12

Click here to load reader

A Holistic Approach to the Measurement of WOM

Embed Size (px)

DESCRIPTION

We present the results of a holistic approach to WOM research, which finds that 90% of word of mouth takes place offline, is primarily positive, is often influenced by advertising, and takes place disproportionately among influencers. We also present an approach to assess the social monetary value of customers, integrating the TalkTrack® data into agent-based simulations. We show that while random seeding of customers is effective, focusing on opinion leaders can increase profitability substantially.

Citation preview

Page 1: A Holistic Approach to the Measurement of WOM

1Copyright © ESOMAR 2009

WORLDWIDE MULTI MEDIA MEASUREMENT 2009

pART 5 / THE pOWER OF SOCIAL MEDIA

A HOLISTIC AppROACH TO THE MEASUREMENT OF WOMIts Impact on consumer’s decIsIons

Ed Keller Barak Libai

INTRODUCTION

numerous research studies conducted over the past few years show that word of mouth is the most important consumer touch point when it comes to decisions about products, services, and brands. media agency Zenith optimedia, for example, released research results in 2008 that concluded, “recommendations from family and friends trump all other consumer touch points when it comes to influencing purchases,” according to a report on the research in the april 9, 2008, issue of Advertising Age.

With the recognition of the power of word of mouth, a growing number of brand marketers – and the agencies that serve them – are investing in marketing approaches designed to stimulate word of mouth advocacy. as a result, word of mouth marketing has emerged as one of the fastest growing media sectors. according to one estimate by media forecasters pQ media (pQ media, 2008), u.s. spending on consumer-generated media is growing faster than any other type of alternative media segment (out of 18 measured).

a wide variety of strategies to encourage word of mouth are being employed. many marketers are increasingly focused on the opportunities afforded by digital media to facilitate the people’s growing desire to express themselves (e.g., social media platforms, online ratings and review sites, etc). others are employing offline word of mouth approaches (e.g., the deployment of word of mouth evangelists for brands, house parties, experiential marketing, and the like).

some marketers are developing integrated approaches to drive word of mouth advocacy, deploying “new” and “old” media in combination. For example, some of the most successful super Bowl ads in recent years have relied on social media and pr to drive consumer awareness and engagement in advance of the super Bowl ad, thereby raising awareness and anticipation, while the ad itself spurs further conversation during and following the game. often there is a consumer promotion tie-in that further extends the opportunity for consumer engagement.

Beyond recognition of the power of word of mouth and the influence it has on consumers, brands are motivated to find ways to tap the power of consumer word of mouth in response to the current economic realities they face. often, out of economic necessity, the marketing community is moving away from strategies that emphasize mass reach to strategies that involve targeting fewer people with a message that is more relevant and compelling and thus more likely to be acted upon. It is a change that emphasizes efficiency and effectiveness over scale.

today’s tougher economic climate also brings with it a demand for roI in marketing, and as a new approach, word of mouth requires such proof. While many people intuitively accept the philosophy of Wom, the need to justify investments in promoting word of mouth programs is a key challenge faced by the word of mouth marketing industry. a critical issue in this regard is the lack of a structured approach to measure the economic

1

Page 2: A Holistic Approach to the Measurement of WOM

2Copyright © ESOMAR 2009

WORLDWIDE MULTI MEDIA MEASUREMENT 2009

pART 5 / THE pOWER OF SOCIAL MEDIA

value of word of mouth. While research can document the rising importance of word of mouth to consumers, and point to strategies to help companies generate more Wom, this is not always enough. In order to justify marketing investments, managers need to understand how word of mouth actually affects the bottom line.

this paper reports on the findings of a unique approach to word of mouth measurement, developed by the Keller Fay Group, and a unique methodology to quantify the impact of word of mouth on a firm’s profitability based on modeling work by professor Libai. (note: throughout this paper, we cite statistics and insights that come from the authors. unless otherwise indicated, all statistics related to the quantity, quality, and drivers of consumer word of mouth come from talktrack®, the Keller Fay Group’s ongoing system for measuring both online and offline Wom; all statistics relating to the economic value of Wom comes from work by Barak Libai, either independently or as co-author with professor eitan muller and professor renana peres.)

Keller Fay’s word of mouth research takes a holistic approach, measuring all word of mouth – both offline as well as online word of mouth; all people involved in word of mouth – both “speakers” (what people say when they discuss products, services, and brands) as well as “listeners”, and the impact Wom has on them; and is the only ongoing study of word of mouth that is representative of the total u.s. population.

In order to measure the economic impact of word of mouth, professor Libai and his colleagues have used computer simulations – including inputs from Keller Fay’s research – to determine the monetary value of an individual due to his or her word of mouth effect on others, or what might be labeled the social value of this customer. In addition to measuring the social value of customers, this paper also focuses for the first time on the word of mouth impact, from a profit perspective, of influencers compared to that of other customers and finds that influencer marketing can increase long-range profitability by as much as 44%.

