Customers Are Talking: the Weird, Alchemic Process of Distilling Insight from Stories

futurelab default header

by: John Caddell

Book clubs are big these days. A group of folks read the same book, then get together and discuss it, accompanied by refreshments (often of the wine & cheese variety).

Besides the social aspects of the book club, there’s something powerful about a group of different people, who’ve read the same story, discussing and deciding what it’s about.

When I try to describe to folks the work I do helping companies gather, find patterns and put to use stories their customers tell, I say that the “discussing and deciding what it’s about” is the secret sauce of the whole process. It’s the part that can’t be automated or left to algorithms to decipher–it requires a diverse group that shares key experiences to root their assessments in common ground.

You can find and collect stories by hand, by computer or by applying algorithms to data sources. (The pluses and minuses of each technique are best left to another post.) But I haven’t figured out a way to computerize the weird alchemy that allows a group of colleagues to distill 100 stories into 10 deep insights in 4 hours, and I’m fairly convinced there may never be a way to automate it.

The mechanics of the process are mundane, at least the way I’ve practiced it. I spread the stories, printed one to a page, around a conference table. Graphs that show the correlation of certain story attributes (the graphs are also stories) are arrayed on the wall. People immerse themselves in the stories. Perhaps they are sales calls. Or complaints. The people read slowly at first, tentatively. Then one person writes something they noticed on a sticky note. Then something else on another. Soon everyone is writing.

They may share thoughts while this happens. I scramble around the room collecting stickies and plastering them to an empty wall. Eventually the pace of writing stickies slows. It’s like cooking microwave popcorn. When the frequency of popping slows down, it’s done.

They cluster the stickies, finding relationships. Then they name the clusters–those are insights. Perhaps they go through another round of clustering and naming, if they have time and energy.

Then we talk. There are 8-10 clusters that stand out. They may be issues their customers are facing–which present opportunities and threats for the company’s products. They may be values the customers hold–which are key to marketing and positioning the company. They’re always interesting, usually surprising, and often unveil conflicting or contradictory views. For instance, a very strong attribute the company or product has often is highly valued by one constituency but viewed negatively by another.

It’s an amazing process to watch, to see a group of people take the same material, view it from different angles, reconcile their assessments, and come up with “the truth” as they see it.

It’s a lot like a book club, without the wine and cheese.

UPDATE, 6 May: Per Stephane Dangel’s comment below, here is a fictional example of a story told by a graph:


This chart tells a story, don’t you think? It’s a crosstabs chart comparing people who bought a product over the phone versus those who didn’t, under two scenarios. In one, the sales representative used established best practices; in the other, he/she didn’t.

I get two themes from this “story.” One, when reps use best practice, more sales are made. Two, best practice is not frequently used.

Both are valuable insights for a company seeking to improve its telesales. The first is probably no surprise (though it’s possible to imagine a situation where using best practice would make no difference in closing sales–that would be a surprise!). The second theme is probably surprising under any circumstance!

Related post:
Another kind of value proposition

[If you’re interested, the new version of Cynthia Kurtz’s "Working With Stories" e-book contains a case study I did on my first story project. Check it out here.]

Original Post: