Interview mit Stephan Shakespeare (YouGov)
Stephan Shakespeare ist Gründer und CEO von YouGov.
marktforschung.dossier: Mr. Shakespeare, YouGov was founded in 2000 and is considered one of the pionieers of online market research. What have been the most significant changings in market research since then?
Stephan Shakespeare: The biggest changes are online surveys and digitized data. When we started out in May 2000, internet surveys were already starting to be used by leading edge clients. But at that time it was considered to be ‘quick and dirty’. YouGov was the first to create an online methodology that was more accurate than any form of traditional research. If it could be accurate, then the other advantages were obvious – faster, cheaper, and the ability to ask more difficult questions. The use of high-response panels added the potential to layer data over time. Then of course there was the passively-collected data – the digital revolution means that we can locate people through their telephones, follow their in-store purchases through loyalty cards, and monitor their online media usage. Suddenly, we realized that traditional forms of research, although still highly relevant, were only a part of a much bigger field of behavioral and opinion research. This is the change that is still gathering pace, and which will change our industry a lot more in the next decade. Simply reproducing traditional surveys online will look just as old-fashioned and clumsy as those old telephone surveys.
marktforschung.dossier: BigData, mobile, social media – what do you think will be the main challenges for international market research in the next ten years regarding methodoloy on the one hand as well as economic questions on the other hand?
Stephan Shakespeare: We have to realize what conventional research is good for, and what big data is good for, and when they can connect. I spoke recently to the CEO of one our largest supermarket stores and he was almost dismissive of the so-called ‘insights’ of survey research compared to the purchase-point usefulness of loyalty-card data. The first makes broad claims of significance, but the second helps make hundreds of concrete decisions which optimize sales. I could see his point of view, but I think he missed something important: the classic survey methodology is still a vital experimental format that is better suited for predicting some types of future trends. Our challenge in market research is to recognize what social and digitized media is good for and what more traditional formats serve better – and where we can get greater value from the combination.
Market researchers feel threatened by big data and they push it away with ridiculous arguments about ‘raw noise’. Of course all data needs to be sifted and structured. I hate it when market researchers start talking broadly about ‘insight’ – let’s face it, clients don’t really believe our claims to provide brilliant insights: we know that because they don’t pay much of a premium for it. When someone claims ‘insight’, take a second look and you’ll always find a fairly obvious conclusion about quite standard data. It’s vanity to talk about insight. Real insight is a very rare event – Darwin had it, Einstein had it, I doubt that I have it. It’s quite brilliant enough to talk about practical usefulness instead – practical usefulness in designing strategies, media campaigns, monitoring and campaign adaptation. That allows us to find the value of the old and the new, and even to bring them together.
marktforschung.dossier: How about Google’s aims to run tools such as „Google Consumer Surveys“? Will companies and agencies originally focused on internetbusiness really harm the market research industry in the near future?
Stephan Shakespeare: The investment of $800m in Survey Monkey, a big chunk of it from Google, is a watershed moment for our industry. That puts it at about the same valuation as TNS – one of the global giants of our industry - when it was bought by WPP. And yet, Survey Monkey occupies the single floor above YouGov’s silicon-valley office in Palo Alto! I believe we will shake out into three segments. At one end you will have self-service and sample providers. That will be fully technologized and commoditized. At the other end you will have what I call ‘designed data streams’ – not raw data but data that is often derived from continuous survey systems (like YouGov’s BrandIndex) or from passive sources (like social media monitoring) which can be rich and complex and high-quality. In the middle, you have conventional market research, which can contribute to both the others. Conventional market research will remain a decent business, but it cannot be a growth business.
marktforschung.dossier: What is your opinon about online panels as we know them today – will they still be there for conducting market research in 2020?
Stephan Shakespeare: Online panels, especially if the panelists are high-response (but not over-used) and stay engaged for many years, can provide data that is richer and more reliable than any other source. But building and maintaining such panels is not easy, and is becoming ever harder, as there are so many competitors for attention. Why should someone give up their time to answer the usually very boring questions of market researchers? Why should they make the effort to answer them with real consideration? A small prize will not be enough. I think it will be necessary to find new ways of making people want to work with researchers. There must be some meaningful pay-back for the respondent. That is one of our biggest efforts at YouGov – we are investing in new ways to make panelists feel they should be part of our company. And it’s one of the reasons we make such an effort to provide interesting stories for the media: people like to feel their effort serves the public good in some way. If we do it right, we will help the industry. The fact is, traditional methods have meant that we have not had to show regard for the quality of the interview – if the person has been randomly dialed, why should the researcher care about the respondent’s experience? You will never talk to the respondent again, so the exercise becomes to squeeze as much ‘response’ out as possible before the end. But on a panel you have to care, because you have invested in the panelist. If they don’t like it, they will stop helping. Panel-based research will end up improving the respondent experience.
marktforschung.dossier: Finally one personal question: as far as I know you have done quite different things in the past such as being school-teacher or running the internet television channel 18 Doughty Street. What fascinates you most about market research?
Stephan Shakespeare: The intellectual growth area of today is understanding how the mind works. How is opinion developed and shared? How and why do we communicate, and what is the effect? How do the connections between minds produce collective trends and decisions? Everything I have done has been in some way connected to these questions. Market research is – or should be – at the heart of this theme. Surveys, often simple, even ad-hoc in the classroom, have contributed vitally to new concepts such as ‘prospect theory’ or ‘choice blindness’ in behavioural psychology. What we do is at the centre of how we understand and design our world. I love being in market research.
marktforschung.dossier: Mr. Shakespeare, thank you very much for this interview!