MEASURING WORD OF MOUTH: QUANTITY, QUALITY, AND INFLUENCES

the source of primary data for this article is an ongoing series of studies from the Keller Fay Group known as talktrack®. every week since June 2006, talktrack® interviews a fresh, nationally representative sample of 700 americans ages 13 to 69 about the “conversations” they participated in the day before the interview. this translates into 36,000 interviews per year, with data collected about 350,000 brand conversations annually.

the surveys are administered online. the participants are presented initially with a two-page diary they use to keep track of conversations in 15 product categories for a single day. a primary purpose of the diary – beyond reminding respondents of the areas that we are interested in studying – is to help respondents to remember the brands they talk about during the study day. the following day they complete a 20-minute questionnaire in which they list the brands that came up in conversation and then answer detailed questions about their conversation regarding each brand.

talktrack® collects continuous data related to the quantity, quality, and drivers of word of mouth. We measure the volume of word of mouth about products, services, and brands, and as such are able to estimate (at any moment in time, or over time) how many conversations per day about products, services, and brands consumers engage in, which in turn allows us to monitor the absolute and relative volume of word of mouth about thousands of brands across 15 different product/service category areas – including food/dining, media/entertainment, beverages, travel services, shopping and retail, automotive, technol-ogy, telecommunications, finance, health and health care, personal care/beauty, the home, children’s products, household products, sports/leisure/hobbies.

In addition, talktrack® collects data related to the medium (mode of conversation, venue, and “sender” demographics); to the message (positive/negative polarity, perceived credibility); and to the audience (demographics of “receivers” and relationship to “sender”). It also identifies the drivers of brand mentions, including

Page 3: A Holistic Approach to the Measurement of WOM

3Copyright © ESOMAR 2009

WORLDWIDE MULTI MEDIA MEASUREMENT 2009

pART 5 / THE pOWER OF SOCIAL MEDIA

customer experience and marketing communications, and the outcomes of those mentions, such as intention to purchase, to get more information (“inquiries”), and to pass along to other consumers what was learned (“relays”).

our method is distinguished from other approaches used to study word of mouth (many of which monitor online conversation via blogs, chat rooms, message boards, and the like) in that talktrack® measures all word of mouth (offline as well as online), all participants (both “talkers” as well as “listeners”), and all consumers (because the study is designed to be representative of the total u.s. population ages 13 - 69, regardless of how active they are in word of mouth).

While the database that has been compiled over a nearly three-year period is extensive and allows for many types of analysis and drill downs, here are five of the most important – and in some cases, surprising – insights into word of mouth that have emerged as a result of this continuous study.

1. Brands are important as social currency: Billions of brand impressions created each day via WOM the most fundamental finding from our research concerns the sheer volume of word of mouth among consumers. there are 3.3 billion brand impressions created each and every day in america via word of mouth. Brands, it is fair to say, are a major currency of conversation.

What is significant about this figure is that it is one indication to marketers of the “size of the prize” if they can succeed in generating more and better word of mouth for their brands. Just a single share point of additional word of mouth on a base of 3.3 billion conversations per day – or even a half share point – would yield more than 10 million additional conversations about that brand each day.

the leading categories for word of mouth are food/dining and media/entertainment, with a majority of all americans talking about these categories on a typical day. Beverages, sports, and telecomm round out the top five most frequently talked about product or service categories. (see figure 1).

FIGURE 1 CONSUMERS TALk ABOUT MANY CATEGORIES (pERCENTAGE OF RESpONDENTS HAvE 1+ CONvERSATION IN CATEGORY ON GIvEN DAY)

© 2009 Keller Fay Group 1

Figure 1

Consumers Talk About Many Categories (Percentage of respondents have 1+ conversation in category on given day)

19%

20%

22%

24%

25%

30%

33%

34%

37%

37%

39%

40%

43%

51%

55%

Travel Services

Household Products

Children's Products

The Home

Personal Care & Beauty

Financial

Health & Healthcare

Automotive

Shopping, Retail & Apparel

Technology

Telecommunications

Sports, Recreation & Hobbies

Beverages

Media & Entertainment

Food & Dining

Base: Respondents, n=37,351 Source: TalkTrack®, January – December 2008

Base: Respondents, n=37,351 Source: TalkTrack®, January – December 2008

Page 4: A Holistic Approach to the Measurement of WOM

4Copyright © ESOMAR 2009

WORLDWIDE MULTI MEDIA MEASUREMENT 2009

pART 5 / THE pOWER OF SOCIAL MEDIA

Less frequently discussed categories – e.g., household products and children’s products – are not unimportant word of mouth categories. the reason these categories fall lower on the list is because conversation about brands in the category tend to be focused among narrower market segments such as adult women and parents. When one looks at results for women and parents, the percent of people having daily conversation rises dramatically. Further, even among the total population, one in five or more consumers have daily conversations about every one of these 15 categories.

We have known since the beginning of talktrack® that word of mouth is sensitive to news or marketing

programs when we look at results at the individual brand level. the economic challenges that began in the fall of 2008 now make clear that the dynamics of conversation are also affected at the more macro level, responding to economic circumstances. For example, word of mouth about financial services brands rose noticeably once the economy took center stage in september 2008. at the same time, we have seen increasing Wom about media/entertainment and sports, as well – precursors of trends toward higher box office sales for movies during the downturn. at the opposite end of the spectrum, we have seen declining Wom about a range of categories that require higher levels of expenditures, such as autos and travel. (see table 1).

TABLE 1 FINANCIAL CRISIS NOT JUST IMpACTING FINANCIAL CATEGORY MORE WOM ABOUT FINANCIAL SERvICES, SHOppING/RETAIL, AND MEDIA/ENTERTAINMENT; LESS ABOUT AUTO, BEvERAGES, AND TRAvEL (CATEGORY MENTIONS AS A pERCENTAGE OF ALL WOM MENTIONS AMONG ADULTS)

(Ranked by point Change Dec 2008 vs. Aug 2008)

12 Months Ending Aug 2008

Sept – Dec 2008 % point Change

Financial 5.6% 6.5% +0.9

Shopping, Retail & Apparel 9.7% 10.2% +0.5

Media & Entertainment 11.1% 11.5% +0.4

Children’s products 2.4% 2.7% +0.3

Sports, Recreation & Hobbies 6.9% 7.1% +0.2

Technology 8.2% 8.3% +0.1

personal Care & Beauty 3.7% 3.8% +0.1

Health & Healthcare 4.5% 4.5% 0

Food & Dining 12.0% 11.9% -0.1

Household products 2.5% 2.4% -0.1

The Home 1.4% 1.3% -0.1

Telecommunications 7.0% 6.8% -0.2

Travel Services 4.5% 4.2% -0.3

Beverages 10.6% 10.2% -0.4

Automotive 7.0% 6.4% -0.6

Base: All Conversational Brand Mentions (12 Months Ending Aug 2008, n=251.773; Sept-Dec 2008, n=81,274) Note: Percentage point change figures are derived from using more than just one decimal place. Percentages may round to just less than 100% because “other” is not shown. Source: TalkTrack®, September 2007 – December 2008

Page 5: A Holistic Approach to the Measurement of WOM

5Copyright © ESOMAR 2009

WORLDWIDE MULTI MEDIA MEASUREMENT 2009

pART 5 / THE pOWER OF SOCIAL MEDIA

Most word of mouth takes place offline How and where do word of mouth conversations occur? While social media and other types of digital media are often associated with the rising importance of word of mouth, our research indicates that fully 76% of word of mouth conversations occur “face to face,” while another 16% happen by phone. In other words, a total of 90%+ take place offline. meanwhile, 7% of word of mouth takes place online, of which 3% takes place via email and the same number via instant or text message (3%), while 1% is via blogs/chat rooms.

these findings are significant, because for many marketers, the monitoring of internet blogs and chat rooms has been used as a surrogate for measuring offline word of mouth. Yet our research finds that these conversations are a distinct minority of consumer-to-consumer interactions about brands. and, because talktrack® provides a basis for measuring all word of mouth, online and offline, and for making comparisons among modes of communications, we now know that not only is there more word of mouth that takes place offline, there are distinctly different dynamics for offline and online word of mouth. offline conversations are more positive about brands, more credible (to those on the receiving end of Wom advice), and more likely to lead to purchase intention. While the trends in online conversation about brands are sometimes indicative of the trend in offline conversation, most of the time they are not. one reason this might be the case is that the demographic profile of people talking about brands online is dramati-cally younger than for offline conversations, with half of all online conversations about brands taking place among teens. (see more on the comparison between offline and online word of mouth in Keller Fay, 2008.)

Most word of mouth about brands is positive Whereas many people are surprised by the finding that most word of mouth takes place offline, our findings about the “polarity” of word of mouth are also surprising to many. overwhelmingly, consumers have positive things to say about brands, by a margin of more than six to one. across all brands in all categories, two-thirds (65%) were mentioned in a mostly positive light, and less

than one in ten (8%) in a negative one. the remainder is pretty evenly split between those who say the conversational mentions of brands was a mixture of both positive and negative comments, and those who say the conversation had neither a positive nor negative tone.

the mixture of positive vs. negative word of mouth varies by category, and by brand. certainly some have a stronger position, and some are weaker. For any individual brand, there is a need to evaluate its individual position relative to its category and how this is changing over time and to plan accordingly.

