Friday, May 26, 2006

Pete Blackshaw's Comment Reponding to "Is Negative WOM More Frequent Online than Offline?"

Pete made some excellent comments in response to my last post "Is Negative WOM More Frequent Online than Offline?". In the interests of promoting dialogue and involving as many others as possible I wanted to "elevate" his comment into a post so that more people might have access to it. Please find his post below (edited only for typos and paragraph spacing to make it easier to read):

Great post, and it truly is an important topic and question. To lend further perspective on my opinion, I do believe consumers are as inclined to provide positive feedback as negative feedback, but as the barriers to providing that feedback increase, the ratio of negative to positive feedback (or recommendation) increases.

This was very clear to me in the early days of developing a consumer feedback portal after leaving P&G. We carefully studied why consumer feedback to companies (and the ensuing "recommendations to others") tended [to be] so negative. The vast majority of consumers (24 or 25) who have a so-called "feedback moment" would fail to provide that feedback because of (1) contact barriers, (2) cyncism that no one would listen or bother to respond, (3) didn't know who to find, etc. Consequently, only the really negative would push through the pipe, and that in turn would cascade (in a highly cathartic manner) to CGM venues like message boards. In so far as it takes time to invest into the feedback process, the emotion behind the negative tends to dominate the exercise.

That said, as CGM venues have become easier and more effortless, we're seeing a growing percentage of neutral to positive commentary, especially in blogs, which are becoming highly transparent mirrors into everyday behavior. Often the "word of mouth" or "recommendation" is incidental or implied, not delibrate.

If what Keller and others suggest is true, I'm even more convinced that companies and brands must be more aggressive in actively inviting -- nay, encouraging -- feedback, and lowering the barriers at all cost to that expression. If a friction-free feedback experience ups the odds of Pampers capturing an evangelist who speaks generously of the brand both offline and via high-impact CGM venues, that's a great payout. I'll think more about this.
Thanks for this Pete! Two comments on your comment:

Your points about customers not able or motivated to tell the company is definitely consistent with other research on this topic. For example, the Customer Dissatisfaction Study released by the Verde Group found that people were 5x more likely to tell friends about a negative experience rather than a company for the following reasons: a) it's not worth customers' time to contact the company; b) customers tried before but had a negative experience in contacting the company; c) customers don't know who or where to contact; d) customers thought it might negatively affect future service; 5) customers didn't think it would do any good. In an interview with Paula Courtney from the Verde Group I learned that about 10% of the people who have a negative experience actually contact the retail store (20-30% in consumer sector).

The other comment is that there is a great research opportunity for a study to be done that demonstrates the hypothesis that 1) there is a growing percentage of neutral and positive commentary in CGM venues, especially blogs, and that 2) this increase is due to CGM venues becoming easier and more effortless to use. Is there some graduate student out there who wants to do a thesis or dissertation on this? ;-)


Thursday, May 25, 2006

Is Negative WOM More Frequent Online than Offline?

I'm trying to make sense of two seemingly contradictory positions about the prevalance of positive and negative WOM (PWOM, NWOM). Let's start first with some recent findings reported by Keller Fay from their TalkTrack methodology (opens into PDF).

Their data show that brand-related interactions are more likely to be "mostly positive" (62%) than "mostly negative" (9%) in terms of valence (aka, polarity or sentiment). And the majority of the WOM episodes represented were offline (92% to be exact, including F2F and telephone).

My own research with college-students (obviously not a representative sample) also supports the finding that PWOM is more frequent than NWOM (and a similar percentage of these interactions were in offline settings too).

Further, Robert East from Kingston Business School, has found that PWOM is 3x more likely than NWOM (largely because people have more opportunities to spread positive WOM). Robert's research also suggests the counter-intuitive finding that PWOM and NWOM are not all that different from each other in terms of impacting brand decisions, though this varies by category. Read his study for more details (opens into PDF).

A contrasting picture, however, emerges from Pete Blackshaw in a recent blog post and ClickZ article on defensive branding. Pete contends that in online CGM venues negative experiences are far more prevalent than positive ones. This is especially the case in search, where people are seeking out information online and considering a purchase. It seems that for certain product categories, such as automotive and electronics (two categories mentioned in Pete's post), the search results are often dominated by negative comments. Further, if it's indeed the case that many people use search for researching brands before they buy, then these negative comments can have a great deal of reach with, and potentially impact on, other consumers. So, regardless of whether or not all online CGM is more positive or more negative (anyone know of comprehensive research on this matter?), the important point according to Pete is that search results, at least for certain categories, often include more negative comments than positive comments (and these are the comments more likely to be seen by inquiring consumers).

