Monday, September 19, 2005

Defining Word-of-Mouth: Nominal, Descriptive, and Evaluative Distinctions

A colleague from the Stockholm School of Economics, Sara Rosengren, recently commented in an e-mail exchange on how I define word-of-mouth in my research.

I am also a bit curious about your definition of WOM. Regarding everyday WOM you use "infomal, evaluative communication (positive or negative) about an organization, brand, product, or service" as a definition. I was wondering whether WOM actually must be evaluative? Couldn't it just as well be only informational (e.g., "there is a sale at Sears")? Russel Belk has written about this a long, long time ago (1971). He then makes a distinction between what he calls "nominal" (just a mention), "informational" (descriptive information), and "evaluative" WOM. I kind of agree that all these types are actually WOM, but often only the "evaluative" part is considered in research. It would be very interesting to hear your thoughts on this issue.

Here is my response to her excellent question:

Here's my take: I think the "informational" versus "evaluative" distinction is a tough line to demarcate from an analyst's perspective. That is, the view of interaction I adopt is a rhetorical view which implies that there is no such thing as "informational" talk; that is, talk is always selective and partial (and thus it can't be "just neutral" or "just informational", if not for the very reason that someone chooses to mention one thing rather than something else or not at all). However, we (as participants in conversations) make a practical distinction between talk that "comes across" as more or less informational or more or less persuasive. For the "WOM Communication Log" survey [used in a research study published in my 2006 Management Communication Quarterly article], there is a question about valence which would allow people to indicate for themselves whether the commentary was negative, neutral, or positive (about 20% of the episodes were reported as involving neutral valence).

The "nominal" distinction is a worthwhile distinction and is especially relevant to the "Total Interactions and WOM Episode" worksheet (the 7-day log where people recorded their number of interactions with people and the percentage of those conversations that included brand-related talk). From initial analysis of actual instances of transcribed WOM episodes we learned that "nominal" mentions are quite frequent, for example in the process of storytelling ("we were eating at McDonald’s when I saw..."). This is a nominal reference, but McDonald's features as scenic element of the conversation rather than as an object of the conversation. To help participants determine which "mentions" to count for the study I decided to only count it as a WOM episode if it was more than just a nominal mention. Of course, I suppose I could have had them count it as a WOM episode if the organization, brand, product, or service was an object in the conversation but this seemed like it would be too confusing. By the way, one thing I'm looking at with the transcribed interactions is how do some "nominal" mentions lead into more extended conversation about the WOM object. (I'm just in the early stages of this).

So, in short, I think Belk's category system can be meaningful and useful (given the point about a rhetorical view of talk) and I should have mentioned it in my article. Perhaps I'll do so in a future article. Thanks for raising this issue!

For those interested, the Russell Belk article discussed in this post is:

"Occurrence of Word of Mouth Buyer Behavior as a Function of Situation and Advertising Stimuli," Proceedings of the American Marketing Association, ed. Fred Allvine, 1971, 419-422.


Creative Commons License

This work is licensed under a Creative Commons License.