FEATURE1 December 2009

New view on an old landscape

Synovate’s innovation director Jan Hofmeyr says it’s time to start asking difficult questions about long-established research practices.


Research: What’s your opinion of traditional brand tracking?

I think there’s a tremendous amount we could do better. I have been heard to say internally that up to 80% of the information that’s been collected is defective – it’s not necessarily wrong, but it could be better. But because of inertia in the industry it’s very hard to get people to change. We’re supposed to be researchers; we’re supposed to care about truth and new knowledge. As a matter of fact, if you look at the pace of change when it comes to the big spending buckets in our industry, there are many things that people are doing today that we first started doing in the 1980s.

?Research: Tell us about your new approach to brand tracking.

The idea comes from a number of observations. The first is that the quality of survey data drops alarmingly after about 22 minutes because respondents really become uninterested. The second thing is that if you have the classical long interviews, you end up with a very fragmented view of customers. You can’t put it all in one place because you’ve made each questionnaire too long. The third observation is that if you look at conventional tracking, close to 90% of what you collect doesn’t change. So clients are paying a lot of money for stuff they already know. If you can identify redundancy, you can be ruthless about stripping stuff out. The fourth strand of it is that, now that we can look at people making decisions under brain-scanning machines, what we find is that they don’t use all of the things we think they use. I know some very well-known FMCG manufacturers who use 80 attributes to find out what laundry detergent people like. The average person uses about three, because the way the brain works is to find efficient decision rules that will lead you to not make too big a mistake. As long as you can identify the three that are relevant to someone, you can predict what they will buy. Worldwide, the survey metrics connect to market share with a correlation of 0.95. The correlations are so good people don’t believe them.

“It’s very hard to get people to change… we’re supposed to care about truth and new knowledge… there are many things that people are doing today that we first started doing in the 1980s”

So what we have is this convergence of different strands, coupled to a very rigorous research programme that allowed us to strip the redundancy out of measurement and get to more valid measures with fewer questions. You move towards a tracking system that keeps a finger on the pulse, rather than giving a full medical examination all the time. If you then see the pulse rate go up or down, you then initiate a very fast full medical.

What it gives you also is flexibility because you haven’t got this big thing. The thin tracking allows you to stay thin for seven months, and then when something happens, you beef it up and you track daily for three weeks or whatever.

?Research: If there’s so much redundancy, why hasn’t this been done before?

It’s a very courageous thing for a [research] company to do because, if you look at classical tracking, you’re saying, ‘Here’s $200,000 spent on my customer stuff, and here’s $200,000 for my brand equity stuff,’ and so on, and now you’re saying, ‘You know what, we can put it all in one for $200,000.’ What company is willingly going to give up the $400,000 for $200,000?

Then on the clientside, paradoxically, you get resistance for two reasons. First of all, currently the clients are organised into those spending silos, and each one wants to preserve the budget that they’ve been given to spend on their area. Now somebody comes along and says, ‘We can do the two in one.’ They don’t really want to hear that. When we get in at CEO or CFO level, the whole world changes. When we go in to our direct client level, there’s phenomenal resistance. If you’ve been tracking a squiggly line for 15 years, you’d be surprised at how resistant people are to giving it up.

?Research: So it’s crucial to make sure clients really buy into this.

Yes. People love the idea of innovation but they hate the reality of it. We’re faced with the classic innovator’s dilemma which is that you’ve got to retrain everybody, you’ve got to be prepared to change your mind, you’ve then got to retrain your clients, it costs money, you’ve got to have a marketing campaign that rolls it out, you’ve got to know what you’re going to do with all that other old stuff.

There’s internal and external inertia that makes it difficult. But if you look at the market research world, you don’t need that many
big companies to think like this to change things and to grow. It’s do-able, but there are significant barriers that need to be overcome.

?”Medical doctors will go to professional conferences in order to learn up until the week before they retire. We don’t do that. We go to professional conferences to get drunk and find out what the competitors are doing”

?Research: Why is now the time for thin tracking? Has it been driven by the recession?

I think it’s a lucky confluence of events. I’ve always been obsessed by trying to get the measurement down, and I think this is a good opportunity to become more aggressive because of the downturn. Accountants are just putting pens through whole chunks of research spending. We’ve got an opportunity to take advantage of what’s happening. It plays to our strength, I would say. Apart from the way the recession is driving efficiency on the survey side, you’ve also got clients wanting to explore the potential for these [new ways] to provide better information more cheaply. The survey side of spending, I just can’t see it being maintained. As a piece of the pie it’s going to get smaller, no question, and as an absolute amount, it could be smaller too. I just don’t see any other outcome.

?Research: So it’s about making your work more integrated and more nimble.

Yes. I also think the job will be more fun. These big trackers are just boring – the respondents would certainly be pleased to see the back of them. The one virtue of them is that on the clientside when you’ve got a big fat tracker you can have cookie cutter outputs. You really can pitch up at work each morning, make sure the tracker’s in field and the charts are on the web, and then go and have coffee for the rest of the day. I think there’s a tremendous amount of laziness in our industry - people don’t want their lives to
be made complicated. I have sympathy with that, but it makes for a boring job.

?Research: How do you rate innovation in the research industry?

