Recently Unilever announced that they intended to be less enslaved to quantitative market research when innovating and would instead aim to take more risks to be innovative and to learn from experience about what works and what doesn’t.

It would seem they have listened to the growing body of evidence that shows that asking consumers what they think is really not as helpful as had previously been assumed.

As it happens I was talking at an event about innovation recently and, as ever, some of the questions asked percolate around my mind and push me to consider new angles and issues.
In my talk I adapted an idea from Scott Berkun, who wrote the book The Myths of Innovation.

He argues, quite convincingly in my view, that each stage of innovation an innovation process has a possibility of going astray. In a seven stage process, working on the generous assumption that there is a 50% probability of success, the resulting probability of succcessful innovation is slightly under 1%.

In other words, around one idea in one hundred might be reasonably expected to become established as a successful product in the market.

My adaptation of this is to factor in the misleading role that soliciting the views of consumers at many of the stages adds (see Consumer.ology for all the details).

Ironically, the very elements of the innovation process that are included to reduce risk actually add it. In essence the research is just as likely to give the right steer (by accident) or wrong.

By my estimation the probability of innovation when market research is applied in a traditional manner is approximately one in four thousand!

So what is the solution?
In life I generally find that there are three broad solutions to any situation, the same applies here:

1. Carry on with the same approach because it’s familiar. This might seem like a rediculous thing to suggest, but human psychology is such that it is far and away the most likely scenario.
2. Abandon market research. Unilever appears to be advocating this route, at least from their press release they haven’t made any attempt to distinguish between good and bad. My opinion is that this is infinitely preferable to option one, but it’s not the route I would advocate.
3. Understand the degrees of validity of consumer research to distinguish between when it’s worth spending money and when it isn’t. And, just as importantly, when spending money, design research that is as psychologically valid as possible. It may have limitations, but provided those limitations are fully understood at the outset, the potential for the results to be misleading can be factored in to the decision making process.

Some research is unquestionably worth doing: for example, understanding how it is that consumers currently behave in the category in question and profiling the underlying psychological drivers of consumer behaviour.

Similarly, live testing, where it can be accommodated, provides a robust and bias-minimising means of understanding the response to an initiative.

Between these two areas, lab tests and carefully conducted experiments can help inform decisions, although they can be claimed to have absolutely no predictive power in thesmelves (although over time a business may learn that it can extrapolate certain results into its own market with a degree of confidence).

Critically, businesses must recognise where the so-called voice of the consumer shouldn’t get within a country-mile of a decision. Research companies will often claim that investigating consumers’ views is worthwhile because it’s just one element in a decision.

In reality, it’s very rare that a company will have the confidence to ignore research that slams a new concept or discount results that endorses one, it is far better that such erroneous information never muddies the waters of innovation.

Image courtesy: Bart