FEATURE3 May 2012
FEATURE3 May 2012
Coke’s Stan Sthanunathan urged the MR industry to embrace technology or get left behind in his keynote speech at The Market Research Technology Event.
But before we go any further, you need to watch the video below. You don’t have to watch all of it, just enough to be amazed by the spectacle of IBM’s Watson computer competing on the game show Jeopardy against two former champions – and winning.
“It is actually quite a scary experience,” says Stan Sthanunathan, Coca-Cola’s vice president of marketing strategy and insights. “Most people will have seen Watson play chess and beat Garry Kasparov, but chess is a mathematical game. It is a game of decision trees and a computer can usually calculate faster than a man. But in this video, Watson is reading thousands and thousands of texts, parsing them, parsing a question and coming up with an answer.”
For Sthanunathan, it’s not much of a stretch to imagine Watson’s capabilities being applied to solving the problem of understanding consumers based on the mass of text-based information they produce.
“IBM is working on computers that are going to be a thousand times faster than Watson. Imagine that,” he says. “In the next five years, we might end up with computers that have the collective intellect of a big city, and then those computers will start developing computers that are smarter than themselves.”
Human researchers can’t hope to compete with the computational abilities of machines. What’s needed, therefore, is a change in skillset and mindset. Sthanunathan explains.
Research: Will researchers be rendered obsolete by these analytical machines?
SS: I don’t think computers can replace humans completely, but they can dramatically enhance the effectiveness of the human being. We still need humans to tell stories. But the role researchers are playing today – in terms of analysing and reporting – might not be there tomorrow. We need to acquire new skills to succeed in the next few years. We need to get more comfortable with higher levels of ambiguity, for instance. It’s also not about statistical significance. Show me one CEO who gets excited about statistical significance. Making logical deductions are what sets humans apart from machines. Machines can’t think and feel – a logical deduction is something unique that humans can provide.
In your Market Research Technology Event keynote you also spoke of the need to swap a ‘conclusive mindset’ for an ‘exploration mindset’. What do you mean by that?
SS: At the moment, we drive towards conclusions with the work we do. If we don’t start expanding people’s thinking then sooner or later there will be machines that will do a fair amount of conclusive thinking for us and we won’t have the skills we need to be effective in this world. Also, researchers see insight as a primary deliverable – that’s the focus of agencies and research people within companies. But we need to move towards insight as a means of driving inspiration and provocation. That’s something a computer can never do. Can you imagine a computer ever replacing a Nelson Mandela or a Mahatma Gandhi?
The ability to inspire – that’s a skill that seems to be impossible to teach. Does that mean we as an industry need to look to bring people in to the business who already have that ability?
SS: Exactly, and that’s why diversity of thinking is something that we all need to focus on. If you surround yourselves with a bunch of people who are exactly like you – who think exactly like you – that drives a very high degree of incestuous behaviour. The more you surround yourselves with people who make you uncomfortable, the better a researcher you will become. I’m not saying that the rigour and discipline we have in the function will go away completely. We need that. If you don’t have that you don’t have any basis on which you can tell stories. So it’s a set of additional skills that we require.
You also spoke about using technology as a driver of research, not just an enabler. Can you explain?
SS: A lot of the time we as researchers focus on how to use technology to collect insights in a better way. Instead we should be thinking about the technologies we have available to us that can help us collect insights that we could never collect before. It’s a slight difference in mindset. You can see this change happening already. It’s all relatively early stage stuff, but there are lots of companies doing lots of interesting things and the sooner they go mainstream the better it will be. Three years ago when I was looking around, it was usually the smaller companies doing this sort of thing, because they had nothing to lose and everything to gain by coming up with a big bold idea. Now I find that a lot of the big agencies are doing interesting stuff: Kantar and Nielsen, for example. This is good for the industry, in my view.
But there are a few end-games that we need to keep in mind, because technology can sometimes be a distraction, especially when you fall in to the trap of constantly chasing after the next new shiny object. We have to focus on a few areas. One is to do with how we use technology to visualise our insights differently so that they make people sit up and ( a) take notice and ( b) register what we have to tell them. The second is to consider how consumers are leveraging technology on a daily basis and how we can leverage that to collect information we could never get before and do away with stupid questions, like asking someone to recall the purchase process they went through a week ago.
Do you subscribe to the idea that we can eventually leave questions behind completely?
SS: Yes, that is one of my predictions. I think that in the not-too-distant future asking questions will become a very small part of the overall industry. There will still be room for it, but it will become a smaller and smaller part of what we do.
Isn’t that the real danger for researchers: not that IBM will come in and take over market research, but that market researchers won’t evolve their offer beyond asking questions to the point where they become sidelined?
SS: Exactly. The technology companies come from a totally different angle. They have no legacies to protect, no business models to take care of. They can cause disruption.