July 2006
← June 2006 | Research Blog Home | August 2006 →
The Holmes Report's decision to summarize the concepts in Fred Reichheld’s The Ultimate Manifesto this Monday might be an indication of the growing value of WOMM (and, therefore, its measurement) in public relations.
As PR professionals continue their quest to develop measurement techniques that illustrate the impact of their work, Reichheld’s Net Promoter Score (NPS) becomes a valuable piece of the evaluation puzzle. The potential for using NPS in crisis preparations, CRM and strategic communication programs makes this book, and its seminal article in the Harvard Business Review, required reading for the client-focused practitioner.
Besides observing changes in Net Promoter Scores, measuring frequency of WOM episodes would also factor well into ROI analyses. Understanding how many stakeholders are discussing a topic before and after a campaign could provide compelling data alongside media monitoring and content analyses.
To further investigate campaign impact, PR research might also investigate WOM participants in relation to specific key messages.
Variables for WOM creators might include:
Facilitation
Where did you learn that (message)? Select all that apply. [List key placements]
Velocity
How soon after learning that (message) did you tell someone else about it?
Distribution Spread
How many people did you tell that (message)?
Variables for WOM receivers might include:
Outcome Rates
After hearing (message), which of the followings actions did you do?
[Inquiry / Conversion / Relay / Re-Creation / No Action]
Source Diversity
From which of these sources did you hear that (message)? Select all that apply.
Credibility Ratings for Source Diversity
Which of these sources was most persuasive in your decision to (outcome)?
In brief, if the campaign goals are to be measured by percent shifts in Net Promoter scores, researchers must first identify what facilitates WOM, track the evolution of information-gathering habits and understand how key opinion groups determine source credibility.
Find more info about NPS and links to related sources Walter Carl's post on this blog.
Products that are easy to discuss and recommend to others are more likely to achieve WOM success than goods solely perceived to be new and innovative. That's a key finding from a new study by Keller Fay Group, in association with BzzAgent. Nearly half (45%) of the study's participants chose to promote products that were "easy to talk about". "Something that is worth talking about" and "Something I'm proud to share with others" followed closely behind at 42% and 40%, respectively.
Learn more and register to read the report
For the first time, social networking site MySpace is the number one ranked web site in the U.S. That's the key finding of new tracking data from Hitwise.
* MySpace accounted for 4.46% of all Internet visits in the U.S. for the week ending July 8, 2006.
* Among top social networking sites in June 2006, MySpace accounted for 79.9% of the total market share of visits.
* Following MySpace were Facebook (7.5% of visits), Xanga (3.8% of visits), Yahoo! 360 (1.1% of visits) and Bebo (.98% of visits).
Learn more
"Studio 60 on the Sunset Strip", a new TV show set to debut on the NBC TV network this fall, is generating the most discussion interest on line. That's the key finding of a new study from Brandimensions. Tracking these conversations offers a glimpse into what shows consumers are talking about with each other.
* "Studio 60" was talked about in 22.2% of online audience discussions.
* The show also received the highest "sentiment" score (the most positive audience feeling about a show) with a 3.81.
* Combining results for all new shows on each net, NBC generated 41 percent of all new show discussion on the Web, with CBS second at 19 percent, ABC third at 15 percent.
Go to Brandimensions web site
Mediaweek story on the report
A mutual fund manager is more likely to buy (or sell) a particular stock in any quarter if other managers in the same city are buying (or selling) that same stock, according to a paper published in the December 2004 issue the Journal of Finance. The effects of WOM were found to be distinct and separate from local preferences. That is, WOM had a greater effect and emerged even when local stocks were completely excluded from the analysis.
Read the paper (PDF download)
Asian American youth are more likely to identify and seek out trends through word of mouth messaging than through any other media channel or network. That's the key finding of new research from New American Dimensions and interTrend Communications. This trend is particularly true for second-generation Asian Americans, who indicated they are twice as likely as their first generation counterparts to learn about trends via word of mouth.
Learn more
The amount of Internet buzz a movie generates is a strong predictor of its box-office take. But it hardly matters whether that buzz is good or bad, according to a study by Yong Liu of Syracuse University.
* Liu tracked 303 postings about a major movie in the summer of 2002 at Yahoo! Movies. Nearly one-third (32.5%) of these were negative.
