Blog

I Have an Affinity for Unbound Thinking Using The KJ Method

What is your go-to brainstorming technique when your aim is to generate some rough ideas that can potentially be built into something valuable?

As a young student, I was taught to web as a visual technique for brainstorming. The web is a form of mind mapping, which is any technique that helps us visualize ideas during brainstorming sessions. Beginning with a topic, the flow adds contributing ideas as a first pass, then adds supporting ideas or topics in subsequent passes.

Discovering the next great idea can be tough. Let’s take a look at the KJ Method for mind mapping.

Relating to the affinity map

The affinity map or affinity diagram is a mind mapping technique to reach consensus, originally based on Jiro Kawakita’s KJ Method applied in Total Quality Management, and is one of the Seven Management and Planning Tools.

The KJ Method partitions a brainstorming session into generic phases of mind mapping and group consensus. Collectively, participants can quickly isolate patterns and concentrate on high priority themes, while lower priority themes can be elaborated when and if necessary at a later time.

A basic mind map through the KJ Method is very effective when you need to get some unbound ideas in play to select the next great product or feature idea. Personally, I like that it encourages individual mapping ideas with natural grouping by themes, or affinity. In other words, the KJ Method rewards thought variety. Finding the common ground is still a great pacifier and place to start, even when it doesn’t feel like progress.

Are you seeing it all? Opening up to the affinity map

These days we rarely find ourselves working on a team collaborating to solve a simple customer problem. Affinity maps help participants see the whole by centering focus on opportunity bonds in complex problems. At the same time, leveraging mixed perspectives and balancing sensitivity to “solutions” and “impact” because the best solution isn’t always from the people closest to the problem.

One application of the KJ Method is to filter data from behavior tracking and user surveys, looking for patterns in interview and research responses. Tackling complexity from an alternate position, where participants would consider different approaches to the problem in lieu of solution-first thinking. While it is easy to automate steps of this process that doesn’t get humans closer to empathizing with their topic and user, and ultimately solving the real user problem à la design thinking.

I must point out that the KJ Method shouldn’t be viewed as an extra process, rather a technique to jostle a stale or plateaued routine for analysis and research. An idea percolator, of sorts.

How to use the KJ Method:

Here are some time prescriptions for each phase of the KJ Method. Timeboxes can generally be modified to your needs as each event is unique; soliciting, organizing, and negotiating hundreds of ideas may not fit within a 3-hour time block particularly if participants don’t regularly collaborate on planning activities such as managing backlogs, release planning, program increments, etc.

Setting the stage: Priming a workshop to unify business and customer outcomes through organic discovery.

Involve the right people! This technique can either be the answer to chaos or create more trouble. Involving internal and external stakeholders such as vendors, customers, and contractors mitigates this risk and gets everyone closer to the distinct ideas that provide real value to the user. Also, have a clearly defined scope, topic, purpose, etc. before the scheduled event to manage expectations.

Phase one: Get to thinking!

Begin with a single focus. Everyone, individually, writes out ideas onto cards. Traditionally, the cards were formal like index cards. Although you could use a digital tool, count on these using stickies for convenience.

Phase two: Get to organizing!

Now move the ideas (stickies) into similar themes; placing the stickies in any order gets the job done. Grouping ideas – really, clustering – helps everyone involved begin to see the bigger picture. Clusters expose the relationship between seemingly random ideas. Just stick with your initial gut feeling on affinities when forming themes instead of deliberate grouping.

Slap a name on your themes to identify an essence of the categories along with a brief description. Split themes into suitable subgroups if you find too many stickies sharing a common theme and clearly define them as they evolve.

Phase three: Get to voting!

Pause and review the ideas and categories by voting on the highest priority themes. I like dot voting for this phase. Once everyone has voted, tally and rearrange the themes by the number of votes. Time or context permitting, an additional voting wave can be done to isolate stickies within the most important themes; this may not be necessary for most workshops.

By now, participants should easily be able to rank themes such as best, great, good, and other. Best is a relative value, the gist is a good enough idea – think MMF and MVP.

3 tips for discovering value through diversity using the KJ Method:

Employ a moderator.
Raw ideas are the only input.
Stack your techniques.

Employ a moderator

I don’t recommend the moderator to be the only person capturing ideas, as a scribe. They won’t have time to record hundreds of stickies while appropriately facilitating the workshop. So, try to keep in mind that this technique brings participants front-and-center. There is, of course, two approaches to weaving this activity: silent or discussion-based decision-making. Which should you use? I employ both, leading with silent decision-making until discussion is necessary, which in most contexts happens in phase three and beyond the workshop [link: anchor].

Raw ideas are the only input

Sometimes affinity maps reach a roadblock: should you filter and rephrase ideas? Raw ideas need to be that – unaltered. When you are using data from past performance, that too should be unaffected by the process. Another thing I like about the KJ Method is that it assigns all ideas an equal, uniformed value at the baseline in phase one. It’s a bottom-up approach. To that point, avoid a line of thinking with hierarchy in mind from the start – you won’t know the patterns until everyone’s stickies are shown.

After dot voting and final priority ranking, then you filter and describe components (phase three), fungibility, and objectives of groups and ideas (i.e. after phase three).

Stack your techniques

Envelop the new ideas into your current schedule or iteration planning through sizing for value and usability. Be sure to qualify the sizing on feasibility as well, given the team’s state of readiness.

Also, I like to snap a photo at each evolution. A record of progress adds transparency to the process and logs the crumbs.

“The group came up with many great ideas. Here is the group disentangling many raw ideas, sorting by affinity. The group organized and prioritized valuable ideas.”

Save that somewhere because your stakeholders might back your idea in the future!

A mind mapping workshop stimulates the team through conversations and interactions that will influence future iterations; “that was a great idea and it got me thinking…” It accelerates the idea-to-plan sequence, especially if you have an integrated team that disagrees often.

Experiment with the KJ Method and keep planning activities fresh!