In an era where customer focus is almost the only rule of marketing, “big data” has come to be discussed more and more as command of customer data has increased and consumer behaviour analysis and market research have been considered together. Companies collect data from every point they can reach and run all kinds of analysis in order to predict customer behaviour and increase sales. But there is also a phrase that never leaves the lips of many company executives: “we have a lot of data, but we don't know what to do with it”.
The situation in market research is not very different. Sometimes we ask participants to answer pages of questions in hour-long surveys and end up with a mountain of data. Sometimes we run seemingly simple studies with 15-minute questionnaires. Market research firms prepare reports for clients based on the data they obtain. They build analysis tables from field data and try to make sense of it using different charts so their clients can understand it more easily.
For market research firms, documents in PowerPoint format running to hundreds of pages containing only such tables and charts are adequate only up to a point. Turning the data obtained into insight is far more vital than presentations of this kind. So what does turning data into insight mean? How does the insight everyone is chasing emerge? Without going too deep, I would like to share my thoughts on how the data held by research firms and/or their clients can turn into insight.
Let us proceed with an example from the lives of our possible ancestors, a very long time ago. Imagine a man in the Stone Age walking with his 5- or 6-year-old child when a dinosaur appears in front of them. The man, moving quickly, escapes; the child, unable to run, is killed by the dinosaur. Based on the data he has, the man probably concluded that dinosaurs are bad and that he should stay away from them.
Now imagine that six or seven years later, the same man is walking with his second child when the child dies from an illness that could not be observed from the outside — and that at exactly that moment a bird is passing overhead. Based on the data he has, the father will conclude that the bird flying overhead is as dangerous as the dinosaur. From that moment on, he will feel the need to hide whenever he sees a dinosaur and whenever a bird flies over him. His belief that both situations are bad, and the effort he spends to prevent the consequences, will only diminish once his observations of birds and dinosaurs increase and he sees that birds are not in fact that bad.
Using this example, I would like to touch on the differences between the concepts of data, information and insight. The harm caused by the dinosaur and the bird can be considered data. When Stone Age man uses that data to decide which animal to flee from, he turns the data into information. If the inference he draws leads him to decide he must flee from both, he has not read the data correctly. More accurately, because he has not made enough observations, he has had trouble turning data into information. Therefore accurate and sufficient data is the most critical element of information. In market research too, studies not conducted with an adequate sample size can provide false information, however beautifully formatted the presentation.
Turning data into information requires more than sufficient observation. You need to be aware of the existence of different dynamics and to read the environment (the market) you are in well. In other words, cause-and-effect relationships must be established well and the sources of those causes explored. Put differently, this is the step at which it becomes possible to decide which data is accurate and reliable, and to see how that decision will affect our lives. It is what makes it possible to use limited resources for the right purposes and reach the desired objectives. Returning to the example, knowing that birds do not cause death and that the real killer was a health problem will also change the actions taken.
At the final stage, we need to turn the data we hold into insight. At this point it is necessary to understand which actions the information we have obtained will lead to, and to reveal the underlying reasons. No information that cannot be turned into action should be treated as insight. Because insight is seeing that the measures to be taken against dinosaurs differ from those against birds, deciding who the main enemy is, setting priorities and then seeing what should be done and when.
Knowing that a marathon is 42km 195m and that runners lose an average of 3 litres of water is data. That runners should be provided with water during that marathon is information derived from the data. But giving runners mineral-rich fluids at the 11th kilometre is insight — all the available information blended together.
Another example: through its fixed internet service, Türk Telekom reaches more than 6 million points (homes and workplaces) in Türkiye via wired or wireless modems. Power cuts in any neighbourhood can therefore be detected even before the electricity distribution companies notice them, because it is possible to establish that all the modems in a neighbourhood have lost function at the same moment. With this data it is possible to determine at what time power cuts occurred in which districts, inform the distribution companies, and — by measuring the trends in these outages — provide input to prioritisation work for infrastructure improvements.
So if the formula for turning data into insight is clear, why is insight obtained in only some research projects? Why do some research reports not contain even a single line of recommended action?
There may be many answers to these questions. But I believe the underlying reason is that qualified human resource is under-used. The technical quality of the research designed to obtain the data to be interpreted (sample size, sample distribution, the type of questions asked and other factors affecting data quality), and the shortcomings in turning the resulting data into information (analysis quality, the cross-tabulations taken, the statistical models used and so on), can leave you at a point where obtaining insight was never possible in the first place. And it is people who decide what should be done and how at every stage of that process. Decisions taken by people who know their craft are what guarantee the insight obtained when the research is complete.
Market research companies, unfortunately, operate within a vicious circle today. On one side there is the spending required to produce quality work (the fees paid to qualified project managers, quality fieldwork agencies and competent analysts); on the other there are the ever-growing numbers of procurement officers (people who decide on research purchasing instead of researchers or marketing managers). Managers of market research firms spend every day trying to meet the demands of people whose expertise is negotiation and who want more for less, while also trying to hit internal revenue and profitability targets. That equation resolves in only one way: through low-cost human resource and automated analysis and reporting systems.
Researchers who can interpret market structure correctly, think about what the data they obtain means, and offer insight with recommended actions naturally command higher salaries. And research firms paying those sums find it hard to submit cheap proposals. This is why many valuable research projects end up reported in a form where only two numbers in the report get looked at — and client companies lose their faith in research.
My suggestion for escaping this spiral is for research firms and the professional association to deliver comprehensive training, together or separately. Training in which experienced researchers explain with examples how insight can be extracted from real data, and explain to less experienced colleagues why this matters. It would also be appropriate for clients to build a decision tree so that, when choosing which research firm to work with, they look not only at price but at firms that employ qualified, capable staff. To bring these two elements together, a certification programme could be created under the leadership of the professional association: just as there is a quality assessment system for research firms, those attending the training could receive a similar certificate. Stating in proposals that at least one or two people holding this certificate will work on each specific project would guide pricing for both the research firm and the client. Procurement units would also gain a data point they could use as an objective assessment criterion.
At the end of the day, what is valuable and what contributes to business results is data that has been turned into insight. And what extracts the right insight from data is qualified people.