Our industry is one of the most conservative you will come across. Look at IT – look how keen they are to throw out the old and adopt the new. Medical doctors will go to professional conferences in order to learn up until the week before they retire. We don’t do that. We go to professional conferences to get drunk and find out what the competitors are doing. A constant refrain in our businesses is, ‘That’s too complicated.’ Do you ever hear an engineer saying, ‘That’s too complicated, I’m not going to build it’? Or an accountant saying, ‘That spreadsheet’s too complicated, do it again’? You don’t.

?Research: Aren’t there good reasons why people in the research industry are so attached to their established ways of working?

The mindset is that what defines a knowledge worker is a commitment to standards and robustness. You assume that your industry will have squeezed mediocrity out and arrived at something that really is robust and can stand the test of time. I just don’t see it like that. In a survey, you’re building models of the mind that are supposed to relate to what happens in the real world. Whatever measures you have should have a direct counterpart in reality, and if you discover that they don’t, if you care about the science of what you’re doing you work very hard to try and bring it closer to reality. When you are shown a clear counterexample to your way of thinking, then you ought to be willing to modify your thinking. That is, for me, what defines a research orientation and a scientific orientation. At this point you have to separate the ideals of science from human beings. There’s one fantastic quote from one of the 1960s science historians, who said that new theories only gain acceptance when those who hold the old theories die.

Take any one of the big five – they all have hundreds of millions of dollars worth of business wrapped up in old stuff. What do they do? How do they change? There’s corporate inertia that’s offset against better theories and better measures – that’s the tension.

?Research: Is the problem that it’s difficult to prove whether research really ‘worked’ or not?

With neuroscience, it’s getting much closer to where we can connect the so-called intangibles to really hard metrics. You can now ask questions that measure whether the brain is crackling. But what do people do when they’re confronted with these things? They say, ‘This is fascinating, it proves that brands are real, and that we’ve always been measuring them in the right way. Most people are looking at all this new stuff and saying, ‘And therefore we were right.’ There is a tremendous amount of reassurance that’s currently having to be done to keep all the 1990s stuff ticking over.

?Research: Where do you see innovation in research at the moment?

Small companies that don’t have legacy revenues to defend. I think the cut [in research spending] has actually not been a cut, it’s been a redistribution. The big old companies are not getting it right, so buyers start to explore the potential of these other sorts of organisations to deliver relevant information.

?Research: Would you say that the risk of not changing is greater than the risk of change?

The important thing for a big company is to get the timing right. If you look at corporate histories, companies do go out of existence, and those that go out of existence are those that refuse to change. But the really great companies get the timing right, like IBM with PCs. When they see that it’s becoming a serious issue, in 1982, they go in strongly and aggressively and they pull the whole industry around. That’s going to be the test. We, I think, are a little bit out front, but a TNS or an Ipsos might just get the timing right.

Interview by Robert Bain


15 years ago

Keep at it Jannie! There are lots of people who agree with you. One additional factor that affects innovation is an "innovation bureaucracy" that now affects the large providers. Now careers are at stake so anything that conflicts with the view of the innovation bureaucrats is cast aside as it can call into question their usefulness / competence if they are not the sole source of ideas and products. Sad really. Anyway, cheers from Ken in Toronto old friend. Keep up the good fight.

Like Report

15 years ago

"Our industry is one of the most conservative you will come across. " I couldn't agrree more with Jan.

Like Report

15 years ago

... is the most conservative... because our juniors are train in rules and do the work within the rules, as it is the only right way to do the 'true' reserach. Thinking outside rules is presumed to be 'something dirty' becuase we are always asked to prove things. To do any inovation, you need 3 things - experience, creative ideas and risk guts - the reserach industry seem to missing the latter 2.

Like Report

15 years ago

The other myth about tracking studies is that they offer a continuous finger on the pulse. Yet imagine the data blips downward this week - and we tell the client. But is it just a blip? So we we aggregate the enxt 2 - 4 weeks of data. No, it's not a blip. Okay then, let's arrange a meeting about this issue. (usual probs occur - half the team are away, the meeting room is fully booked etc, the agency can't make it) so then a meeting is held 2 weeks hence. Okay now - 6 weeks have elapsed since the blip was reported. So is this we finger on the pulse exercise, or is the whole process more glacial than we imagine? Tracking studies can get over this through fatter samples, more focused sample universe etc - but I've not heard many researchers express any interest in upsetting their surprisingly glacial cash cows. Perhaps innovation comes - as this article seems to frame things - only when our sacred cows are put under the budegeting axe. What a shame we aren't more proactive.

Like Report

15 years ago

My personal favourite research "best practice" is when agency staff deliberately repeat the same mistakes year after year in order to ensure that the tracker results do not change! When the mindset is that the tracker results should not change, and that when they do it is more than likely to be a blip, what is the real purpose of collecting such data? Surely, it is nothing more than to reassure the client research team that they are doing their jobs correctly whilst making money for the agency team to be lazy in their thinking. My personal mantra for the research world is to be insight-oriented and to think about what the results mean, rather than process-oriented, collecting data in the same way as before.

Like Report

11 years ago

Holy crap. This is from 2010... Amazing how when you look back in time the issues remain the same. Jan, your my hero... This article has solidified my mind to join the revolution of change!

Like Report