* WOM offered significant explanatory power for both aggregate and weekly box office revenue particularly in the early weeks after a movie opened. The effect of WOM was magnified by the sheer volume of it, and not from its valence (the percentages of positive and negative messages).
Read a summary
Read the New York Times write-up
Podcasts have become a popular medium for spreading WOM, with usage among BtoB technology buyers both significant and growing. That's the key finding of new research from KnowledgeStorm and Universal McCann.
Among the key findings:
* 65% of respondents listened to podcasts for both personal and business interests.
* 72% downloaded or listened to podcasts on technology topics on more than one occasion. 23% did so "frequently."
* Nearly 60% reported that information on business or technology topics, currently delivered as white papers or analyst reports, would be more interesting as podcasts.
* 55% would be more likely to consume white papers and analyst reports if they were delivered as podcasts.
* 27% reported podcast content already influenced work-related purchase decisions.
Learn more
A paper published in the Journal of Advertising Research in December 2004 revealed four email profiles used to identify those more likely to pass-along emails. Advertisers tend to focus on reaching and connecting with viral mavens, but some infrequent senders who received few but forwarded were found to represent another potentially overlooked but valuable conduit for marketing messages.
The four types of profiles included:
* Infrequent Senders 1: Those who neither received nor forwarded many pass-along emails.
* Infrequent Senders 2: Those who received few, but forwarded a large percentage of emails.
* Viral Mavens: Those who both received and forwarded many emails.
* Infrequent Senders 4: Those who received many emails, but forwarded none.
Read the report (PDF download)
There are 8.68 million bloggers and 7.16 million social networking service subscribers (SNS) in Japan as of March 2006. This was the key finding of a report entitled "Information and Communications in Japan" published by the Japan Ministry of Internal Affairs and Communications.
Other highlights of the report include:
* Over a one-year period, the number of bloggers increased 2.6% (up from 3.35 million)
* The number of SNS users increased 6.5% to 7.16 million over the same one-year period.
* 60% or more people study about products from blogs and other Internet sources before they purchase them.
Learn more (English)
Summary of the Ministry report (Japanese)
A new study from crmmetrix entitled "BaroBlogs: The Barometer of the Blogosphere" finds that France is one of the leading blogging countries outside the US and its corner of the blogosphere is growing fast. Key findings from the study include:
* 26.7% of the French online population visit a blog at least once a month.
* 18.8% have posted a comment on a blog and 8.1% have created there own blog.
* 92.2% agreed blogs enable the greatest freedom of expression.
* 75.7% thought blogs create a closest possible relationship between people.
* 62.9% considered blogs more critical than any other source of information.
Learn more
Word of mouth is the top influencer on young TV viewers in deciding what to watch. That's the key finding of a new Magid study. Messages most likely to be passed on are likely to be funny and entertaining, scary, or appealing to lifestyle. The study identified "super disseminators" as the core group that passes on messages about TV programs, spreading WOM to six or more people on a regular basis.
Learn more
Weak and strong social ties play different roles in WOM referral behavior at both micro and macro levels. That was the key finding of a paper on WOM published in the Journal of Consumer Research in December 1987. Highlights of the paper include:
* At the macro level, weak ties displayed an important bridging function, allowing information to travel from one distinct subgroup of referral actors to another subgroup in the broader social system.
* At the micro level, strong ties were more likely to be activated for the flow of referral information than weak ties.
Learn more
Blogs and podcasts have become strong vehicles of consumer-generated media and word of mouth marketing, so naturally they have attracted the attention of advertisers. Currently, blog advertising leads the way, but podcast advertising should surpass that by 2010.
Those are the key findings of a recent report by PQ Media, which concluded:
* Podcast advertising is projected to grow at about 154.4% annually to $377 million in 2010.
* Blog advertising will grow more slowly, but a still-healthy 78.4% to $300.4 million in the same amount of time.
Learn more
In the video game business, word of mouth has become one of the most important factors influencing individual purchasing decisions. In fact, word of mouth specifically passed from friends and family ranked as the number one factor in game buying decisions in all stages of purchase. That's the key finding of a new survey presented during MI6: Marketing Interactive '06.
Survey highlights include:
* 25% of game buyers said friends and family are the most prevalent source of initial game awareness.
* Friends and family are even more of an influence to those those who bought the game as a gift (39%)
* Positive WOM from friends and family is a top purchase requirement for all types of games.
