Market research is the compass of the business world, the map to buried treasure, the beacon in the storm of uncertainty. Or is it? Only if you know where you want to go.
Customer focus starts with getting to know and understanding the customer needs, which is often carried out by market research. We’ve already discussed when organisations don’t use the already existing learnings. Yet, the opposite can be a problem, too. When it becomes a practice over time, organisations can run it out of sheer habit. They are used to asking customers so much that they ask them even when it makes limited or no sense.
Chloe was a market researcher in a multinational company that operated in several countries in Europe. She had been running research for more than a decade, so she could see good and bad examples of market research.
Typical situations when market research is misleading:
There are times when the options are filtered to two, so you have to decide which one is the better. Or should you? In some cases you can only have one solution like if you cannot differentiate a message due to system constraints. In other cases, you might run both solutions and it is more about priority or allocating more resources to the better performing one.
Chloe was asked to perform a research on two different messages in a newsletter, which makes no sense. The worst result in terms of finding the right message is 50-50. In this case, you can use both messages. If you can target the message the right way, you will get better conversion. But if you know nothing about your audience, a random delivery will result in an average conversion.
And what if the result of the research is something different, like 60-40 or 75-25? In situations like this, you rarely or never get a 100-0 result, which, instead of you, could definitely decide what to do. But in the case of 75-25, will you use the 75% of your audience and limit your business opportunity by 25%? I don’t think you should.
So what can you get to know from market research? Nothing relevant for the messages, while for the best, you might find something for targeting. Therefore the real question is not what the message should be, but who the target audience for the different messages is. You should create the questionnaire accordingly.
Instead of the research, you can run both messages in a smaller group of customers to see which one is performing better. Then you can use that version and spread it in the entire customer base.
Measuring exact customer behaviour is better than asking them about it.
It is even more disturbing when market research is carried out to make a decision instead of the decision maker.
Imagine an orchestra conductor timidly asking each musician, "What notes should we play next?" The result would be cacophony, not a masterpiece. An orchestra thrives under a conductor's clear vision and direction.
Yet, once Chloe was asked to measure the customers’ opinion about a discount the company was about to terminate. It made no sense. Terminating a discount is obviously a negative act for the customers, you’ll get no smarter when you ask them about it. You could say that ‘Okay, but we can get prepared where it matters’. And I say that preparation is inevitable and you should get prepared for all the possible outcomes, no matter if they have big or small impacts in such a sensitive market move.
We often delude ourselves into thinking that we can know exactly what the customer wants through research. In most cases, several options receive a significant share, meaning that it is not enough to do one thing well, but rather several things well. If we already know the main options, then it is pointless to research further; we should be prepared for all of them, as we will simply reach different customers with each.
Marketing wanted to define segments within the customer base, so Chloe was asked to run a segmentation research. When you do so, it is worth knowing the ups and downs of such a process. The problem was that a huge part of the customers was rather passive in terms of using the service actively and being involved in co-creation. That’s why all Chloe could involve were the active and engaged customers.
When you investigate a specific group, you will get to know that specific group only and you won’t have general learnings. So you must not run any activity for the entire customer base based on that specific group. You will do nothing, but fail because of the horrible misfiring and conversion.
To remember how to think about the results of research, think of the cows in the countryside as the joke goes. Three people are travelling by train in the countryside, a common person, a teacher and a scientist. They notice a black cow in the field. The common person says that every cow in the countryside is black. The teacher says that there is one black cow in the countryside. And the scientist corrects them all that there is at least one black cow with at least one side that is black in the countryside.
When you want to know general things about your market (e.g. mixture, purchase power, volume), you will need to explore it with the right questions that cover all the major and relevant parts. But if you go deeper and zoom in to more specific segments, you will get to know the details of that given part. Both are okay if they are in line with your intention. Just be conscious.
No surprise that when the marketing department got the result of the active customer group and then developed a campaign for the entire customer base, the conversion was extremely bad, they simply wasted a lot of money.
International market research comparisons can be misleading. Consumer preferences, economic factors, and even legal regulations vary greatly between countries. Not to mention the cultural impact on how people interpret questions and scores. Take the 1-5 scale that means one thing in a country where school grades are the same (1-5) and means something else in a country where grades are A-F. Results from one market might not translate to another.
So when Country A has a score of 45 on a scale of 0-100 and Country B has a score of 50, you cannot say that Country B is better in that regard. You have to compare each country to their own performances, like year over year, to evaluate if there is progress.
We’ve seen that in some cases, market research makes no sense. While in other cases, once it made sense, but after a while, it provides nothing new. Running too many surveys for too long can lead you to fatigue where you get nothing relevant any more. Let’s see 2 examples for that:
Chloe had to measure every little customer satisfaction activity, but for what? In the end, the company had a lot of numbers. However, executing similar measurements offers you to find the underlying patterns and connections so that you don’t need to execute a new one for each occasion.
For example, you can learn that if a customer sees the caring activity, they’ll get more satisfied, and those who even use it, they’ll get the most satisfied. After this learning, it doesn’t really matter if the satisfaction score increases by 5 or 6 points. All that matters is how you can make an impact. Knowing that, you can make conclusions based on other, non market research metrics that are already available in your systems, e.g. number of users, because more users mean more satisfaction.
To understand it properly, it’s worth knowing that there are different types of market researches:
The factory operation of market research is when numbers are generated all the time, but real learnings and insights are missing or not revealed any more. Realising it can lead to modifying the design of the research or even cancelling it. For example by its nature, if run properly, there are no huge changes in tracking research, but if there is any without a major change in the operation then it is a fault of data for sure. It means that after understanding the topic thoroughly, you might consider stopping the measurement.
Remember, market research is a compass, not a GPS. So stop imitating customer focus with l’art pour l’art market research and take responsibility for the business decisions.