However, the overwhelmingly positive nature of word of mouth is extremely important for marketers, for several reasons. First, it means that we should think of consumers as primarily supportive of brands and companies, in the sense that they want to help connect good brands with good friends. While it is true that stopping a friend from making a bad choice is a helpful act, the most helpful recommendation also offers a replacement choice, and perhaps several. second, these findings suggest that the oft-cited “risk” of participating in word of mouth is likely overblown. the greater risk for marketers likely resides in not engaging in a conversation that is happening with or without the marketer’s participation.

WOM inputs: Marketing communications drive WOM amid the recent growth in popularity of word of mouth marketing, the field is often described as an alternative to “traditional” media and marketing channels. While word of mouth does represent a philosophical breakaway from a one-way, top-down communication model, it does not necessarily mean the abandonment of traditional media and marketing channels. Indeed, about half (48%) of all brand-related conversations include a reference to some kind of media or marketing that was seen or heard by at least one conversational partner.

these media and marketing references run a wide gamut: advertising, editorial and programming, company websites, point of purchase, coupons and other promotions, etc. By medium, television drives the largest number of

Page 6: A Holistic Approach to the Measurement of WOM

6Copyright © ESOMAR 2009

WORLDWIDE MULTI MEDIA MEASUREMENT 2009

pART 5 / THE pOWER OF SOCIAL MEDIA

conversations (15%), followed closely by the internet (11%). By media/marketing type, advertising emerges as the most referenced in Wom conversation (20%). We estimate that there are 716 million daily conversations about brands in the u.s. that are “advertising-influenced,” and this is probably a conservative estimate of advertising’s role in word of mouth because it only counts conversations where advertising is specifically mentioned by one or more participants. as a result it does not include occasions when an advertisement indirectly or unconsciously motivated or provided content to a conversation about a brand but where nobody directly references advertising as a source of information in the conversation.

this leads to the question, when advertising influences word of mouth, do we see different levels of efficacy? the answer is yes, in one important respect. Wom that is “ad-influenced” is about 20% more likely to bring with it enthusiastic brand recommendations. Fully 46% of all ad-influenced Wom involves a strong recommendation to “buy or try” the brand versus 39% of other Wom. meanwhile, Wom that doesn’t involve a reference to advertising is far more apt to have no recommendation at all (31%) compared to ad-influenced Wom (18%).

so, while many consider Wom to be part of “new media,” we now see that “traditional” media and marketing channels can also be counted among the important

“input” tools available to marketers interested in driving word of mouth on behalf of their brands.

Influencers: At the center of the word of mouth conversation In the six years since The Influentials (Keller and Berry, 2003) was published, interest in word of mouth marketing has grown rapidly and with it has come increased demand for more and better insight on the role of influencers in stimulating word of mouth conversations and brand advocacy. With the passage of time has come a counter movement as well, in which some critics have questioned the validity of “the influencer model.” (see for example, Watts 2007.)

new research on the impact of influencers on firm profitability will be discussed later in this paper, but in this section we want to share data on the volume of word of mouth among influencers vs. others to help illustrate the disproportionate role they play in word of mouth.

We find clear and consistent evidence that influencers (a group we at Keller Fay label conversation catalysts™) have a far greater-than-average involvement in word of mouth. compared to the average american, they have 80% more conversations each week about products and services, and are 130% more likely to engage in brand-specific word of mouth. (see figure 2).

FIGURE 2 INFLUENCERS: ENGAGED IN FAR MORE CONvERSATION (AvERAGE # WEEkLY CONvERSATIONS AND BRAND MENTIONS*)

© 2009 Keller Fay Group 2

Figure 2Influencers: Engaged in Far More Conversation (Average # weekly conversations and brand mentions*)

80

145

56

129

Conversations Brands

Total Public Conversation Catalysts™

*Based on research in US *Based on research in the United States

Page 7: A Holistic Approach to the Measurement of WOM

7Copyright © ESOMAR 2009

WORLDWIDE MULTI MEDIA MEASUREMENT 2009

pART 5 / THE pOWER OF SOCIAL MEDIA

according to our projections, conversation catalysts™ are responsible for one-third of all the brand impressions every day in the united states via Wom, a level more than two and a half times their demographic representation. this translates into about one billion brand impressions that are being created each day in the united states as a result of word of mouth conversations involving influencers. their opinions are sought out by their friends and family, and they are far more prolific than the average american in spreading the word about products, in category after category, for brand after brand. among consumers who are influencers at the category level (e.g., Financial catalysts, or auto catalysts, etc.) word of mouth levels about brands can be as much as five times as high as it is for the general population. a detailed analysis of the word of mouth of conversation catalysts™ is found in Keller, Fay, and Berry (2007).

