User Segmentation through Field Research

Through field research, the user's behavior is observed in the real user's environment.

A lot can be discovered by simple observation or interview.

It discovers the problems the user is experiencing, the problems that the users are not aware of, but that are slightly different but have the potential to cause problems, and the problems that will be encountered soon. Or, if you add something, you find things that can make the user more efficient and effective. The book, IDEO's Pleasant Innovation, shows it better than the papers.

It is also possible to understand the user and the user's behavior for a certain phenomenon.

By the way, can we find out the user's behavior type through field research?

Usually, in order to do behavior-based user segmentation, quantitative segmentation is performed through a survey, and then, differences between segmentations and similarities between segmentations are found through field research. The most important thing is to understand how the people in each segmentation behave specifically, and to find out what the problems and needs are.

However, since a structured questionnaire is used when conducting a survey, the questionnaire is initially created with some hypotheses. To create a questionnaire, you need to understand users and their behavior. Therefore, in my opinion, if there are no existing theories or research results, there is a way to perform qualitative segmentation through field research and then quantitatively survey. So, even if we find behavioral patterns through the field research we are doing this time, we cannot be sure that they are representative, so we thought that we could use them as basic data for a survey later.

So, can we assertively divide user behavior types through field research?

The word assertive is used to mean that the whole is so. In other words, can it have representativeness in terms of quantitative research?

We searched for papers that divided user types through field research. There were a lot of field research papers at HCI 2005 this year, so I looked them up.

There were three papers that classified user types through field research, whether they were diaries, in-depth interviews, visiting interviews, or circumstantial observations.

Although it is stated that field research was conducted in the paper, there was no method or causal relationship on how the results were analyzed. Since it is not a methodological paper, it is difficult to digest all of this in Chapter 6, so it seems that space was devoted to writing the findings.

However, no one mentioned the validity of whether the usage behavior divided through the field research was the real usage behavior. It's about "real" usage behavior that I use well when talking about grunge in the population. Quantitative researchers will raise the issue of sampling while talking about representativeness.

I was looking for an ACM thesis, but I didn't have enough time, so I just started thinking with my head.

As I was thinking about what to do, I remembered the scene where the renowned physicist, Reed Feynman, gave advice on a number of questions over the phone.

So, I dared to call and ask Professor Park Ji-soo, who took a lot of anthropological approaches.

In fact, before going to the UK, Professor Jisoo Park attended a workshop held at the HCI Society and greeted him as he passed by, and it was only the first time he greeted him properly at the HCI 2005 held in Daegu. I wanted to hear the results of Professor Jisoo Park's research in England, but I was sad that I couldn't. He was in the hallway and asked for an explanation, and he gladly took out his laptop and explained it. And I asked the questions I usually had, went into another field research session, told me about my thesis on ethics, and gave me homework (?) to do research on ethical issues in field research.

Professor Jisoo Park is a person who has taken a lot of anthropological approaches to actual product development, so I think he thought that he would be familiar with various situations. Again, he had a lot of experience, and the professor explained in detail.

While talking about success stories through field research, I started asking questions in earnest.

1. Can you uncover user behavior patterns through field research?

you said you can I mentioned again the representativeness question I asked in Daegu in February, and Professor Park Ji-soo said it could be done. I still have regrets about representativeness, so I thought that I should think more about whether it is possible to bring out representative results even with qualitative research. Or whether it is necessary to talk about representation from the beginning. (I've been thinking about this issue, so I'll write a separate post)

2. How can we extract user behavior?

Look at your diary or interview, jot down actions and cluster them. This is a bottom-up approach. Then, after looking at the things tied upside down, you come up with another group and approach it again from the top-down.

After hearing the explanation from the professor, I began to be convinced of the method. I think it's good to call.

Before I called Professor Jisoo Park, I was thinking about looking at my diary and drawing an affinity diagram to reveal user behavior patterns. I was thinking about whether I could use the affinity diagram used when brainstorming or when dividing findings by type after usability test, when looking at a diary and drawing out behavior patterns. My problem was how far should I look at the basics of behavior.

I hypothesized the user behavior process in the domain, and reviewed and supplemented it with a domain expert. And I coded it according to the action process I drew in the diary. Then I defined the basic types of behavior in the way I thought and the way my professor suggested, and started looking for patterns.

Again, it's a headache. You look at the user's profile, understand the user's basic background, extract the main actions, and then look at the diagram again, one by one, and put the whole thing in your head. Then we extract the major issues again and put them into the analysis table. Then, we find that the behavior is repeated again, and even if it appears once, we select the behavior that is repeated with other informants. Write this on a post-it note and stick it on the wall.

However, as there were more than 5 people, groupings started to form a little, and when there were more than 10 people, a post-it with the representative’s name on the behavior type started to stick.

The reason I didn't think of this from the beginning is because when I said, "Some people use this way," it was assumed that each person has only one usage behavior.

But, on the way home from work yesterday, does a person have only one personality? There are also multiple personalities! Is there a law that says that a person has only one behavior? I thought that one person could have several behaviors, depending on the purpose or circumstances.

It may be said that it is one in a big way, but I thought that if I went down a little, one person would show several behaviors. In fact, I was showing different behaviors depending on the situation in the diary, but yesterday I tried to find a behavioral pattern that could define this person in one word.

Does one person have only one usage pattern, or does it have different patterns depending on the situation? To what extent can you say that it is one usage pattern per person? This on its own could be considered a dissertation. Maybe someone studied it. Of course, since my work is to develop a product, not research to write a thesis, I cannot depend on the value of research.

After thinking about this, I think I decided to analyze each specific behavior of the information provider.

How to deal with a visit interview is before a specific protocol has been developed, so you need to think about how to link the diary study with the visit interview. There are several on-site interview methods, so you have to choose.

While conducting field research, I have never made a type for each usage behavior through field research. I still don't know how to prove it, but I have the idea of ​​going to the persona as a final product.


0 If you like the article, please click the heart~ It will be a strength to bloggers (SNS/login/advertising is not related)

Articles you might like

Contact map: customer management on the map
Contacts at a glance on the map, location-based contact management, view nearby contacts


Add a Comment

Email addresses are not disclosed. Required items *is indicated by

This posting is part of Coupang Partners' activities, and a certain amount of commission is provided accordingly.