As part of The Basement's discussion about the "To Tell Or Not To Tell?" report, Spike Jones, from Brains on Fire, made the following comment on Matt Galloway's post regarding the the validity of the report to the broader world of word-of-mouth marketing (WOMM):
Hey Matt,I wanted to respond to Spike's comments about the purpose of the study, its funding, and the broader generalizablity or applicability to other business models because these are all important issues in any research project, but especially with academic-industry collaborations. But first I want to thank Spike for being a "fan" and say how excited I am to use the Brains on Fire case study for Rage Against the Haze in my WOM class this summer (to learn about their case study go to the Brain's On Fire website, click "Work" then "WOM" then the first case study in the lower-left-hand corner) :-)
Good post. And, as you probably know by now, I'm not a BzzAgent fan, but I am - on the other hand - a Dr. Carl fan.
Here's my question about the study: Do you think the results would have been different if they didn't use BzzAgents? Something triggers my radar here. The BzzAgent model is a unique one, that much I agree on, but I can't help but think that this study was funded to validate that model instead of looking at the world of WOMM outside of BzzAgent.
Purposes of the Study. There were three purposes for my most recent collaboration with BzzAgent: 1) To test a new set of outcome metrics that are more consistent with the WOMMA Terminology Framework (credibility effect related to other information sources, inquiry, purchase, use, and pass-along; by "new" I mean that my first study didn't include these); 2) knowledge of institutional identity (meaning were the conversational partners aware that they were talking with a person affiliated with a marketing company and did this make a difference in the outcome metrics?); and 3) differences between the everyday WOM behavior of agents/program participants and their institutional (or campaign-related) behavior. (Additional articles or research reports will discuss findings from the other areas beyond disclosure of institutional identity).
I mention these multiple goals because they are informed by a theory I'm developing about organized WOM (more generally), the testing of a specific business model, as well as a test of the WOMMA Terminology Framework (to my knowledge this is the first academic-industry study to explicitly situate the study in the Framework's terms; disclosure: I was on the drafting committee for this Framework). So two things here: 1) yes, one purpose of the study was to test aspects of the BzzAgent business model, but it would be incomplete to just say that; and 2) the sword cuts both ways in terms of validation as part of an academic research project. That is, findings might invalidate the business model as well as validate it. Results get reported whether they are favorable or unfavorable to that model (and the TTONTT report has both).
Funding. Initial funding came from an internal grant I received from Northeastern's Research & Scholarship Development Fund (thanks to the Provost's office), and this covered the cost for the conversational partners' portion of the study. BzzAgent covered all the costs on their end related to the Agent participation. Of course, even if the company funded the full study as part of an academic research project then the same principles of reporting results that would validate or invalidate a particular business model would still apply. Exceptions sometime apply to any proprietary information but this factor gets disclosed as it becomes relevant and doesn't compromise the integrity of the results for this study.
Speaking of funding, I have another parallel study ready to go that seeks to investigate the same outcome metrics and WOM communication dynamics for "everyday people" without any organized WOMM program involved. So far I have only had funding for university students so if any company or foundation out there wants to further our understanding of everyday WOM through a nationally representative sample, just let me know! :-)
Generalizability or Applicability to WOMM. Any study has limitations and the unique business model investigated here is no exception (I discuss some of the more relevant limitations on p. 20 of the report). To validate the findings they need to be replicated with the same business model at a different time, and with different business models (and I'll be the first to acknowledge the value of doing so). Further, the results need to be peer-reviewed as is standard practice with academic work, which is why TTONTT is called a "summary report" and Footnote 1 states that academic journal publication constitutes official reporting of the results. And in specific response to one of Spike's questions: I do think we should expect to see differences in some of the outcome metrics and how program participants are viewed (for example, as existing customers, as category experts, etc.) depending on the business model, so this is another reason why additional models need to be studied.
But here are some ways I think the study does have applicability beyond the specific BzzAgent business model:
- I think the importance of disclosure applies to WOMM firms that have their own ongoing community of participants (like BzzAgent or Tremor or Vocalpoint) as well as those who work with different sets of category influencers for each campaign (like Matchstick or M80 or Brains On Fire) [Addendum, 02/11/2006: Please see Spike Jones' comment to this post about my grouping of these companies]. Further, I would suggest that companies who work with their own existing customers as part of organized programs (without an outside company to help them) would still benefit from campaign disclosure. (I mention some of this in point #10 on p. 18 of the report.) I think disclosure issues may be slightly different based on different models, but this will be the subject of a later post.
- I think the report points to the importance of the existing relationship as a factor in source credibility and perceptions of an ulterior motive. I'll be developing this in an academic article to tie in with research on the Persuasion Knowledge Model and the Social Consequences of Interpersonal Influence Model (opens into 3MB PDF file).
- It also points to the fundamental principles of all WOM, whether it is organic or organized: relevancy, trustworthiness, and goodwill.
- It provides a test of the WOMMA Terminology Framework and contributes new terminology to refer to "Participants." Specifically I think the Terminology Framework needs generic terms to distinguish between those who participate in organized WOMM programs or have an affiliation with a marketing company and those who do not. I propose we use "everyday people" and "program participants" (I used "agents" as a generic term in the TTONTT report but think "program participants" is even more inclusive).
- I think there are some other issues as well but stay tuned for more research on those!
Thanks to Matt and Spike for the opportunity to contribute additional thoughts!
Tags: WOM word of mouth Word-of-Mouth Marketing buzz marketing disclosure