Affinity mapping is taking every single thing your users said, putting each one on its own sticky note, and grouping them until themes appear. You can do it alone, and you can get a lot of good insights doing it with a group.

It is the step between having interviews and knowing what they told you.

Start with verbatim statements

One statement per sticky, in the person's own words, not your summary of them.

The rule for splitting them is that each sticky holds one idea. "I don't usually keep track after the baby is 14 months or so" is its own sticky. "Always put it in writing when I give medication, for safety reasons" is its own thing. That is not related to the food they ate, so it does not belong on the same note.

Keep the duplicates. When I did research I would see the same comment come up a lot, and I would grab a few of them, so that I could show on my big canvas that it is not just one person, that there are multiple people, and that they talk about it differently. Some will be really angry about it, some will be like, it is okay. I want all of those.

Do yours before you give it to AI

If you gave all these statements to an AI and asked it to summarise them into themes, it would give you themes, and they would be correct. That is the trap, not the objection.

The problem is what it does to you. Once those themes exist, if you were not involved in making them, there is a good chance you stop seeing anything else. You have blinders on and that is all you are going to see.

So do yours first. Then, if you want, give the AI the same ungrouped list and compare. Your insights come from your own experience and your own way of looking at the world. An AI is working from what is essentially the common denominator, which is worth seeing second, not first.

The value of this exercise is not the list of themes at the end. It is what happens in your head while you sort.

How to actually do it

Work on as big a screen as possible, so you have room to read the stickies and room to place them. I would work on a big monitor.

Start on a random canvas rather than in a tidy structure. Unless you already have the experience, use a canvas format so you can just be random. I still need the random canvas myself. If you are the kind of person who can keep organising this in your mind, you can go straight to organising.

Then read. Just start grabbing.

Here is what that sounds like in practice. What do you use for photo sharing, Tiny Beans. Put that one over there by itself. Look at the next one. Here is another one about photos, put those together, since they are both talking about photos.

That is the whole mechanic. Read one, find where it goes, and if it does not go anywhere yet, leave it alone. Groups form out of the pile rather than being decided in advance.

Then step back and look for trends

Once you have everything up and you are looking over it for insights, you are thinking. You are not really writing anything down yet.

You are looking for high level trends. There are issues around communication and people remembering details, that is a trend. Another one is scope creep in the job. Another is transportation. High, high, high level.

Expect a few of them to split. A lot of the time what you thought was one problem turns out to be several once you really dig into it, and the pile is where you find that out rather than in the interview.

Those trends are what become your problem statements, so it is worth sitting in this stage longer than feels comfortable.

Running it with a group

In the old days you would do this on a whiteboard with real sticky notes, and people would walk up and move them.

The important part is the start. Before anyone does anything at all, give people 15 or 30 minutes, depending on how many stickies there are, to review in silence. They are getting the overall sense that you already have, because you are the one who collected them.

Skip that and the first person to speak sets the themes for everybody.

A note on tools and on skipping it

These sticky notes get really long, and depending on what kind of person you are that can stress you out. It bothers me, because I like things organised, geometric, lined up, so for this activity I prefer Mural.

And if you have a small enough set of research that you can hold it all in your head, you do not have to do this. Do it if it makes sense. I needed to do it to make sense of all of mine.

Keep reading

What Is a Persona and How Do I Build One From Real Research?

How to Write a Problem Statement for Your App

How to Talk to Users When You're a Developer and Find It Deeply Uncomfortable

Learn this with me

You learn affinity mapping in class 2 of From Passion to Product, my free six-week live course, on the statements you gather in your online research, and you do it again in class 4 on your own interview notes. Twice, on two different piles, because the second one is the one that is hard.

It is a group exercise in class 2, on real statements rather than a practice set. One rule I teach with it: do not use AI for the grouping stage.

The App Strategy Workbook covers the whole course and every step in it has the exercises and the templates, so you can do this now instead of waiting.

Join the waitlist to hear when the next free cohort opens, or start with the workbook.