A Word on WoM Metrics

futurelab default header

 by: Stefan Kolle

        As I pointed out in my previous post, in my opinion we are measuring the wrong things in marketing, and many of us are still living in a delusional world of measurability. Let me take this discussion into the realm of Word of Mouth (WoM).

        Right now, I have no indication whether someone visiting my webpage, or viewing my TV ad is responding positively to it – the occasional clickthrough aside. I don’t even know if he or she is actually watching, or the TV is just on in the background, or the webpage that she spends 10 minutes on is actually in the background to a Youtube video. And perhaps most importantly – I don’t know in how far it actually matters that THIS particular person is watching my ad right now, instead of sitting outside with her friends sharing experiences about brands they interacted with. Which one would be more impactful? Let me state the obvious – active WoM, i.e. someone talking about me to their friends, co-workers of family, has a far higher impact on behaviour then any disembodied advertising.

        And now add the next layer – does this person MATTER? In other words, is she an influential? Will she indeed be talking about her positive and negative experiences – and if she does, will others listen? There is an interesting corollary between the propensity of people to share their experiences and how many people they know. So IF she is likely to talk, she will even do so to ever more people.

        Let’s add yet another layer. WoM and loyalty research (including NPS) tends to focus on actual buyers of a product. But this leaves out some important factors, such as whether someone voluntarily bought or not. In other words – I bought that car only because my company has a corporate agreement with this brand. I have no personal affection with it. How will this influence my emotional stance towards this brand? The other way around – I am the biggest fan of this brand – but I can’t afford it (or am not yet of car-buying age – let’s not forget the power of 16 year old boys in the car selling process). So I’m a major advocate for the brand – but because I’m not actually a buyer you don’t count me?

        If I try to add all these elements into an overview, it looks rather crowded – especially given my limited graphic talents J But take a look at this arrangement I made of (totally fictional) data. If I knew all these additional data, I could start fine-tuning my WoM efforts in a very powerful way.


        For instance, for this fictional brand, of the voluntary buyers, most fall in the positive, but non-influential zone. Let’s start looking why this is the case – are they a bit ashamed of their association with the brand? Does it just happen that the psychographic profile of our typical buyer is that of a shy person that is reluctant to speak out about his preferences? Once we know these things, we can look into remedies – enhance pride of ownership, or motivate the shy to speak up.

        In this same example, of the involuntary non-buyers, many turn out to be positive influentials, i.e. even though for some reason they can’t actually buy our brand, they are our biggest fans. So we need to work with them, give them storytelling tools, invite them into our family instead of disregarding them and mistreating them because they don’t fit the actual buyer profile.

Here I also come back to my little setup in my previous article. – by taking the Relevance, Reputation and Attention of a person into account, I get a better feeling for whom I want to focus my attention on.

        And now that we have thought of all these different ways of looking at WoM and influentials, I’m going to add the final insult. The one thing we really can’t measure, and which also undermines such useful metrics as NPS (which I am a fan off for the current lack of anything better) is our tendency, if not to say talent, to adapt our stories and recommendations to the listener. In other words, when I talk to my father about phones, I will promote a completely different brand then when talking to my brother – as I know they have different requirements, preferences and abilities.  So I am at the same time a promoter for Nokia and for HTC.  I might even stay a promoter TOWARDS my father of Nokia while having a troubled relationship with the brand myself (which I don’t, just giving an example, dear Nokia social marketing department). So if I were asked to give my NPS score on Nokia, I might give a negative score – while at the same time actively promoting the brand. Confusing? That’s humanity for you.

Whoever manages to solve this metric deserves the Nobel Prize in Marketing. Obviously only the year after mine….

        And now, coming full circle to the start of my argument – as long as we realize that we don’t really know what to measure yet, let alone HOW to measure it, we can start thinking in new ways. The search for new models and metrics will go on for a while, but in the meantime, let’s work with what we have. Let’s focus on identifying our most likely influentials, and how to reach them. And let’s not get blinded by the question whether we are reaching 22% or 25% of the population. Maybe it’s 25 people we need to reach, not 25%.

So – any brand or agency out there that wants to explore this further? Let’s talk. We might change the world a little for the better.