the Blink test

I met yesterday with a former colleague who runs a word of mouth agency. To whip up a souffle of research ideas. I was stung by his dismissal of several of my suggestions that they were too dependent on word of mouth online which he assured me only covers 10% of the recommendations that take place. So here's the product of my thoughts on the train journey home – what if I had to design a research product for the other 90%? I shall start with Malcolm Gladwell's Blink – how is it that human beings draw conclusions quite so quickly about those they meet. And combine it with his earlier book the Tipping Point where he posits that for an idea to travel there need to be 3 people per node – a superconnector, a maven and a persuader. So my task would be to devise a research method which would identify these people as soon as possible.

What makes this difficult for research is that it wants to do everything as fast as possible within a single process – people don't work that way. We don't need to evaluate every branch on the tree about whether we could put our weight on it without it breaking. But somehow we arrive at a shortlist so when faced with an immediate dilemma we have a cache of resources we could draw on.

So let me take the simplest form of research – interviewing people leaving a party where there is a lot of networking asking them if they met anybody new, what they talked about and the impression they formed of that person. We would need the topic list. And if Gladwell is right we should be able to assign a score for that person about whether they are experienced (have talked to a lot of people about the topic or could if they wanted to), whether they were authoritative (told me at least one thing I didn't know in such a way that I was impressed about the quality of their knowledge – it wasn't just a random fact they picked up somewhere). And lastly they were practical – a persuasive person isn't somewho who is a super salesperson – they illuminate the problems I have so I am convinced they could help me find a solution. That's how a persuasive Nissan owner could still be helpful to a hapless Toyota owner.  I think we should measure for all 3 dimensions for every individual met but pay careful attention to what is perceived to be the person's dominant ability.  This is not predictive – but it does give an indication about the weight which might be placed on that branch were it needed. Problems don't come by topic  – the accelerator on the Toyota is an obvious one but it might be a Nissan which had problems with the gears. Human beings are good at general search where machines usually fail because they require literal connections.

The final part is that of momentum. Do you walk out of a party convinced that X is a genius who you will definitely call the following morning. Or do you form that conclusion in a week's time? We would need to have a follow up interview to measure if any of the individuals you were impressed with have shifted their position. I would expect some drift – but I would be very interested in why someone has been seen to be much more experienced, authoritative or helpful than a week before. We have a contamination problem of course. The very act of interviewing people leaving a party has forced them to come to a conclusion much more quickly than would normally be the case. The solution would be to recruit people from the same party but only to interview them a week later (with prompts) to see who they remember.

Will that do for a train journey?  Works for me.



Twitter Updates
Photo Albums
Planners drinks night No1 BucharestTV stations JeddahPutna Sihastra Monastery 2011Grindleford Grinch April 2007Willie Williams exhibition

Designed by Matthew Pattman