Learn more
While personal experience still carries the highest influence on consumers, word of mouth and employee opinions are not far behind. Those are some of the key findings of a recent survey from Harris Interactive.
* 92% said personal experience with a company is the most credible.
* 85% said WOM communications is credible.
* 84% found the opinions of company employees believable.
Learn more
s
For WOMMA's WOMBAT 2 conference, Ann Green of Millward Brown, Ed Keller of the Keller Fay Group, and Greg Wester of VoodooVox all co-authored a presentation with the latest research on how word of mouth is being adopted, who is using it, and how it's affecting consumer behavior.
Due to popular demand, this presentation is now available for free download from WOMMA.
This is a cut-and-paste bonanza! All the facts, stats, and data you've been asking have all been assembled in one place. Don't miss it.
Download the report (PDF download)
A recent research paper on the Dynamics of Information Access on the Web finds that 36 hours is the average amount of time it takes for half of the total readership of a given article to have read it. This is far longer than many media researchers might have suspected-- especially given that we know frequent updates are one the biggest predictors of blog readership. It made me wonder what the half-life of WOM is. When does WOM become stale and what are the predictors of endurance?
Interestingly, this research was conducted by a physicist at the University of Notre Dame (Albert-Laszlo Barabasi ), but the write-up of the article gives a faulty citation (Physical Review E, June). I'm still looking for the actual research to determine to what extent we could make an analog for WOM. Regardless, there are numerous implications at a conceptual level: how long can we expect others to heed a WOM unit? Does the reception of WOM occur in bursts or in a more even trajectory? Does the perceived hit-rate affect its trajectory-- that is, would WOM have a longer shelflife if you perceived others still found it "fresh"?
Burst is reporting that nearly 3/4 of all Summer travelers go online during the planning process. This is interesting to us, specifically because of work like Yahoo's Trip Planner.
Travel is one of those categories where we know--intuitively and from personal experience--WOM is key to decision making. What Yahoo has done is bake that insight directly into the overall experience. That is, bringing the individual stories into the decision making process and the purchase funnel.
When we think about what drives WOM, we might also consider phenomena that are similar, but not quite WOM (which we can define generally as C2C communication regarding products, services and brands). What I am suggesting is considering established research in other--RELATED--domains, then thinking about these other research findings in the context of WOM. A few areas that come to mind:
- gossip
- rumors
- memes
- story-telling
Here is just a taste of some insights that seem to be very relevant for us WOM-types:
Rosnow, R. L. (1988). Rumors as Communication: A Contextualist Approach. Journal of Communication(Winter).
This is a classic rumors paper and is an easy read. He summarizes four general predictors of rumors:
1./ Personal anxiety
2./ General uncertainty
3./ Credulity
4./ Topical importance
AREN'T ALL OF THESE RELEVANT TO WOM? AND WOM MARKETING?
This paper has four general conclusions:
- Importance of the rumor is the most controversial factor. (Although we might guess that imporrtance is key, there is contradictory evidence that importance predicts rumor transmission.)
- Anxiety takes precedence over credulity when anxiety is very high or very low (due, in both cases to the amount of attention paid to the credulity). In other words, high anxiety leads to rumors, low anxiety leads to no talk.
- Since rumors are contextually dependent, situations that invite rumormongering for some will not for others. (Similarly, there are both contexts and consumers for which WOM is more prevalent.)
- Rumors can become reality (i.e. uncertainty => certainty).
Heath, C. (1996). Do people prefer to pass along good or bad news? Valence and relevance of news as predictors of transmission propensity. Organizational Behavior and Human Decision Processes 68(2): 79-95.
In this paper, Heath highlights the anecdotal evidence that exaggeratedly bad news may propagate in the marketplace of ideas, more so than good news - even when they are equal in terms of shock value.
DOESN'T THIS MIRROR THE (MIS)CONCEPTION THAT NWOM DOMINATES PWOM?