THE SOCIAL vALUE OF WORD OF MOUTH, AND OF INFLUENCERS: THE IMpACT ON FIRM pROFITABILITY

Introduction In this section we present a way to measure the word of mouth monetary value of opinion leaders. It demonstrates how the empirical data gathered in comprehensive systems such as the Keller Fay Group’s talktrack®, can be integrated into simulations to enhance our understanding of the non-trivial way in which customer word of mouth turn into monetary value. similar to the holistic approach in getting empirical word of mouth data, our approach to social value is holistic, and takes into account multiple ways in which social interactions create profitability.

the need to justify investments in promoting word of mouth communications of customers, and opinion leaders in particular, is a key challenge faced by the word of mouth marketing industry. a large number of market research firms and agencies help their clients find ways to build brand and online communities, and to identify, recruit, and affect opinion leaders in the hope that they will further influence their social systems. this development is supported by a number of well known books that have drawn attention to the importance

of opinion leaders, also referred to as influentials, influencers, connecters, catalysts, and hubs and by a plethora of cross-discipline academic research since the 1950s that has examined the role of opinion leaders in the spread of new ideas and the growth of innovations.

However, the validity of this practice is still in question. a well-known example is a recent study by Watts and dodds (2007) based on computer simulations. In this study, they argue that influencers do not necessarily start contagion processes that differ greatly from those begun by ordinary social system members. more broadly, they suggest that the impact of opinion leaders has been much exaggerated, and that influencers are not required for social epidemics. Firms are further advised that marketing strategies should not focus on finding influencers, but rather should be directed toward helping large numbers of ordinary people to reach and influence others like them (Watts 2007). these claims have attracted wide media attention, as well as debates among marketing professionals in this industry, with the question still open.

a critical issue in this regard is the lack of a structured approach to measure the value of opinion leaders to the firm. the broad cross-discipline academic literature on opinion leadership has historically focused on the characteristics of opinion leaders and their communication behavior, i.e., identification of opinion leaders, how many others are affected by them, and why. In order to justify marketing investments, however, managers will need to further understand how opinion leaders actually affect the bottom line. We label the monetary value of an individual due to his or her word of mouth effect on others the social value of this customer. We take two steps. First, we suggest a general approach for the analysis of the social value of customers. then we use this approach to examine the value of opinion leaders in the context of a new product’s growth. similar to Watts and dodds (2007), we use computer simulations to examine this question. However, from a monetary perspective, our conclusions differ. the main insights we suggest are:

• the social value of a customer can be assessed as the effect that the absence of this customer will have on

Page 8: A Holistic Approach to the Measurement of WOM

8Copyright © ESOMAR 2009

WORLDWIDE MULTI MEDIA MEASUREMENT 2009

pART 5 / THE pOWER OF SOCIAL MEDIA

the firm’s long-range profits, beyond that customer’s direct lifetime value. • customer social value stems from two main sources: additional customers that would not have purchased the product, and the acceleration of the purchase process. acceleration is an under-explored phenomenon, but can drive much value. • Individual-level simulations are an essential tool for understanding the role of word of mouth (Wom) in firms’ profitability. • Focusing a Wom customer recruiting program on opinion leaders can increase the net present value (npV) of the firm’s revenue considerably, due to acceleration of the adoption process. In our simulations, seeding the market with opinion leaders has increased long-range profitability between 6% and 14% over the alternative of approaching random customers. When compared to an alternative of a no seeding program, the opinion leader program provided long-range surplus value of 11%-44%. In more recent analysis, we find that a competitive environment further enhances the importance of opinion leaders.

Customer Social profitability Before getting to the value of opinion leaders, we need to consider how to measure the social value of customers in general. Here we tie the measurement of social value to what turns out to be the central metrics of modern marketing: customer Lifetime Value (cLV) and customer equity.

cLV represents the expected profit or loss of a firm from a customer, measured by the present value of the expected customer’s cash stream over a future relationship time period. an important input for cLV is the discount factor, which helps translate future money to today’s value. the idea that a future cash stream needs to be discounted is a basic element of financial management, and as we will show, has an important role in calculating the social value of opinion leaders. the discount rate is not only a function of interest rates in the economy and the return on alternative investments, but also of the risk associated with the specific industry.

typically, managers will not be interested in the profitability of a single customer but a group of them, or all of them. customer equity is the sum of the firm’s lifetime values from a group of customers. of special interest is the total customer equity – the customer equity of all present and future customers. In a sense, total customer equity is the ultimate measure of marketing, i.e., all marketing efforts are aimed at increasing total customer equity.

