Monday, May 19, 2008

New Academic Research on WOM: The Role of Disclosure in Organized Word-of-Mouth Marketing Programs

As a follow-up to my earlier post, I am excited to announce the release of my latest journal publication on word-of-mouth marketing communication. This piece was published in the Journal of Marketing Communications titled "The Role of Disclosure in Organized Word-of-Mouth Marketing Programs". If this title sounds like a familiar topic for me you'd be right as it is the the academic version of my industry research report "To Tell Or Not To Tell?: Assessing the Practical Benefits of Disclosure for Word-of-Mouth Marketing Agents and Their Conversational Partners”" published in January 2006. At the time the research received a lot of attention in mainstream and social media being cited in Advertising Age, The Atlantic Monthly, The Boston Globe, and Business Week, among others.

The publication of the academic journal article also comes out at a fortuitous time and venue. By timing, I'm speaking of the latest legal and public policies coming out of the UK concerning the issues of transparency and disclosure in advertising and marketing practices By venue, I mean that the editorial board of the journal has a heavy contingent of UK and European-based scholars.

(By the way, see the Word of Mouth Marketing Association's position on this important topic. It turns out this new UK legislation was not targeted at organized word of mouth marketing programs in particular but it cast a much wider net in terms of unfair marketing and advertising practices in an effort to protect consumers from deception and other unethical practices).

OK, so to the content of the article. Here's the abstract:

Prevailing views of organized word-of-mouth (WOM) marketing programs suggest that disclosing corporate affiliation reduces perceived credibility and hampers campaign effectiveness. To test this view we surveyed WOM marketing agents and their conversational partners (CP) after a WOM marketing episode. Results indicate that when disclosure occurred – defined as when the CP was aware they were talking with a person participating in an organized WOM marketing program – agents were rated as more credible, CPs had fewer negative feelings about the agent’s corporate affiliation, and CPs told more people about the brand being discussed. These counter-intuitive results can be explained in part by the existing personal relationship between the agent and CP and invite us to consider how personal relationships may moderate the impact and potential business advantages of disclosure in organized WOM marketing programs.
If you read the To Tell or Not To Tell? report I STRONGLY encourage you to read the academic version of the article because it goes into a lot more detail and nuance about the research, the results, and the limitations.

The main difference is that the academic venue afforded more of an opportunity to underscore the importance that the underlying relationship plays in explaining the counter-intuitive results of the research. In fact the results that support the business case for disclosure -- higher perception of source credibility, higher relay rates when disclosure occurs (meaning more people were subsequently told about the product), and minimizing the potential for "backlash" if a person doesn't disclose but then the person they're talking with later finds out they were part of an organized marketing program) -- can be explained in large part by the pre-existing relationship between the people talking. The act of disclosure played a role in explaining the phenomena but not as much as the type of relationship between the people talking.

This implies that the same results may not be as salient when you're talking with someone you don't know very well, or at all, and disclose that you're participating in a marketing program. In these situations, because people may not know anything else about you as a person and your motives, then this disclosure may indeed diminish the program participant's credibility and diminish the perceived sincerity and effectiveness of the recommendation. And this is precisely the reason why there are consumer protections in place, and that's because non-disclosure may indeed make a difference in how people perceive the brand-related communication, even if there aren't any spurious motives (as in many things, perception is the reality).

But if you have an existing relationship (for example, friend, family member, co-worker) the act of disclosure is welcomed or a non-issue because this bit of information is contextualized by the history of all the other interactions the people have shared.

I believe there are a lot of significant implications to this research and underscores the importance of understanding the many contextual features that affect how people interpret each others' communication.

If you think you'd find this article interesting you can download it from the publisher's website here or a pre-press version from my download page (but if you're going to cite the paper be sure to cite the published version).


New Academic Research on WOM: The Conversational Geography of Word-of-Mouth Communication and Marketing Practices

We've all heard about the tortoise and the hare and how the tortoise eventually, slowly but surely, arrives at the finish line, before the speedy hare. Well, two tortoises have finally crossed the finish line representing academic publication for some of my research on word of mouth marketing communication practices. It was a long journey but they have finally arrived.

These two pieces represent a significant milestone in my academic publication record because they are the first two pieces where I did not use a colon in the title. :-)

Actually they are important for other reasons which I'll explain briefly below. Actually, I'm going to split these into two posts and discuss each article separately.

The first piece that was just published appears in Communication Quarterly and is titled "The Conversational Geography of Word-of-Mouth Communication and Marketing Practices". Here's the abstract:

This study was a test of the utility of a diary-based methodology for revealing how word-of-mouth (WOM) marketing agents perceive their campaign and non-campaign-related WOM communication episodes. A modified version of the Iowa Communication Record, originally designed for presenting the “geography of everyday conversation,” was the base for the collection of 2,088 self-reports of the agents' WOM episodes. Data were subjected to a principal components factor analysis. The resulting factors—communication quality, value, impact, relational change, and conversational control—served to gauge differences attributable to the institutional nature of the WOM, sex of respondent and conversational partner, relationship type, and day of week.
This piece is actually "Part 2" of my "What's All the Buzz About?" research published in Management Communication Quarterly. There are (at least) three really interesting results in this article.

The first interesting result was that agents perceived a higher amount of change in their relationship after a campaign-related word of mouth conversation when compared to their non-campaign conversation. Subsequent analysis revealed that the direction of this relational change was positive; that is, participants in the agent-based programs reported feeling closer to their conversational partner after the conversation where they were discussing a campaign-related product. Since agents generally perceived the WOM episodes to be of high quality and of some value to their life it is not surprising to see some levels of greater relational closeness occurring after the brand-related conversation. But another possible explanation is that agents felt like they were helping their conversational partner by providing them with relevant information which led them to feel closer to the other person. But more research needs to be done to explain this finding.

