Market segmentation research

I have never been involved in market segmentation research, because my duty is to intervene after that people have been identified, after their needs have been defined and after the market research institute has prepared the questionnaire. At that point I come in with my army of telephone interviewers and we begin to go searching the public segment.

But I’ll tell you what could happen. Once the CATI machine (Computer Assisted Telephone Interview) gets started, we realize there is an excessive rate of refusals by the respondents, which means they deny the opportunity to proceed with the interview. There is an upper limit of “abandoning” that is of XX%. Once this threshold gets exceeded, it is clear the sample that the institution has given us does not work properly, it does not match the questionnaire.

I wish to say, as an aside, that the sample is usually provided by the institution that has acquired those phone numbers in absolutely legal terms, but we are more often asked if we have got already a sample of eligible respondents. In that case, yes, we have got a huge database, but we also have a new almost infallible methodology that we add to support. I called it Cube Survey. Our Facebook Manager activates a sponsorship on the blue social media, by basing the targeting on the segment described by the research institute, and we ask to the selected people through a form, to provide us with the phone number and availability to be interviewed. That’s the end of my aside.

It may also happen that during the first interviews, the supervisor – who randomly listens behind the scenes – realizes that the interview does not go well, it has no rhythm, it jams up, and there is just something wrong! If we are asking a 47-year-old woman how satisfied she is with the new Hyundai’s lightened steel shock-absorbers from 0 to 10, she will not be able to give us an answer. If we ask it to a 20-year old van driver, he will give us an evaluation in a few seconds. Sometimes our interviewers – who know the timing and fluidity of a good CATI – indicate us which questions do not work. In these cases, we immediately point the problem out to the institute because it is necessary  to replace or eliminate immediately the question, or if the sample is completely wrong, we need to stop everything and ask for a new sample or use ours with the help of Cube Survey.

Market segmentation research is the basis

That is how this article is called, because, guys, it is really the first step any business needs to take. Even in the Business Model Canvas it is one of the three pillars, together with Value Position and Key Partners. I would like to summarize here, for those who do not know what the types of segmentation are. I’ll group them together into three categories, to simplify.

• Prior
• Usage & Attitude
• Needs

Prior (most used)

For the purposes of market analysis, we make hypotheses based on theories without scientific evidence. For example, let’s assume that women in the age group between 40-50 years don’t know what lightened shock-absorbers are compared to a 20-year old van driver. These hypothesis can however vary in time, because the user who’s not on target could suddenly learn what shock-absorbers are, for example by of an epic fail which is all over the news (remember the Monclair case when the whole world after a few days became expert on goose feathers?) the “prior” segmentation should be regularly monitored. The trouble with this segmentation is that we cannot deepen the needs of the consumer.

Usage & Attitude

Let’s get into the psychological field. In this case, demographic data won’t help at all, what we need here is preference data (scaling) in order to proceed with a cluster analysis. In this way, with telephone interviews, we can understand characteristics and possible developments of a market. Data is collected based on the use of a product type, frequency and use patterns, dissatisfaction, user and non-users opinions, known brands, and so on. This kind of market research is usually used in the “consumer” sector, but it works perfectly even in B2B, and is repeated annually to monitor market trends.

Segmentation Based on Needs

The market can be subdivided according to the needs of the consumer. This kind of segmentation is used to decide whether to create new products or to try to improve and push existing products. Often, this type of research uses joint analysis to separate groups according to functional performance. With the cluster analysis, the goal is to determine trends that are hidden behind product performance.

Knowing which segmentation to use is often critical as the analysis itself , because it is guided by the cost and goals of decision makers.

Which segmentation should be used?

Knowing what kind of segmentation to use means like having a compass of business goals and market research that will be carried out. The mistake of the buyer personas may mean failing completely at company level. That’s why it is crucial, especially for those who are about to open a company or launch a product, to conduct a market research in advance. And if you want to play it safe, choose the CATI method, it’s infallible!

If you need more info about CATI, write me e.armato@iff-international.com

Ennio Armato (Branch Manager Italy)