While WOM researchers have started to look at PWOM vs NWOM in a more systematic way and determined that NWOM does not always dominate, Heath considers the Q in the context of news in general. In a series of three studies, Heath concludes that:
- People typically prefer to pass along central rather than extreme information. (This seems to contradict WOM research, which seems to consistently indicate that consumers who are MORE satified or MORE dissatisfied talk more (see Anderson 1998 Customer Satisfaction & WOM, Journal of Service Research))
- When confronted with extreme information, people prefer to pass along news that is congruent with the emotional valence of the domain in question. This means that in emotionally negative domains, people are willing to pass along bad news even when it is exaggeratedly bad. At the same time, however, people transmit exaggeratedly good news in emotionally positive domains. SO CONSUMERS MAY BE MORE LIKELY TO GENERATE PWOM REGARDING FAVORABLE DOMAIN PRODUCTS (LIKE VACATIONS, FICTION BOOKS, ETC) AND NWOM REGARDING UNFAVORABLE DOMAIN PRODUCTS (LIKE MORTGAGES, FUNERAL SERVICES, ETC) - sound reasonable?
I hope I have provided you with some ideas that will help you de-mystify the phenomenon!
Individuals' receptivity to both positive and negative WOM information is determined largely by its "fit" with their prior evaluative positions. This is the key finding of a 1989 study published in Advances in Consumer Research. Once an individual assumes an evaluative position toward a product, he/she will begin to filter information it. To the extent the information "fits" the evaluative position, the WOM will be accepted.
Learn more
A customer who receives satisfaction during a phone call to a customer service call center hangs up with more customer loyalty than one who is originally satisfied with the product. That satisfied call center customer also makes more future purchases and is more likely to refer friends and family to the product or service. These are key findings of new research from the Center for Customer-Driven Quality.
Call centers can be a resource for a groundswell of goodwill and resulting word of mouth, and the opportunities are growing as the number of calls to increases exponentially. Currently, 60,000 call centers in the United States taking 30 billion calls per year. By 2010, that number will double to 60 billion.
Read the press release
Consumers with a high overall media consumption, particularly in network, print and electronic media, are most influenced by WOM in home improvement purchases. That's the key finding of a study by BIGresearch, which also defined several different categories of media consumers, including:
* Ravenous: high overall media influence, high network, print and electronic consumption.
* Old School: high mass media influence, low online media influence, and average overall media consumption.
* Persuadable: high mass and search media influence, average promotional media influence.
Source: 8 Media Consumption Clusters Discovered From Analysis of BIGresearch's Simultaneous Media Survey, March 2006
Read an associated column
How does the behavior of influencers compare with non-influencers? Jupiter Research analyst David Card recently looked at a set of influencers -- so-called "cool" high school girls (the influencers whom others try to emulate) -- and found that they don't differ much from others in terms of viewing personal pages online, playing games and using email. However, they are in fact more likely than others to be attracted to online activities focused on music, fashion and photos. Card also found that these "cool" influencers were significantly more likely to respond to online advertising by making an online purchase or by registering for a discount or sweepstakes.
Source: Jupiter Research
Read an associated story
David Leonhardt wrote an interesting column about the value of Google Trends in the Business Section of last Wednesday's NY Times. He comments on the ability of this powerful source of CGM (Google searches) to serve as a "real economic phenomenon," beyond what John Batelle once referred to as a "database of human intentions."
I was wondering what WOM researchers think about this and the projected outpouring of other companies coming up with their own methods of capturing search terms. The article conceives of this source of data as "an enormous online focus group," which is interesting given that most of us WOM researchers throw around that terminology to refer to our drastically different database.
Thus, my underlying question: how are search terms related to buzz? Might search terms mediate the relationship between buzz and sales?
In its truest sense, search terms are not WOM Units. So, what power do they really have to translate to sales? It's sort of a question of what has more predictive prowess, actions or feelings? So often do we tell clients that absolute levels of buzz are unimportant without qualifying them with the associated sentiment; popularity is so often tempered by tone. Would Google Trends be more predictive of intentions if they were to incorporate the overall mood (via language) of the internet on a given day when a given search term spikes?
Some folks may have missed this due to the 4th of July holiday here in the States, but Matt Creamer over at Advertising Age wrote an article entitled "Do You Know Your Score?" [print edition, cover story and p. 24 for July 3rd, 2006].
The article does a nice job of talking about how the use of the Net Promoter Score (NPS; the likelihood that a person will recommend an organization, brand, product or service [OBPS] to a friend or colleague) is becoming a standard across many different industries but especially among the word-of-mouth marketing community. The NPS score is calculated by subtracting the number of people who would not recommend the OBPS, or detractors, from the number of people who definitely would recommend the OBPS, or the promoters.