The second interesting result was how participants in the agent-based marketing program perceived that they were more likely than their conversational partners to initiate interaction, to decide topics, and to end the conversation for products that they were talking about as part of the organized program when compared to brand discussions that weren't part of a marketing campaign. This makes a lot of sense since it's the program participant who has access to information about the latest and greatest products, but it could also be problematic because a hallmark of everyday conversation is a sense of "mutual conversational control"; that is, where the parties perceive they are contributing equally to the conversation. I have unpublished research (an even slower tortoise!) that shows that their conversational partners also perceive that the program participants exercise more "conversational control" during campaign-based conversations, but the jury is still out the implications of this (for example, if their partners perceive the agents exercise more "conversational control" during the interaction does this negatively affect perceptions of credibility or is this an understood norm in relationships when one person may have greater knowledge or experience about a topic than the other).

The third finding I wanted to highlight concerned the relationship between the program participant and their conversational partner(s). In comparison to weak-tie or acquaintance relationships, program participants perceived their WOM conversations with strong-tie relationships (best friends, romantic partners/spouses, and relatives) to have higher conversational quality, more value to their present and future life, and more of an impact on their attitudes, feelings, and/or behavior. This is important for two reasons: 1) the higher perceptions of quality, value, and impact might partially explain existing findings showing that people are more likely to engage in WOM conversations with strong-tie relationships. Second, the results pose a challenge for organized attempts to stimulate WOM in social networks. Existing research by Dr. David Godes and Dr. Dina Mayzlin has shown how firms need to stimulate WOM in weak-tie relationships to generate incremental WOM that spreads beyond the WOM that has already occurred within a social network. But it is these weak-tie interactions that agents perceive as having less communication quality and value, which as stated above, may be one reason why they occur less frequently. If one goal of a WOM campaign is to get the word out about a brand, product, or service to as many people as possible (or at least as many people who will find it relevant), and if weak-tie relationships are important to that process, then it will be necessary to create opportunities where people find the interactions with weak-ties more rewarding.

If you think you'd find this article interesting you can download it from the publisher's website here or a pre-press version from my download page (but if you're going to cite the paper be sure to cite the published version).

And stay tuned for my other article that was just published!


Thursday, May 08, 2008

Introducing the Conversation Value™ Model and Preview of WOMM-U Sessions

As mentioned in an earlier post I'm speaking at WOMM-U today and tomorrow about WOM measurement and how to scale WOM programs. On the WOMM-U website WOMMA bills my session as "Your CMO says: 'If I gave you X dollars more for WOM, what would I get?'"*

The purpose of this post is to give a preview of how I plan to address this topic and it coincides with some exciting news with ChatThreads regarding the announcement of a new WOM ROI metric. We call it "Conversation Value™" and it's a measure of the revenue impact of consumer conversations.

I've been working on this model with Dr. Barak Libai, who has been a visiting professor of marketing at MIT and a professor at Tel Aviv's Faculty of Mangement (and also a fellow member of WOMMA's Academic Advisory Board).

The Conversation Value™ Model incorporates consumers' WOM behavior into a life-time value (LTV) model, which allows companies to quantify the bottom-line value of each conversation about a brand.

The two categories of inputs are LTV-related and WOM-related. (We feel that if you calculate the value of a customer just based on a LTV model that you actually underestimate the value a customer brings to the firm because it doesn't take into account their WOM.)

For LTV inputs we calculate the average value a customer brings to the firm through their own purchase behavior, adjusted based on a discount rate that takes into account the time value of money (money now is worth more than money later).

The WOM inputs are collected via ChatThreads' analytics platform. The inputs include generational relay rates (the number of people told from Generation 0 to Generation 1, Generation 1 to Generation 2, etc.), which is a measure of reach, as well as generational purchase rates (the percentage of people who report purchasing the product or service at each generation).

The value of each conversation, or "conversation value", is calculated by combining the life time value and WOM referral value and dividing this by the number of conversations with unique people. The "net present conversation value" is computed by subtracting the costs for the marketing initiative from the conversation value figure.

You end up with a dollar amount, like $1.20, for example, and this number means that each time a person had a conversation with a new person as part of a marketing initiative (whether it's an advocacy or influencer WOM program, or a more traditional event marketing or sampling program) the company made $1.20. The value could be a negative number as well which means the initiative failed to generate a positive ROI. Companies can track this number over time and work to optimize their initiatives in order to increase their conversation value.

Interestingly, conversation values can range dramatically by cateogry and by each of the inputs to the model, such as the profitability of each unit sold and generational relay and purchase rates.

Something else that's really cool is how you can run simulations to understand what happens if, for example, each program participant reached just one more person at each generation, or purchased X% more. Or what happens if you decided to scale the program larger (for example, by engaging additional program participants). Additionally, you can use the model to look at cost per conversion so that you can compare an initiative designed to generate WOM to other media channels. This will give media planners, buyers, and marketers a clearer sense of how to allocate their resources.

We are developing Conversation Value™ Models for ChatThreads' clients and will be releasing some research reports and white papers in the near future. For the time being you can come to my interactive sessions in Miami. Hope to see you here!

* BTW, the hypothetical CMO's statement above may make it seem like you can just buy WOM or have a successful WOM program by just throwing around some additional money. I don't think this is what the conference organizers intended when they wrote their short blurb, but I'll certainly dispel this notion when I discuss it today. What they were getting at, I think, is understanding the impact of resource allocation and how to go about it in thoughtful and strategic ways (or, this is how CMOs probably should be talking about it anyway!).