customer profitability measurement has historically focused on the profit to the firm from customer transactions only. to differentiate it from other profitability measures discussed next, we label the classical customer equity measure direct customer equity. In contrast to direct equity, we label as the social customer equity of a group of customers the impact on profitability of that group because of their word of mouth effect.

to examine the social value of customers, we focus here on a case wherein a firm sells a new product (much of the customer-related Wom that we see in the market is associated with the growth of some form of a new product), has some current customers (which we label adopters), and aims to acquire more. We use the term new product in a broad sense: It can be a durable good, a service, or even a behavior that the firm wants to drive (e.g., online banking). It can also be a new version of an existing product, yet still new for customers. It is the uncertainty and possible risk associated with new products that makes Wom a major driver of their growth.

consider a current customer named mr. smith. mr. smith has the potential to provide the firm with “social value” via one of three major avenues:

Incremental customers. mr. smith can help the firm to acquire, via Wom, customers that otherwise would not have adopted. since each customer acquired has a direct lifetime value, one can argue that the social value of smith is the sum of lifetime value of these customers. this approach is probably the one most widely used by marketing professionals, though often it is not the lifetime value of the extra customers being considered, but rather only short-range (e.g., next year) profits.

Page 9: A Holistic Approach to the Measurement of WOM

9Copyright © ESOMAR 2009

WORLDWIDE MULTI MEDIA MEASUREMENT 2009

pART 5 / THE pOWER OF SOCIAL MEDIA

there are two challenges to the incremental customers approach. the first is that, in practice, it typically counts only the customers that mr. smith has actually affected (the “first degree of separation”). Yet, these customers may in their turn add other incremental customers, and the process can continue further on. therefore, the value of mr. smith to the firm is actually higher when taking into account further degrees of separation. However, measurement of the full “ripple” effect is not trivial. the reader may want to look at Hogan, Lemon, and Libai (2004) for an approach to dealing with the calculation of a full ripple effect.

the other challenge to incremental customers relates to their use. It may be argued that in many cases, people will eventually adopt the product without mr. smith. customers are connected to multiple social networks, and are affected by mass media, so eventually, if the product fits them well, they will get it anyhow. therefore, it is not clear if it is justified to allocate the lifetime value of customers specifically to smith.

Marketing cost reduction. a certain version of the incremental customer approach is to argue that if not acquired via Wom, customers would be acquired via marketing actions such as advertising. (there are other marketing actions such as direct mail, store promotions, and salespeople. For our purpose here, all are grouped under the label “advertising”). therefore, instead of the lifetime value of the incremental customers, the savings in advertising are added (Kumar, peterson, and Leon 2007).

one could ask of course if advertising actually substitutes for word of mouth. In the absence of mr. smith, some consumers may eventually adopt based on word of mouth from people other than mr. smith. For many, advertising may not be relevant at all: the large body of research on new product adoption suggests that a considerable part of the market may not adopt a new product without some kind of social effects by others. While advertising can clearly support the Wom process, it is not clear that it can replace it.

Acceleration. the third approach has had less exposure to date, yet can be of much effect. the basic idea is that

by providing Wom to others, a customer accelerates the process of adoption of the new product. this takes into account the fact that that even if we assume that all customers eventually adopt, the absence of smith’s Wom will slow down the process as it will take some customers time to be affected by alternative sources.

Given that customers may eventually adopt and their lifetime value taken into account, one might wonder if the delay due to the absence of one individual could have a notable impact. Hogan, Lemon, and Libai (2003) have used the acceleration effect to calculate Wom value at the aggregate level, and have shown that it is indeed sizeable, especially in the early part of the product life cycle.

to see the intuition, consider a case wherein mr. smith “disappears” and stops affecting others via word of mouth. this will alter the whole ripple effect otherwise created by mr. smith. not only might adoption by those close to smith in the social system be delayed, but further degrees of separation may be delayed subsequently (of course, the extent depends on the exact structure of the social system and the nature of the product). For each adoption that is delayed, the lifetime value of this late adopter is affected via the discount rate. overall, the effect on customer equity can be substantial.

Following the above, because of the possible long-range effect of Wom, and due to the fact that in the absence of a particular individual other means of communication may compensate, we suggest that the way to analyze the social value of customers is to “take them out” of the system, and observe the effect on profitability. thus, the social value of a customer can be assessed as the effect that the absence of this customer has on the firm’s customer equity, beyond its direct lifetime value. the measurement of customer equity before and after mr. smith “disappears” will take into account the acceleration effect, as well as any incremental acquisition or advertising savings, and in all degrees of separation.

