Not Insight but Foresight

To celebrate 75 years in the business, we have invited BMRB managing directors and chairmen, past and present, to write about issues concerning the market research industry. The third in this series of articles is by our former chairman, John O'Brien, who worked at BMRB from 1983 until 2004.

I have been growing steadily annoyed at the slow re-branding of market research as Consumer Insight. The image conjured up is one of a prophetic character who, when faced with a mountain of survey research data, can magically extract that gem of knowledge and re-direct the future marketing strategy.

It is a model which is fundamentally flawed. I would like to set out two themes about the current research world that have contributed to this re-branding and then consider the implications.

Theme 1

There is absolutely no truth in the widely held belief that Florence Nightingale invented the Pie Chart. She was a member of the Royal Statistical Society and did use a variant of the chart to great effect in changing the nursing regime for war wounded, but the Pie was invented at least 15 years before she was born. However, there is no doubt that she would have been upset by the way that the Pie Chart has been transformed into probably the least useful method of data illustration known to man.

Despite the fact that it is widely used in the business world to present statistical data, users find it difficult to compare slices within a pie and between pies. Pie Chart “chefs” counter this by adding percentages and labels to the chart segments - in effect over-laying a statistical table on the uninformative graphic.

So why has the pie chart become ubiquitous in the business world? The answer is because it is easy to do with modern software and the “chefs” think it looks good.

The same combination of factors has created the pervasive cult of the PowerPoint presentations full of gimmicks, transitions, charts, colours, fonts….. The presentation authors put their creative energy and focus into making the presentation look good, with little real concern for the message they are trying to put over. In fact, so often there is no message, no clear theme, just a long sequence of flashily illustrated facts. It is the triumph of style over substance.

Why might it be that the message of a piece of research is so hard to find?

Theme 2

Over the past 30 years it has become cheaper and cheaper to carry out market research and the work can be done quicker and quicker. In the early days of survey research, a national survey would take at least 3 months to plan and complete. Every question asked added not just to the cost of interviewing, but also to the substantial cost of data preparation and analysis.

When the final results for every question in the survey had to be added by hand, by a room full of clerical staff using 5 bar gates, and questionnaires had to be separated and counted to give results by sub groups, a lot of though was given in advance to which questions should be asked, why they were asked and how they would be analysed.

Now we can carry out a national web based survey in a day from start to finish. It is easier to add questions into the survey as requested rather than spending time arguing about why each one is there. It is easier to tabulate every thing by everything else knowing that the key results will be produced somewhere in the resulting analyses, rather than spending time arguing about exactly which tabulations are the critical ones to illuminate the research objectives.

So the message is hard to find because it is buried!

And the solution?

Find someone who can make sense of it all and provide insight, someone who can decipher the meaning of the survey?

NO!

John Samuels, perhaps BMRB’s greatest researcher, changed my view on swimming as an exercise when he explained that you must never count the number of laps you have swum or stokes per lap or anything else which occupies the brain in fruitless work. Decide how long to swim and then use the time to think. He explained that it is difficult to think. Even as little as a minute thinking is hard on the brain.

And that is the key insight into our industry problem. Because it is hard to think and easy to get on with a survey, we put it off the thinking for as long as possible - until the survey is done, the data analysed and the mountain of printout is sitting there.

Do the thinking up front. Debate the reason for every question. Divide questions into categories - critical questions, important questions, descriptive or background questions. Keep the later group to a minimum. Reject any questions which do not fall into the other two categories.

For critical and important questions document in advance what you think the results will be (to within 10%). Document in advance what you will do differently if your expectation is confirmed; what you will do if the result is much higher and what you will do if it is much lower.

If you cannot document anything you will do differently, cancel the research and go back to the drawing board with those people who can.

The net result of doing the thinking early and not researching if you don’t know what to do with the results is a massive benefit for market research all round.

Fewer interviews will be conducted and respondents will feel less hassled. Those interviews that do happen will be shorter and more interesting leading to improved respondent experience. Research agencies will not waste their time producing mountains of statistical tables and hours of PowerPoint presentations. All involved will have a clear understanding of what will happen depending on how the results come out.

There will be no need for insight because FORESIGHT will have mapped the paths and set up the contingency plans.

Commenting is not available in this weblog/section entry.