Impact of Marketing- Induced Versus Word-of-Mouth Customer Acquisition on Customer Equity Growth,” Journal of Marketing Research, 45/1 (February 2008): 48-59.

26. Ina Garnefeld, Andreas Eggert, Sabrina V. Helm, and Stephen S. Tax, “Growing Existing Customers’ Revenue Streams through Customer Referral Programs,” Journal of Marketing, 77/4 (July 2013): 17-32.

27. Robert East, Kathy Hammond, Wendy Lomax, and Helen Robinson, “What Is the Effect of a Recommendation?” The Marketing Review, 5/2 (Summer 2005): 145-157.

28. Ed Keller and Brad Fay, The Face-to-Face Book: Why Real Relationships Rule in a Digital Marketplace (New York, NY: Free Press, 2012).

29. Mitchell J. Lovett, Renana Peres, and Ron Shachar, “On Brands and Word of Mouth,” Journal of Marketing Research, 50/4 (August 2013): 427-444.

30. For an example of such questions, see Robert East, Kathy Hammond, and Wendy Lomax, “Measuring the Impact of Positive and Negative Word of Mouth on Brand Purchase Probability,” International Journal of Research in Marketing, 25/3 (September 2008): 215-224.

CALIFORNIA MANAGEMENT REVIEW 59(2)90

31. Exemplary measurement scales to be considered in this context are the Likert scale (e.g., 1 = extremely unlikely, 2 = unlikely, 3 = neutral, 4 = likely, and 5 = extremely likely) or the Juster Purchase Probability Scale (0 = no chance, almost no chance; 1 = very slight possibility; 2 = slight possibility; 3 = some possibility; 4 = fair possibility; 5 = fairly good possibility; 6 = good possibility; 7 = probable; 8 = very probable; 9 = almost sure; and 10 = certain, practically certain).

32. David C. Edelman and Marc Singer, “Competing on Customer Journeys,” Harvard Business Review, 93/11 (November 2015): 88-100.

33. Eyal Biyalogorsky, Eitan Gerstner, and Barak Libai, “Customer Referral Management: Optimal Reward Programs,” Marketing Science, 20/1 (Winter 2001): 82-95.

34. See, for example, Robert East, Kathy Hammond, and Malcolm Wright, “The Relative Incidence of Positive and Negative Word of Mouth: A Multi-category Study,” International Journal of Research in Marketing, 24/2 (June 2007): 175-184.

35. Jacob Goldenberg, Sangman Han, Donald R. Lehmann, and Jae Weon Hong, “The Role of Hubs in the Adoption Process,” Journal of Marketing, 73/2 (March 2009): 1-13; Oliver Hinz, Bernd Skiera, Christian Barrot, and Jan U. Becker, “Seeding Strategies for Viral Marketing: An Empirical Comparison,” Journal of Marketing, 75/6 (November 2011): 55-71; David Easley and Jon Kleinberg, Networks, Crowds, and Markets: Reasoning about a Highly Connected World (New York, NY: Cambridge University Press, 2010).

36. Duncan J. Watts and Peter Sheridan Dodds, “Influentials, Networks, and Public Opinion Formation,” Journal of Consumer Research, 34/4 (December 2007): 441-58.

37. For a discussion of mavens versus opinion leaders, see Caroline Goodey and Robert East, “Testing the Market Maven Concept,” Journal of Marketing Management, 24/3-4 (April 2008): 265-282.

38. Goldenberg et al., op. cit. 39. Christophe Van den Bulte and Stefan Wuyts, Social Networks and Marketing, Relevant

Knowledge Series (Cambridge, MA: Marketing Science Institute, 2007). 40. Michael Haenlein and Barak Libai, “Targeting Revenue Leaders for a New Product,” Journal

of Marketing, 77/3 (May 2013): 65-80; Libai et al., op. cit.; Mohammad G. Nejad, Mehdi Amini, and Emin Babakus, “Success Factors in Product Seeding: The Role of Homophily,” Journal of Retailing, 91/1 (March 2015): 68-88.

41. Lucy Tesseras, “The Rise of Social Influencers,” Marketing Week, January 28, 2016: 26-27. 42. Michael Haenlein, “A Social Network Analysis of Customer-Level Revenue Distribution,”

Marketing Letters, 22/1 (2011): 15-29. 43. Raghuram Iyengar, Christophe Van den Bulte, and Thomas W. Valente, “Opinion Leadership

and Social Contagion in New Product Diffusion,” Marketing Science, 30/2 (March/April 2011): 195-212.

44. Florian Dost, Jens Sievert, and David Kassim, “Revisiting Firm-Created Word of Mouth: High-Value versus Low-Value Seed Selection,” International Journal of Research in Marketing, 33/1 (March 2016): 236-239.

