“This is never going to work. I don’t get it. Why are we doing this?”
“Don’t worry, Cody.”
I’m teaching one of my 16-year-old son’s friends how to do an affinity analysis on some data he’d collected for a high school “critical thinking” project. Sometimes I’m not sure how I get myself into these things. But here we were.
That semester, Cody had picked topic he was interested in, how teens felt about religion, created an interview protocol and a survey, and collected about 60 responses. Three days before the due date, he showed up at my door asking for help (hey, we’re talking teens here—three days is actually PROACTIVE in high school). His statistics seemed contradictory. His initial hypothesis, that teens who actively questioned their faith would be more satisfied than those who had blindly accepted theirs, didn’t seem to hold water. He had both questioning teens and accepting teens, but there didn’t seem to be much of a correlation with satisfaction or happiness.
I suggested we do an affinity analysis with his open-ended question responses and interview findings, and helped him create about 600 little yellow sticky notes, a separate nugget of “customer” data on each one. We had just started the affinity, and we had small groups of stickies all over my basement wall. The tempting sounds of my son playing Xbox were in the background. Cody was getting anxious.
“I don’t see where we’re going. Can’t we just go back and look at the stats?”
“Dude. Chill. Just go with it.”
For those of you who’ve never experienced it, an affinity is a way to structure qualitative data so that it reveals the underlying issues. You start with just a blank wall and your data points on a pile of sticky notes. Then you start putting stickies on the wall, clustering data points that seem to “go together.” After groups form, you label the “go-together-ness,” or affinity, that the groups entail. You can also repeat this grouping exercise with the first level labels, and so forth. In the end, you have a hierarchical organization of labels that identify the main patterns and issues in your data—they tell the real story of the information you collected. The really cool thing about affinity analysis is that the issues surface directly from the data points—you’re not sorting the stickies top-down into preconceived bins, you’re letting the issues emerge from the ground up.
“It’s all just random groups—this is aggravating.”
“Cody. I do this for a living. Trust the process.”
Cody was going through the emotional gyrations that every one of our client teams goes through when they’re first exposed to affinity analysis. When we start, everyone is energized by the possibilities of four blank walls and the data lying in neat piles of stickies. But once the analysis starts the team frequently descends into an emotional sinkhole—the one Cody was in now. Chaos reigns as people walk around the room putting stickies up, making groups, and then taking them apart again and reforming them elsewhere. Soon the walls are covered in seemingly random groups of two or three or twenty sticky notes. You read a new note and remember something on this wall that was like that… but wait, someone moved it. So you put it somewhere else it seems to go, but you worry that it’s not quite “right.” You wonder aloud where that one item about X went. Who messed up your sticky note group around that?
During this phase, most people get at a little frustrated—it seems random to the untrained eye and it also seems like “organization” is the last word anyone would use to describe what is happening to all of that neat, clean data. You can’t see a way that this is ever going to come together. You worry about the logic of what you’re doing, and you doubt the whole process.
What happens in the emotional sinkhole is that you are being asked to think in a new way, a way you’re not used to. In grouping and regrouping the stickies, your brain is analyzing what it is seeing, recognizing and evaluating patterns in the groups. You are not simply looking at the data notes and mentally testing them against what you already know. You are not using deduction, which makes up the bulk of most people’s formal schooling and professional training. Instead, you are thinking bottoms-up, recognizing order in what seems like chaos.
You are using induction.
Inductive thinking is a key component of design thinking. Most people are not comfortable in this mode, and the angst teams feel at this point is a real indication of just how dominant deductive, linear patterns of thinking are. Cody was feeling exactly this anxiety as he tried to get his head around inductive thinking (and resist the lure of Xbox at the same time).
The unease is palpable during the initial phase of affinity. But then something happens. Somewhere during the second day of one of our affinity building sessions, the sticky note groups stabilize and you move to a different part of the process—labeling the groups. Now things become more comfortable, as you move back to a more analytical mode of thinking. The patterns start to become more obvious, and you start to see the light at the end of the proverbial tunnel. Usually this little emotional kick carries the team to the end of the analysis on a high note. You can almost see the relief on our team members’ faces as they get to return to a more familiar way of thinking and they see the payoff. Cody was no exception.
“Hey—now I can see four bigger issues why kids like and don’t like what religion does for them.”
“See?”
“That wasn’t in my stats. This is cool.”
Unfortunately, this difficulty in adopting a new way of thinking has far-reaching implications. In fact, I think the little emotional rollercoaster that we take teams on in a two day affinity analysis is the same journey that organizations and even whole companies go on in adopting design thinking into their business processes. But whereas we have the luxury of an affinity analysis that is time-boxed to a couple of days, organizations trying to adopt design thinking as a way of doing business face a longer and more complex problem. It’s harder to play the trust card and just power through.
I think this might be why it is so hard to institutionalize design thinking—and its close cousin, user-centered design—in organizations. Too often, people and teams lose faith and then interest, and efforts falter amidst the prevailing doubt. Even in the C-suite, people are comfortable with linear, deductive logic, but the inductive ideation from observed behaviors and the iteration that characterize design thinking seem to break this logic. It’s easy to doubt and lose faith in the process, sometimes even for those who are trying to champion design thinking in their organizations.
It may also be why so many companies struggle implementing Agile development processes; Agile, like UCD, is characterized by iteration and emergent design from small pieces. All of these diverse practices require a level of comfort with induction and iteration that is neither commonly found nor commonly taught.
The truth is that for the majority of us have a hard time trusting induction and iteration. It feels inefficient, meandering, and random. It feels “illogical.” It’s hard to trust the process.
It’s unfortunate, because the payoff is significant. Teams we work with invariably say the affinity is the point where they really start to have innovative insights into the design problem. And some of the most innovative companies in the world are the ones who have successfully integrated design thinking deeply into their business. It might be hard to trust the process, but it pays.
And Cody? In the end, he trusted and finished his affinity. In doing so, he found a rich set of reasons why his respondents were satisfied with their faith and how they got there, nicely augmenting his statistics and explaining the reasons and motivations behind the numbers. He turned out to be quite good at induction, actually. He pulled the “A.”
Now he wants help on Physics. Sigh.