I'm sure this ‘research’ goes to help Bzzagent sell word of mouth mouth marketing on the basis of (conversational) reach, which is a big plus seeing that they have teamed up with Group M in the UK 
Just out of curiosity did you use a control group this time as the self-reporting model is hugely susceptible to bias (e.g. Hawthorn Effect), particularly when agents are directed and incentivised to spread the word. Obviously, your two person measurement methodology provides some safety check, but even here I would guess that reliability depends on sample sizes, who was chosen and how, etc … and whether this research could actually ever be replicated.
What intrigues me here is that your 'research' seems to be suggesting that relay rate of Generation 2 is some kind of constant regardless of other variables, such as who instigated the conversation; the relationship of the recipient to the person who instigated the conversation; the product category being talked about; etc.
This would suggest that having a one night stand with Kate Moss is the same as having a night in front of the adult channel … in terms of the experience for the recipient and therefore their likelihood to tell anyone about it (or not) … as well as the likelihood of the recipients to spread the word based on whether they knew the person telling them or whether they cared.
Guess this means that Dave and those that mimic his model round the globe can claim that your research now PROVES (rather than your initial research provides interesting findings for further investigation) that not only do ‘agents’ have on average 15 initial conversations (as shown by self reporting), but at the second round there’s another 3-5 conversations generated (as confirmed by two person measurement methodology).
So now even an 18 year old at a media planning/buying companies can sell word of mouth marketing by saying that if you buy a campaign with 1,000 agents then the campaign CAN potentially generate (although I bet they’ll say WILL) 45,000 to 75,000 conversations.
For any 18 year olds at media planning/buying companies that’s 1000 agents times 15 initial conversations times 3 to 5 second round conversations.
Don’t forget to charge a high CPM rate by claiming that conversations are x times more effective than an impression, even though everyone knows that a picture paints a thousand words
Maybe it’s just me but I thought word of mouth is about the quality of the conversations not the quantity!
Hi Justin,
Thanks for your comment. I'd like to take the opportunity to reply in some detail.
I'm not sure of the reason for the scare quotes around "research" (it strikes me as a bit presumptuous really). Additional measurement surrounding word of mouth is something that a lot of companies are asking for these days, so I imagine that any company who implements the G2X WOM Tracking Methodology (brand clients and agencies alike) would benefit. In fact, one of the goals is for this methodology is to become an industry-standard, and to have a third-party firm conduct the WOM measurement. This is precisely the purpose of ChatThreads, and the new company is already working with different companies on behalf of their clients to provide independent, third-party validated measurement services for word of mouth across media channels. To learn more you can visit ChatThreads.com Discussing individual company's business strategies is probably best done in a separate venue.
This is an excellent question about a control group. No, a control group was not used in this initial study and this is something that future will work will address. And you're precisely right that one of the reasons I used the dyadic approach was to correct for biases of self-report, and in fact it can work quite well to address some of the limitations. The full manuscript will discuss this in more detail. And yes, sample size and replicability is extremely important (as well as response rates and accounting for non-response bias). In fact, the study is currently being replicated now with programs from a broader range of product and service categories.
The G2 relay rate should NOT be interpreted as a constant across all situations and campaigns. This would be an oversimplification. This is a point made in the PDF version of the executive preview which is why I asked people to download it (I'm not sure if you've had a chance to read it yet). Based on the research I have done G2 relay rates depend on a number of factors, including campaign design and execution, product category, the participants involved, etc. And this is why it's important to use the methodology with each campaign and not rely on an aggregate measure.
While I wouldn't venture to guess what a one night stand with Kate Moss would be like, we are in 100% agreement about the variability of the G2 relay rate. And this is what's so exciting, I think, about this methodology. That it can provide a reliable measure across category, company business models, and media channels. In fact, the methodology is designed to work with any brand experience that has "talk value" -- that is, that will generate WOM conversations. So, this methodology is not just for "word of mouth" companies, though this is where it is emerging from, but it is designed to track and measure word of mouth across a range of experiences. Further, it can be used to track how word of mouth spreads between online, offline, and digital venues as well, which should provide a lot of really interesting insights.
"Proves" is a very strong word when it comes to social science and market research, so I think it's important to use that word with a lot of caution. Actually, one of the great things about this methodology is that it can be used to do third-party validation of campaign results, including results collected through self-report data. This is a win for brand clients who work with WOM media platforms, agencies, and service providers, as well as the agencies themselves because they can have their results validated by a third-party as is done in other industries.
