Affinity Diagrams

The craft of surfacing themes

Dedicating time for you and your team to distill the research collected will produce the insights needed to power the next steps in your product's development.

Updated June 15, 2019

Without effective deadlines, research activity could continue forever, but research alone won’t help determine the actionable steps your organization should take next.

Like competitive analysis, affinity diagrams inject a natural pause in the cycle of research to begin the distillation of what has been discovered and the discussion regarding what should be done.

Affinity diagrams help illuminate themes that would otherwise be trapped in the data collected. Whether on your own or within a team, the practice of breaking the data out into a form that can be easily interacted with forces everyone involved to review the information from a fresh perspective. While this exercise can be conducted alone, it is best experienced as a team as a group dynamic naturally begins the process of building consensus around a problem that should be solved.

If you’ve done the research, all you need to conduct an affinity diagramming session are a few office supplies, space to display your findings, and time.

Getting started

Affinity diagrams are the reason that the lowly post-it note has become the unofficial icon of User Experience design. UX is often visually stereotyped as a group of designers huddled around a post-it note covered wall. While we both know that there is much more to design than this one moment, if you’ve ever visited a startup or tech incubator, you’ve seen this stereotype reinforced as we tend to leave behind wall-after-wall of post-it notes. Not that I’m complaining — I personally respect those little colorful squares as an incredibly valuable indicator of potential success for our projects, companies, and clients.

It’s time to gather your data and anyone who assisted in the process to begin!

Surface observations

The first step in when creating an affinity diagram is the act of transferring information from your various notebooks and spreadsheets into a form that is easier to manipulate.

This is where a massive stack of post-its will come in handy.

On each post-it, write down one observation, fact, or quote that was important. It does not matter if this fact appears multiple times, in fact, you should count on that likelihood. You’ll have a step later for pruning everything back, so just focus on getting all the essential information surfaced.

Observations Source:

A couple of suggestions to make this easier on everyone.

First, if you have access to color post-it’s, try to use them with purpose. So it is easy to track where the information came from, use the colors to indicate which method generated the insight. Yellow for stakeholder interviews, blue for competitor research, etc. If no logical breakdown for color is possible, make sure everyone knows that the post-it color carries no meaning. This will allow people to focus on getting their information on the wall.

Next, use markers/sharpies for writing on the post-its. You want people to be able to read the notes from a few feet away, so be sure to write large enough for people to read from a distance.

Finally, set a time limit — particularly if you are working with a large group. You won’t be posting every observation to the wall, only the items that are of particular importance. A time limit forces people to begin the editing process naturally. As you’ll see in the next step, if someone later discovers they forgot something important, it still has a chance to be discussed.

Concept grouping

Once you’ve finished the initial task of surfacing your observations, it’s time to begin the task of surfacing themes. If you are working with a group, be sure to allow time for the team to review the board before consolidating the information further. You want people not only to feel like they are part of the process but also to take in the scope of the work that has been performed to date. No matter how much time you’ve had to dedicate to data collection, you probably have more than you realized.

Grouping Source:

As the research lead, you or a product manager should lead the next step as you begin grouping the information into common concepts or themes. You can start this process at any place on the board, but make sure you have enough room to begin creating new categories for information.

Start by taking a single post-it and creating a category for it. For instance, if the note says something about ‘forgotten password’ perhaps the group is ‘profile settings’ or ‘security.’ As you’re making these categories, you’ll sometimes receive additional information from the person who created the note — be sure to hear them out and add detail to the post-it as warranted. As more notes are reviewed, you’ll sometimes find new categories and supporting concepts that belong to existing groups. You’ll continue this process of review and grouping until you have worked through all the items posted.

Now that your information has been modified to reflect common groups, open the floor for discussion with your team. Make sure to timebox this portion of the meeting so that you don’t run out of time and be sure to call upon people specifically if you find one or two voices dominating the conversation. As stated in Conducting Meetings, making sure everyone feels welcome requires effort and encouragement.

Create action items

Not every cluster that forms will produce an actionable task that your product team can act on.

A few groupings will point directly at the work of other teams. It is important to share your findings with the teams around you. Other groupings will serve as reminders to provide your project with a moral compass to protect customers from abusive actions witnessed in other platforms. A significant number of clusters will produce clear actions like “must solve for offline usage” or “desires private wishlist,” which map directly to potential user stories.

Action-items Source:

After your action items are created and the group has wrapped up the meeting, it’s time for you to take steps to document the gathering properly. At a minimum, you want a photographic account of the wall of post-its groupings for later review and a document housing the action items that the meeting produced.

This is a great moment in your project because you finally have some sense of what you are going to build — but don’t get trapped in this moment. There two big ways this can still fall flat.

  1. The client or company rejects your recommendations
  2. Unclear or vague action item language robs your observations of impact

Let's start with the first one as it is the far bigger threat.

Just because you observe something doesn't mean that your company or client will welcome your observation. Perhaps your mandate is “improve the quality of the customer experience.” That's a fine observation, but not everything observed is actually something the client will want to act on. Lowering prices is likely to make any customer happy, but if your client serves a high-end market, price alone isn’t expected to make a dramatic difference to their customer regardless of competitive positioning. However, providing tools to better communicate with the customer would be a fantastic way to provide better service.

Beware of guidance like “improve quality,” which is too vague to be actionable. Your goal isn’t to prescribe solutions at this point, but you should provide enough guidance to allow your team to move forward.

Everyone on the team will refer back to this step in the process as the discussions around your project are just beginning. Having your research and decisions memorialized in a form easy to consume builds credibility for the project and makes it easier to onboard new people as the project shifts over time.

Associated Exercises

Name Section
Using data to find the path UX Design
Discovering the scent of a product Project: Fresh Market

Receive daily feedback and weekly meetings with Chris Courtney by signing up for monthly mentorship today!

Sign up today!