The role of opinion leaders to examine the question of opinion leaders, we need to build a social system wherein an individual is an

Page 10: A Holistic Approach to the Measurement of WOM

10Copyright © ESOMAR 2009

WORLDWIDE MULTI MEDIA MEASUREMENT 2009

pART 5 / THE pOWER OF SOCIAL MEDIA

opinion leader, and then “take him out” to see the impact. However, herein we are interested in a comparison to an ordinary person in the social system. so the differential effect of opinion leaders is the customer equity difference between a case wherein they are part of the social system, and the case wherein non-opinion leaders would be part of the social system in their stead.

the two dimensions whereby opinion leaders have been historically characterized are connectivity and persuasiveness (see for example Goldenberg et al, 2006). the first is related to how many others they have in their social system; the second considers to what extent they have a stronger impact on individual others, for example through their expertise or their interest in the subject. We will use both dimensions to construct an “opinion leader” in the following analysis.

We will examine the case of a group of customers connected in a social network that together form a “social system”. a firm introduces a new product, and this social system gradually adopts it. From the firm’s perspective, an individual’s adoption means a certain expected lifetime value. due to the discount rate, later adopters are worth less today.

potential users become adopters following one of two influences. First, there is a certain probability that an individual will adopt based on advertising. second, via word of mouth, each adopter has a probability of affecting non-adopters in his or her social system.

Beyond advertising, the firm may choose to create a seeding marketing program that will impact word of mouth and accelerate adoption. the firm may choose a number of potential adopters and affect their adoption behavior via word-of-mouth programs, i.e., by giving them samples and/or exposing them to information about the brand, or promotions. the expectation is that if these customers adopt, it will create Wom processes that will bring further adoption. We label this group of targeted customers ‘seeds’.

the firm faces two alternatives in this regard: one is to acquire potential customers to the program randomly.

the other is to target influencers, who have a stronger effect on others via their large social networks, as well as their persuasiveness. some methods to get to such opinion leaders include programs for identification of and communications with influencers; recruitment of influencers as agents in Wom programs; and the establishment of “brand communities”. such efforts may be costly; to justify them, one needs to understand to what extent the social value of an influencer is higher than that of a random customer.

the simulation we present next is part of an ongoing research effort to explore the social value of customers. as part of this effort, simulations will be used to examine optimal investment in Wom programs, the interaction of Wom and advertising, the impact of competition on the social value of customers, and related issues. Herein, however, we examine one question only: Given that a single seller that decides to affect the adoption of a group of customers, what will be the differential effect if it does so via opinion leaders vs. random customers? We do not look at the costs of these programs or any other marketing effort, so we will answer questions based on optimal investments. our simulation will focus on the acceleration effect, i.e., we assume that the market potential has been correctly identified, and thus eventually all potential customers may adopt the new product.

We consider a social system with 2,000 members. to assign members to social networks in a realistic way, we were assisted by the Keller Fay Group’s talktrack®, discussed above. using talktrack® data on the distribution of social networks, we built a similar distribution of social networks for our analysis.

the members of this social system, whom we label influencers or opinion leaders, are a group of customers in the top 10% in terms of their social system sizes, which have been found to be about three times the surveyed average. We also assume that they have stronger brand-related impact. using the talktrack® studies, Keller Fay found that catalysts, the group that influences others most, discuss brands twice as much on average compared to others. Following these findings,

Page 11: A Holistic Approach to the Measurement of WOM

11Copyright © ESOMAR 2009

WORLDWIDE MULTI MEDIA MEASUREMENT 2009

pART 5 / THE pOWER OF SOCIAL MEDIA

and interviews with professionals in the Wom marketing industry, we decided to assign a double persuasiveness score to opinion leaders.

other parameters we used relate to the average effect of advertising and Wom on individuals, i.e., what is the probability that an individual will be affected by Wom or by advertising in a single period (say a year), and ultimately adopt the product? We relied on previous academic research in this area that established a range of value for parameters in such simulations. For details, see Goldenberg, Libai, and muller (2001).

We used a yearly 10% discount rate that is consistent with previous studies of customer profitability. For simplicity, we assumed that an adopting customer brings the firm one unit of lifetime value (because we do not consider costs here, the exact sum is of less importance).

Results the main results are presented at table 2.