There are several advantages to using the NPS to track performance (whether it be product or service quality, customer service interactions, or the effectiveness of a word-of-mouth marketing campaign) including that it's easy to understand, it's straightforward to calculate, simple to administer, many companies are using it, and it has been tied to revenue growth (tested in both US and UK companies).
Fortunately the article also mentioned some of the drawbacks or cautions about using the score as the "only" or "ultimate" question. When I discuss the NPS during presentations I discuss its wonderful utility as well as four critiques:
1) Is revenue growth the only financial indicator that matters?
2) Is the number of promoters, versus likelihood to promote, a better indicator of revenue growth?
3) Does NPS tell you what’s working and what’s not working?
4) How precise is NPS?
First, besides revenue growth, other indicators are also important such as operating cash flow (the “lifeblood of the firm”), cash flow volatility (indicator of financial risk), Tobin’s Q (the relative value of such intangible assets as knowledge, human capital, brands, and relationships), and the price-to-book ratio (the ability to generate cash from assets). These other indicators are discussed in Neil Morgan and Lopo Rego’s letter-to-the-editor in the April 2004 Harvard Business Review (they are marketing professors at the business schools of UNC-Chapel Hill and University of Iowa, respectively). They conclude that revenue growth doesn’t necessarily equal success in these other metrics and thus companies shouldn't use NPS as the "only" thing to focus on in increasing shareholder value. They and others have also argued that customer satisfaction scores are also correlated with important indicators in certain industries.
Second, Morgan and Rego argue that there is evidence that the number of people promoting may actually be a better indicator of revenue growth. However this is not consistently supported as Paul Marsden and colleagues, who have studied NPS among UK companies, did not found this in their research (learned via e-mail correspondence).
Third, an important concern is that NPS doesn't tell you what's working and what's not working. It's important to know why one has a high or low NPS so that an organization can make appropriate changes, or build on their strengths. Some people worry that organizations will stop at just measuring NPS and won't do enough follow-up research to understand why they're getting that score.
Finally, some people question how precise the NPS metric is. Pissed-off detractors who give an organization a "0" on the 0-to-10-point scale (0 - extremely unlikely to recommend, while 10 is extremely likely to recommend) are weighted the same as "passive" folks who give the organization a 5 or a 6 (this critique is discussed in a Business Week article). To be fair here, though, Fred does talk about how to address each of the three categories -- promoters, passives, and detractors -- differently when trying to improve one's performance and thus NPS score.
In their critique of the NPS in their HBR letter-to-the-editor, Morgan and Rego write that before an organization adopts the NPS, they should ask two questions:
1) What customer feedback measures best predict business performance in my industry?, and
2) What dimensions of my business performance am I trying to maximize?
According to them, an organization should only adopt the NPS if the answer to #1 is the number of net promoters, and to #2, revenue growth and nothing else. They also emphasize the importance of tracking what’s driving the recommendation as well as the number of customers making the WOM recommendations.
However, another way of thinking about whether a company should adopt the NPS is offered by Dr. Laura Brooks from Satmetrix. She argues in a blog post that a company should adopt the NPS if it actually motivates employees (and the organization as a whole) to change their behaviors.
And for a response to why some people hate the NPS, be sure to check out this blog post from Fred Reichheld.
Stay tuned for WOMMA's Measuring Word-of-Mouth, Volume 2 where we'll see how different WOMM companies are using the NPS in their businesses [disclosure: work group editor for this research book].
NOTE: Cross-posted from Word-of-Mouth Communication Study blog
Suresh Vittal has joined Forrester, from SPSS it seems, and wades right into the deep-end, asking the critical questions about the measurement of consumer-generated media:
Will CGM increase in relevance as a customer insight forum?
Who, within the organization, will own the process and tools of analyzing CGM?
How will the learnings be integrated into the current enterprise analytic platforms?
How does this change the roles of creative, analytics and brand marketing?
An interesting aspect of his post, is the companies that he points to as potential measurers, companies like his previous gig, SPSS. These are pretty serious analytic tools, capable of diving not only structure, but also trends. I haven't heard these vendors brought up before in the CGM/WOM space before, but it introduces some interesting ideas. Of course, I don't know if these vendors have the spidering ability of companies like BuzzMetrics or Cymfony. That is, getting the data into the system may be difficult.
A conversation is forming, down in the comments section.
← June 2006 | Research Blog Home | August 2006 →
This is old content! The Research Blog has been retired. For the new stuff, visit 