OK, so let's assume the following for the sake of argument: PWOM is more prevalent in offline venues and online CGM-related search results are frequently more negative. What explains this difference? Is there something about people's communication patterns via online and offline media? Is it somehow easier to share PWOM offline? Is it somehow more gratifying to complain online?

Existing academic research by Sundaram et al. (1998) in Advances in Consumer Research shows that there are four primary motivations for spreading negative WOM: 1) altruism (to help ensure others don't get burned); 2) anxiety-reduction (telling someone else about the negative experience allows the other person to validate the person's reaction as reasonable and appropriate and gets the issue off our chest); 3) advice-seeking (where one person has a negative experience and seeks the aid of another to help them decide how to respond); and 4) vengeance (wanting to get back at the company).

All of these motivations could also apply to CGM/online WOM. And research by Bailey (2004) in the Journal of Marketing Communications certainly suggests that the internet does contribute to negative consumer-to-consumer articulations because it provides a forum where angry customers can make complaints and have a receptive and global audience to validate their negative experience. These articulations invite "that-happened-to-me-too" responses from others and these "me-too" comments may in turn raise the rankings of such posts in search engine rankings (this is speculation on my part but it seems plausible). Coincidentally, see an exchange between my students about bad cell phone service on our class blog and others responding with "me-too" comments (this example includes trashing one service provider while praising another.)

Plus don't forget that WOM articulations communicate more than just information about the product, but also about our own identities (this is what I like and don't like and this is the type of person I am and am not). The internet is a huge forum for this kind of identity-expression (or perhaps more accurately, identity-construction) -- just think MySpace and other social networking and community forums -- and on a societal level we seem to increasingly identify ourselves by our consumption and brand-related experiences.

And further, if you add in the fact that most companies don't do a good job of providing responsive and engaging outlets for people to complain it makes sense that consumers will take things into their own hands and perhaps express themselves where they'll get the largest and most receptive audience (which leads Pete to argue, and I agree wholehearedly, that effective WOM/CGM programs need to include the consumer affairs folks as well as the marketing folks).

But then if you grant the difference between online and offline WOM regarding an issue like positive versus negative sentiment wouldn't it then follow that online WOM isn't necessarily representative of offline WOM and vice versa? Or maybe it's that online search results aren't representative of offline WOM? I'm anxious to see research that looks at CGM across a range of different categories to determine the relative prevalence of comments with positive versus negative sentiment. Is anyone aware of research that does this?

And if it turns out that online WOM does skew more negative than offline WOM then it seems that online isn't representative of offline WOM as many claim. An implication of this, then, is that if companies are only paying attention to monitoring online WOM (which seems extremely important to do and we'd be foolish not to pay serious attention to it), and if consumers are primarily coming across negative articulations in search results, then nobody is getting a full picture of consumer-to-consumer WOM about brands just by looking online.


Tuesday, May 16, 2006

TalkTrack and the Conversational Geography of WOM Project

Advertising Age reports yesterday on some exciting findings from The Keller Fay Group's TalkTrack methodology (disclosure: consulting relationship).

Actually before I go on, does anyone notice how there's almost always a WOM-related story in AdAge on Mondays when the print version comes out? It's like WOMMA's version of WOM Wednesdays except it's Advertising Age's "WOM Mondays" (doesn't have quite the same alliteration value though [er, actually consonance since alliteration only technically refers to the use of a vowel at the beginning of a word]). It's great to see a WOM-related story almost every week nowadays in this publication.

Back to TalkTrack... one of the reasons I'm very excited by this is because of the parallels with my academic work that falls under the heading of the conversational geography of WOM project. Briefly the argument that I have made is that effective WOM communication is based less in the marketing of a particular brand and more in the everyday conversations in relational networks, since it is those everyday conversations and relationships that form the basis for the influence characteristic of WOM episodes. Thus, in order to understand WOM successfully, you have to take into account A) a mapping of the everyday conversations, and B) the relational histories of the people talking.

Methodologically speaking, you can accomplish A and B through a combination of different methods. To track the conversations you can use 1) a diary-based method, which I've done with the Word-of-Mouth Episode Survey (and what TalkTrack is doing for brand-related interactions on an ongoing basis with a nationally-representative sample), as well as 2) recording actual conversations and WOM episodes in order to transcribe them and analyze them for how WOM episodes emerge on a sequential and inferential basis (using insights from conversation and discourse analysis).