45. Haenlein and Libai, op. cit. 46. David Godes and Dina Mayzlin, “Firm-Created Word-of-Mouth Communication: Evidence

from a Field Test,” Marketing Science, 28/4 (July/August 2009): 721-739. 47. Dost et al., op. cit. 48. John E. Hogan, Katherine N. Lemon, and Barak Libai, “What Is the True Value of a Lost

Customer?” Journal of Service Research, 5/3 (February 2003): 196-208. 49. Tesseras, op. cit. 50. Andreas M. Kaplan and Michael Haenlein, “Users of the World, Unite! The Challenges and

Opportunities of Social Media,” Business Horizons, 53/1 (January 2010): 59-68. 51. Andreas B. Eisingerich, Hae-Eun Helen Chun, Yeyi Liu, He (Michael) Jia, and Simon J. Bell,

“Why Recommend a Brand Face-to-Face but Not on Facebook? How Word-of-Mouth on Online Social Sites Differs from Traditional Word-of-Mouth,” Journal of Consumer Psychology, 25/1 (January 2015): 120-128; Keller and Fay, op. cit.

52. Michael Trusov, Anand V. Bodapati, and Randolph E. Bucklin, “Determining Influential Users in Internet Social Networks,” Journal of Marketing Research, 47/4 (August 2010): 643-658.

53. Lovett et al., op. cit.; Jonah Berger and Raghuram Iyengar, “Communication Channels and Word of Mouth: How the Medium Shapes the Message,” Journal of Consumer Research, 40/3 (October 2013): 567-579.

54. David R. Bell, Location Is (Still) Everything: The Surprising Influence of the Real World on How We Search, Shop, and Sell in the Virtual One (Boston, MA: New Harvest, 2014).

Seeding, Referral, and Recommendation: Creating Profitable Word-of-Mouth Programs 91

55. Ibid. 56. Peeter W. J. Verlegh, Gangseog Ryu, Mirjam A. Tuk, and Lawrence Feick, “Receiver

Responses to Rewarded Referrals: The Motive Inferences Framework,” Journal of the Academy of Marketing Science, 41/6 (November 2013): 669-682; Liyin Jin and Yunhui Huang, “When Giving Money Does Not Work: The Differential Effects of Monetary Versus In-Kind Rewards in Referral Reward Programs,” International Journal of Research in Marketing, 31/1 (March 2014): 107-116.

57. Gangseog Ryu and Lawrence Feick, “A Penny for Your Thoughts: Referral Reward Programs and Referral Likelihood,” Journal of Marketing, 71/1 (January 2007): 84-94.

58. Lisa J. Abendroth and James E. Heyman, “Honesty Is the Best Policy: The Effects of Disclosure in Word-of-Mouth Marketing,” Journal of Marketing Communications, 19/4 (September 2013): 245-257.

59. Schmitt et al., op. cit. 60. Emanuel Rosen, The Anatomy of Buzz Revisited: Real-Life Lessons in Word-of-Mouth Marketing

(New York, NY: Crown Business, 2009). 61. Sinan Aral, Lev Muchnik, and Arun Sundararajan, “Engineering Social Contagions: Optimal

Network Seeding in the Presence of Homophily,” Network Science, 1/2 (February 2013): 125-153.

62. Hinz et al., op. cit. 63. Lars Groeger and Francis Buttle, “Word-of-Mouth Marketing: Towards an Improved

Understanding of Multi-generational Campaign Reach,” European Journal of Marketing, 48/7-8 (2014): 1186-1208.

64. Haenlein and Libai, op. cit. 65. Libai et al., op. cit. 66. For a discussion of some implications of this issue, see Barak Libai, Eitan Muller, and

Renana Peres, “The Role of Within-Brand and Cross-Brand Communications in Competitive Growth,” Journal of Marketing, 73/3 (May 2009): 19-34.

67. For a discussion of word-of-mouth (WOM) programs in a business-to-business (B2B) context, see, for example, V. Kumar, J. Andrew Petersen, and Robert P. Leone, “Defining, Measuring, and Managing Business Reference Value,” Journal of Marketing, 77/1 (January 2013): 68-86; Mahima Hada, Rajdeep Grewal, and Gary L. Lilien, “Supplier-Selected Referrals,” Journal of Marketing, 78/2 (March 2014): 34-51.

68. See, for example, East et al., op. cit. 69. Martin Williams and Francis Buttle, “Managing Negative Word-of-Mouth: An Exploratory

Study,” Journal of Marketing Management, 30/13-14 (2014): 1423-1447. 70. Sarit Moldovan and Jacob Goldenberg, “Cellular Automata Modeling of Resistance to

Innovations: Effects and Solutions,” Technological Forecasting and Social Change, 71/5 (June 2004): 425-442.

71. Mohammad G. Nejad, Mehdi Amini, and Daniel L. Sherrell, “The Profit Impact of Revenue Heterogeneity and Assortativity in the Presence of Negative Word-of-Mouth,” International Journal of Research in Marketing, 33/3 (September 2016): 656-673.

Copyright of California Management Review is the property of California Management Review and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.