Thinking differently about word-of-mouth

The current approach to WOM is to try to stimulate positive WOM while addressing or countering negative WOM. A sort of “accentuate the positive, eliminate the negative and don’t mess with Mr In-Between” strategy.

But what if we could do it a different way?

This idea stems from a conversation I had back in February with Martin Kelly and Andy Cocker of Infectious Media. Since that time I’ve chatted it through a couple of times with various interesting people. It’s not properly thought through yet, but following a chat a couple of weeks ago with Ketchum London’s new Head of Digital, the excellent Fernando Rizo, I’ve decided to put the idea out into the public domain to gauge what (if any) interest there is and whether I should continue to work on it.

“Word of Mouth” is hard to do well

I’ve read lots of word of mouth marketing case studies (there’s a great list over at WOMMA) and it strikes me that WOM is hard to do well for a few reasons. I don’t want to go into these in too much detail, but here are a couple of the structural issues:

  1. Unless I’m a journalist, an A-list blogger or media personality or have some kind of platform, I probably have a very low reach.

    Despite everything pointing towards personal contact being the best impetus for positive word of mouth, most word of mouth campaigns compensate for my low reach by trying to get me to self-service my relationship with the brand and the campaign.

  2. “Viral” distribution just doesn’t work the way most people seem to think it does; and this is particularly true when it comes to WOM.

    While I’m quite likely to tell stories about my personal experience of a brand and fairly likely to tell stories that involve a mutual friend, I’m much less likely to tell stories about other friends’ experience, and not likely at all to tell stories about friends-of-friends.

    Furthermore because of the ‘clumpiness’ of most people’s social graphs, geometric progression (the “I tell two people and they each tell two people and so on” effect) just doesn’t happen.


One of the many reasons that WOM works is a thing called homophily — which roughly translates to “birds of a feather flock together”, or “you can tell a man by the company he keeps.”

I’ve written about examples of this before: for example, my analyses of twittering US Congresspersons and Westminster MPs which showed that one can predict with some reasonable degree of accuracy the political colouration of any given twitter account based on their mutual friends and follows (if you want to know more about the methodology, it’s worth reading Robert Hanneman’s chapter on cliques and subgroups.)

But there’s another side to the homophily coin; the social pressure to conform to the group’s norms.

Why I bought an iPhone

Let’s take a concrete example of this: it often seems to me that everyone I know has an iPhone. I made a conscious decision a few years ago not to buy an iPhone, but I’ve finally succumbed to the pressure.

While it’s true that several people have take time out of their busy schedules to tell me exactly how good the iPhone really is, I’m far more affected by what I perceive as the omnipresence of the iPhone. Everyone, it sometimes seems, has one except me.

Of course, on those occasions when I head down to visit my family in rural Hampshire, I’m reminded of the obvious truth: most people don’t have an iPhone. But that’s not the way the world appears to me.

How about turning it on its head?

What if we could identify people who are under social pressure to buy our products — who are being influenced by what their friends are doing and saying?

Let’s say for example that Apple and O2 (iPhone’s carrier partner in the UK) could work out that a significantly higher proportion than average of my Facebook friends access the service using the iPhone client, and could target me with special offers and rates to push me over the edge.

Instead of looking for WOM influencers, why don’t we look for areas of high potential — and target those people who are likely to be receiving lots of WOM stimuli?


  1. says

    This is some impeccable thinking.

    My first thought is to think about how we can work the Diffusion of Innovation concepts into this, which one would have to do if you were trying to launch a product with this WOM methodology.

  2. chris says

    Difussion of Innovation. Interesting you raise that because has WOM in terms of technology (if we accept that tech uptake is the greatest driver) evolved beyond early adopters at this stage. And if so. Has it yet reached laggards? Michael Tchong from Ubercool talks about how the language of technology drives the language of society but yet to reach the widder society, the laggards whose conversations still sit more in the real world, the pub, the bingo hall, etc. Is that where we need to engage them rather than in the digital world of facebook and twitter And while early adopters drive the sales of technology does the mass market who drive broader consumption really treat digital as a trusted or rather do they even access it as source of wom influence…

  3. says

    Hi Mat – great post. I suppose, just to poke a whole in the premise…would the act of targetting you be scalable (from a marketing budget point of view)? One could argue that you are a ‘moving target’, no? I suppose that you could counter and say that the effort required to ‘tip’ you is less than one would normally deploy to get an influncer to like a brand.

  4. Andres says

    The iPhone example is a bit unusual in that there seems to be an implicit assumption that if I’m surrounded by iPhone users they’ll all have positive things to say about it.

    In my world (a few thousand miles away from rural Hampshire, and probably less saturated with iPhone users) my local WOM would most likely be telling me about the latest Palm phone or Blackberry. It’s a de facto standard in my company (a comms security thing too dull to go into) and while one is free to request another kind of phone, most of the people I know have a shiny new Palm phone.

    And all have conflicting opinions. So:

    1) In the analysis which looks for people “under social pressure to buy product x” the neighbours/influencers currently using “product x” are only likely to have negative opinions where the switching cost of the product is relatively high. If I buy a car or phone that annoys me, I’m more likely to stick with it: and give negative WOM. If I am a current user of dominos pizza, it’s most likely because I’ve settled on that brand and have something positive to say.

    So in the above example, which represents a relatively high switching cost, perhaps 02 should add an additional filter which looks for *former* iPhone users too. If someone has (*gasp*) moved from an iPhone tarrif to another (or has switched from O2 elsewhere) then one might want to flag them as a potential source of negative WOM, muddying the pool surrounding your prospect.

    2) I still think that conflicting WOM makes a target a reasonable prospect because at least they’ll likely have a decent level of awareness -a knowledgebase to plug into after you’ve hit them with your brief message.


Please tell me what you think.