But you also have to account for another kind of history beyond the conversational history that plays out turn-by-turn in a specific episode, and that's the relational history of the people talking. For this, you can use 3) a "relational ethnography" approach (see Carl & Duck, in press; contact me for a pre-press version). What you do here is ask people to record their everyday conversations, which will include WOM episodes, transcribe these interactions, and then interview the participants about what they see going on from their "insider/participant" perspective (this insider perspective will supplement the "outsider/analyst" perspective). Specifically you want to ask the participants in the WOM episodes about how they make sense of their own interaction and to provide insights that an outside analyst wouldn't have access to since s/he is not an "insider" to the relationship. Here, then, you're gaining insights into how the relational connection and history serves as a basis of influence in the WOM episode.

Now, from my perspective, TalkTrack is extremely valuable because it will give insight into the first of the two points I mentioned: the mapping of everyday conversations. And there are many practical applications of this, some of the most important from a commercial perspective, arguably, are the ability to see trends in the conversations, to get a sense of what percentage of the conversations in a particular industry or category a particular brand occupies, insights into polarity, who's making recommendations, where these are taking place, what people are saying, etc. (see my presentation in Hamburg Germany about how this information can be used).

Another benefit is that TalkTrack, or a longitudinal, diary-based approach more generally, can be used as an industry benchmark for how everyday, organic WOM episodes play out. Companies can then compare their outcomes from organized programs to this more general data. Even more specifically, companies that specialize in designing organized WOM marketing programs can then compare data they gather from participants in organized WOM marketing programs to what non-participants (or "everyday people"; the type TalkTrack is tracking) are doing.

I will actually be writing about this for my contribution to WOMMA's Measuring Word-of-Mouth Volume 2 research book (disclosure: WOMMA Advisory Board and work group leader for this book). Specifically I'll be comparing findings gleaned from TalkTrack to findings of WOM marketing agents from my research collaboration with BzzAgent.

The next phase of my academic research agenda will be to utilize the relational ethnography approach for the analysis of WOM episodes.

Stay tuned for the details!


Monday, May 15, 2006

The WOMBAT Strikes Back -- WOMBAT 2 in San Francisco

WOMMA's Word-of-Mouth Basic Training Conference is coming back for more. WOMBAT 2 will be in San Francisco June 20-21, 2006.

I won't be able to make this one unfortunately as I'm teaching my summer class on WOM, buzz, and viral marketing communication (and missing two class periods in a row during the summer term is like missing a full week in a regular semester). I'm bummed that I'll have to miss the keynote speech by Shel Israel and Robert Scoble, authors of Naked Conversations :-(

But hopefully there will be some good blog coverage of the event that we can all follow!

Disclosure: Advisory Board Member of WOMMA

Friday, May 05, 2006

Inviting My Research Blog Readers To Participate In My Class on Word-of-Mouth, Buzz, and Viral Marketing Communication

I would love to extend an invitation to readers of my research blog to also check out my teaching blog. I am gearing up for my new special topics course on Word-of-Mouth, Buzz, and Viral Marketing Communication and want you to participate via the blog.

I have been building this course from scratch over the past few weeks and am very excited about it. We will be learning about the academic research on word-of-mouth marketing and communication and also learn from some of the industry leaders who are building this field. We have a number of wonderful guest speakers joining us and we will be collaborating with some great companies in the space as part of a class project (visit the class blog to learn more!)

Students will also be blogging about the concepts and their experiences in the course and I hope to do a few podcasts of lectures as well.

So, if you're one of the 74 people out there that Feedburner tells me are subscribing to this feed (thanks so much by the way!) or if you're one of the 450 unique visitors per month that MapStats tells me about, then please add my teaching blog feed to your readers today!


Thursday, May 04, 2006

Mapping the Conversational Geography of Word-of-Mouth in China

The conversational geography of word-of-mouth project is traveling to China!

A few months ago I was contacted by Paul Clark, an Assistant Professor at Okanagan College in British Columbia. In addition to his work in Canada, Paul teaches classes in Shanghai, China and is the Director for Shanghai Direct, a consulting firm that helps companies wanting to enter the Chinese market. He has followed my research on the mapping of everyday word-of-mouth conversations and asked if I was interested in collaborating on a small study with his students at Shanghai University.

After discussions about the best way to integrate the study into his Advertising and Promotions course we decided to have the students (n=70) use the Total Interactions and Word-of-Mouth Episode (TIWOME) worksheet to track the percentage of total conversations that include brand-related talk for a 7-day period. It will be fascinating to compare the results with the students from the U.S. population.

His students begin tracking their conversations next week and we hope to have it collected and analyzed within a couple weeks after that. Stay tuned for details!

My goal is to track the conversations all around the world using the diary-based TIWOME and Word-of-Mouth Episode Survey methodologies so if there are other professors who want to collaborate on similar projects please let me know.