Entrepreneurial capitalism and natural selection both have one thing in common, as a means of finding opportunities – they are entirely pragmatic in deciding what succeeds and what is killed off. In science, in government, in large firms or control economies – the last two are often surprisingly similar in their decision-making; a policy or a theory is more likely to be adopted when it comes dressed up in a fine logical narrative. Such organisations are therefore constrained in their actions by what appears to make sense. No such constraints apply to entrepreneurs or to evolution. Natural selection and market experimentation only care about what works. If it works, it survives; if it fails it is eliminated – through extinction or bankruptcy.

This may explain why so many entrepreneurs are cognitively odd, even downright thick. Yes, your chances of failure are increased if you ignore conventions and norms – but so are the odds of  success. 

One aspect of capitalism that most offends our sense of fairness is also one of its most important qualities; it rewards people merely for being lucky. Should you stumble on some lucky discovery by a mixture of pig-headedness and good fortune, you will be rewarded regardless. Many good ideas are like aspirin – people find it works without having the faintest idea how or why. 

This unplanned system of experimentation has costs in terms of inefficiency and a high failure rate. The upside is that it gives the world things that would never pass muster if they were first exposed to the steely gaze of reason.

Imagine for a second standing in front of a committee/peer-review group/board of directors and making the case for investment in the following great ideas:

Red Bull: ‘Gentlemen, what we have here is an egregiously expensive drink in a small can which tastes – I think we can all agree – somewhere between weird and disgusting.’

Wikipedia: ‘.... and the best part of all this is that people will write the entire thing for free.’

McDonald’s: ‘...and people will be forced to choose between three or four items.’

No sane person would ever have predicted the success of these duck-billed playtpuses of the consumer world.

The problem that bedevils organisations once they reach a certain size – typically when they start employing market research companies – is that a narrow, conventional logic is the natural mode of thinking for the risk-averse bureaucrat or executive. There is a simple reason for this – you can never get fired for being logical. If your reasoning is sound and unimaginative, even if you fail, you have not been guilty of the sin of subjective thinking and it is unlikely you will attract much blame.

The problem this creates is that logic always gets you to the same place – and exactly the same place as your competitors. The Ogilvy Change mantra is: ‘Test counterintuitive things, because your competitors won’t.’

Both biologists and businessmen understand the need for deviant and oblique strategies. For all the Darwinian narratives of nature as ‘red in tooth and claw’, what is striking about nature is how rarely two species compete directly for exactly the same resources. 

What does this mean for market research? I would propose that, when making recommendations, people desist from ever proposing a single best option. Instead, present several possibilities and rank them on two dimensions – the expected outcome and the variance. As with investments, there is no ‘best’ option: it all depends on the investor’s appetite for risk. 

Generally there is a trade-off between safety and potential gains. Something with a lower risk of failure is unlikely to make you spectacularly rich – like betting on red at roulette. Institutional decision-making – remember ‘No-one ever got fired for buying IBM’ – leans heavily towards the safe, low gain, low variance option, where there is the least chance of blame.

But we should not compound this bias by filtering out the weird and the eccentric – the high risk, high variance option – before it even has a chance. Present it as the risky option, yes, but present it nonetheless. 

Rory Sutherland is vice-chairman, Ogilvy & Mather, UK