Create More - Segmentation

Marketers love segmentation. Going back to Michael Porter and Competitive Strategy. Can you target all of them (eg Kelloggs cornflakes target audience - anyone with a mouth)? No? Then target some of them. Then along come the data people who can dice us and slice us in ways that not even the humble customer can understand. Segmentation rules. Or does it? There's a nasty little assumption here that the basic unit of the individual can be allocated and reallocated - and where there's a grey area we can use Bayesian mathematics (what?) and derive a probability score.

There are many reasons for which we value Wendy Gordon. And not the least of these is that she once made the point that people say and do different things in different contexts. She christened these differences needstates. I'll give a very practical example. Mum is walking down the haircare aisle of the supermarket. She is replacing the household shampoo which is nearly empty. Just before she left the house her son shouted down the stairs could she get some shampoo for him - because he refuses to use the family one and keeps threatening to use hers. And she then spots an attractive new product which is making extravagant promises about improving body and shine and moreover is on introductory promotion. She leaves the supermarket with 3 shampoos. So which segment of the shampoo market is she in?

No data culture means that things get very complicated very fast. We can call her the household decision maker which is useful because we can send her direct mail on everything. We can do multiple assignments so now there are at least 3 of her - as purchaser for each segment. But the trouble is that we don't know which needstate she's in and just being able to identify which needstates people have doesn't mean you can switch to the needstate you want them to be in when you reach them.

There is another way. Its called being part of a media audience and it is largely squandered by the media owners themselves bless them in their haste to be noticed by as many people as possible at the most attractive cost per thousand. Effective media channels polarise needstates. You are in one mode when you read the questions page of Glamour and another when you are listening to the radio in the car. Media channels are perfect mood setters and switchers. In fact I would go further to argue that media measurement ought to be targeting needstates rather than individuals. What is the point of aggregating people who find the communication boring or even irritating? If we were to aggregate needstates we could encourage media owners to fine tune their programming/content to make it even more effective. Isn't it bizarre that within a magazine all the pages cost the same to buy ads for even if the needstates they are conjuring are entirely different and have different values, I'm not talking about the financial pages here - though I might but about the horoscopes page. Or the medical problems pages. Why shouldn't you pay a premium for targeting the right needstate? And so much more reliable than the blunt instrument of a name and a postal address.

There's a variation here which would also be worth a look - and that is looking at households as a basic unit and arranging needstates around them. Particularly if purchasing decisions are complex involving more than one family member and frankly everybody is out there buying and not always for themselves. But the principle is the same:

A segmentation initiated by the customer is always more reliable than one which is imposed by a marketer. Pull is more reliable than push.

Pushing it a little further it means that the segmentation managed by the customer is more reliable. Take age for example. What is more important the number of 25 year olds in the target or the number of people who want to be treated as 25 year olds? In the last census a significant proportion of the population registered their religion as Jeddai. To make the most of this information you don't go out and find out if Jeddai is a legitimate religion or not. It must be because that's what they chose to identify themselves as.

Surely this will result in customer data turning into fiction? Well yes. And? The internet has shown that given the medium surfers are happy to adopt different and multiple identities which bear very little objective reality. The question is whether marketers have the confidence to market using information which may be fictional but is how customers choose to portray themselves. The trouble with fiction is that there are usually several versions around. Compared with the objectivity of segmentation where the customer fits into one segment only. The trouble with objectivity is that there is only one version of it. Which may be considerably less than the truth.

So in a nutshell beware of segmentation which is out there and not in here. Prize ways of teasing apart the multiple personas of households and the needstates of individual members. We keep on saying that we are driven by our customers - let the lunatics run the asylum for a change.