To your point about 18 year-olds at media planning firms... Caveat emptor. Any company who would buy from an 18-year old who just throws numbers around in the way you describe probably deserves what they get (nothing against 18-year olds; I don't want to be age-ist). But it's important to interrogate the methodology and assumptions of any research.
You're absolutely right about your point about quality of the conversations. The methodology I developed can provide some very interesting perspectives into the dynamics of word of mouth, both quantitatively and qualitatively. For example, by understanding the locations where people talk, the activities in which they engage, the relationship between the parties involved, other media forms referenced, the content of the conversations, etc. companies can better understand the qualitative aspects of WOM phenomena and generate actionable insights to optimize their programs and their product or service experience.
In addition to quality, it's also about the impact of the conversations. This will be developed more in the full manuscript but the methodology also tracks inquiry, use/trial, and purchase intentions and behavior.
I look forward to when the paper is released so you can get a fuller picture of what this methodology can, and can't, do.
Walter
Maybe scare quotes on ‘research’ are presumptuous. Nonetheless I think it’s fair to point out that your ongoing work in this area appears to have supported a number of aspects of the Bzzagent-style business model, or maybe more accurately has been re-framed to support this type of business model.
I think you need to be more categorical in what you say CAN and CANNOT be inferred by your work rather than have me come and challenge you and those that make inferences that are not supported by your findings.
Personally I find the work of Keiningham, Watts et al more interesting because they challenge the status quo rather than simply support the more anecdotal and ad hoc claims of practitioners, and in Duncan Watts case even ‘gurus’ like Gladwell.
Not only do they have a lot more to lose by challenging the status quo as far as their reputations are concerned, but also do more to take things forward by providing some antithesis.
But well done for setting up your research consultancy, and I genuinely mean this … although it will be interesting to see how you get on against the likes of Millward Brown, Ipsos et al, who are also active in this space.
However, forgive me if I can’t help drawing certain conclusion as a result of this news seeing that you didn’t support me when I asked you to intervene when I was being slandered and financially penalised for questioning WOMMA’s position on disclosure.
I’m just not sure how the stifling of debate can be considered to be either academic nor ethical (however you attempt to reframe what happened here), but that’s for your conscience to reconcile.
Maybe we could revisit the debate about disclosure (perhaps with Emanuel Rosen who has been researching this as part of his new book), although I’m pretty certain that both you and Emanuel will do your utmost not to engage me on this one publicly. Shame because I’m a great believer in dialectic and freedom of speech.
Anyway, I’m glad that in future your work will address my comments about using control groups and I hope it will be addressed by using them. Why not start using control groups on your current investigations into programs from a broader range of product and service categories?
I’m also glad you are stating that G2 relay rate should NOT be interpreted as a constant across all situations and campaigns. I agree it would be an oversimplification, but that hasn’t stopped people who were given a sneak peak of this study making the inference. Seeing that we in 100% agreement about the variability of the G2 relay rate then I’d recommend you state this more categorically given my comments about the inferences already being made by those that have been given your initial finding in advance.
I 100% agree with you that when you say the following:
"Proves" is a very strong word when it comes to social science and market research, so I think it's important to use that word with a lot of caution.
That’s why it’s a bit surprising to see (Dr) Paul Marsden who co-authored the Connected Marketing book with me claim that his ‘Advocacy Drives Growth’ article in Brand Strategy magazine is “The paper that proves the economics of buzz” (see here). I’m sure Tim Keiningham and other academics would strongly disagree with Paul’s claim.
My point about 18 year-olds at media planning firms is really a criticism of using reach as a measure of success for word of mouth marketing campaigns, which I believe will only serve to help commoditise this industry … just in the same way that reach based metrics have commoditised other marketing approaches such as digital advertising and viral marketing by driving down price and therefore the quality of executions. As such, there are some potential unintended consequences of your study that are rather worrying.
Lastly, I’m glad you think I’m right about the quality of the conversations being more important than the quality, and whether it’s your methodology or a more independent one that ends up being used as an industry standard to assess value here remains to be seen. However, I also look forward to when the paper is released so that I can get a fuller picture of what this methodology can, and can't, do.
Congratulations on the release, Walter - now I know why I haven't seen many updates to the blog...
Thanks David!
I came to your blog just when I was surfing on this topic. I am happy that I found your blog and information I wanted.While people may have different views still good things should always be appreciated. Yours is a nice blog. Liked it!!!
Commenting by HaloScan