Looking at table 2, the left-hand column reflects the size of the seeding program as a percentage of the total population. We consider programs that use from 0.5% to 7% of the population. column 2 presents the customer equity (long-range profitability) if there is no Wom program. column 3 is the customer equity if the program recruits customers randomly (among them, of course, possible opinion leaders). column 4 is the customer equity if the program includes only opinion

leaders. columns 5 and 6 are the improvement of the two programs over the no-program option.

a number of insights can be gained from these results:

a) Wom programs to encourage early adoption have a considerable effect on profitability. even if random customers are chosen, the increase of customer equity over the no-program option is substantial. of course, one would have to examine the costs of such programs to determine net profit. b) However, getting to influencers in the Wom program yields higher customer equity than do random customers. the absolute difference between the two options increases for larger programs (3%-7% of the population) compared with small programs (0.5%-1% of the population). For the larger programs, the difference between the two options in terms of customer equity is around 14%. c) acceleration matters. note that we did not make any assumptions that customers will not adopt without the program. thus, we believe our results are conservative regarding the contribution of the Wom program: If there is incremental customer acquisition, we expect the word of mouth effect to grow. d) We see a diminishing effect of additional members in the program. While the profits from the Wom programs increase with the size of the program, the effective-ness of each dollar invested goes down. consider the

TABLE 2 LONG-RANGE pROFITABILITY UNDER vARIOUS SEEDING OpTIONS

1 2 3 4 5 6

proportion of Seeding

Customer Equity No program

Customer Equity Random Seeding

Customer Equity Influencer Seeding

Improvement Random Seeding

Improvement Influencer Seeding

0.5% 880 923.9 979.3 5.0% 11.3%

1% 880 961.1 1033.4 9.2% 17.4%

3% 880 1040.4 1157.4 18.2% 31.5%

5% 880 1100.2 1225.6 25.0% 39.3%

7% 880 1146.7 1268.5 30.3% 44.2%

Page 12: A Holistic Approach to the Measurement of WOM

12Copyright © ESOMAR 2009

WORLDWIDE MULTI MEDIA MEASUREMENT 2009

pART 5 / THE pOWER OF SOCIAL MEDIA

improvement in influencer seeding in table 2. With a 0.5% program, customer equity improvement is 11.3%. With a 1% program, improvement is 17.4%, which is less than double. the meaning is that the social value of different customers is not additive to the social equity of the group. the reason may relate to possible overlap among program members in terms of social networks. this diminishing returns phenomenon can help marketers plan what would be the optimal investment in Wom programs. e) the next step is to move the measurement to a competitive scenario in which a person may be affected by more than one brand. Initial simulations we con-ducted confirm the basic results we present here also in a competitive environment. In fact, that relative role of opinion leaders programs is even stronger under competition, which helps to explain the ubiquity of such efforts.

CONCLUSIONS

using simple simulations integrated with empirical data from the talktrack® system, we find that seeding the market with opinion leader programs can have substantial impact on the long-range profitability of firms. Hence, generalized claims that influencer programs are a waste of money may not reflect well the market reality. one take from the above is that in order to realize the potential of the social value of customers, marketers need to take into account the full effect on customer equity, and not use proxies and short-term measures, which may be misleading.

References Goldenberg, Jacob; Donald R. Lehmann; Daniella Shidlovski; and Michal Master Barak (2006), “the role of expert versus social opinion Leaders in new product adoption”, Marketing Science Institute. Paper [06-124].

Hogan, John E.; Katherine N. Lemon; and Barak Libai (2003), “What is the real Value of a Lost customer?” Journal of Service Research, 5(3), 196-208.

Hogan, John E.; Katherine N. Lemon; and Barak Libai (2004), “Quantifying the ripple: Word of mouth and advertising effectiveness”, Journal of Advertising Research, 44(3), 271-280.

Keller, Ed and Jon Berry (2003), the Influentials: one american in ten tells the other nine How to Vote, Where to eat and What to Buy. new York: Free press.

Keller, Ed, Brad Fay and Jon Berry (2007), “Leading the conversation: Influencers’ Impact on Word of mouth and the Brand conversation.” Measuring Word of Mouth: Current Thinking on Research and Measurement of Word of Mouth Marketing, pp 173-181. Womma:

Keller, Ed and Brad Fay (2008) “comparing online and offline Word of mouth: Quantity, Quality, and Impact,” Measuring Word of Mouth: Current Thinking on Research and Measurement of Word of Mouth Marketing, pp 91 – 103. Womma.

Kumar, V.; Andrew Petersen; and Robert P. Leone (2007), “How Valuable Is Word of mouth?” Harvard Business Review, 85(10), 139-146.

PQ Media, Alternative Media Forecast: 2008-2011. Found at .http://www.pqmedia.com/alternative-media-forecast-2008.html

Watts, Duncan J. (2007), “the accidental Influentials,” Harvard Business Review, February, 2007, pp 22-23.

Watts, Duncan J. and Peter Sheridan Dodds (2007),”Influencers, networks, and public opinion Formation”, Journal of Consumer Research, 34(4), 441-458.

The Authors Ed Keller is CEO, Keller Fay Group, United States.

Barak Libai is Associate Professor of Marketing, Recanati Graduate School of Business, Faculty of Management, Tel Aviv University, Israel.