The anti-pattern pattern for ideation

by Brian Doll on July 19, 2017

In software, we love patterns. Inspired by Christopher Alexander’s book, A Pattern Language, developers began formalizing patterns they’d seen and cataloging them. Many of us have likely read or at least skimmed Martin Fowler’s book cataloging patterns for enterprise applications.

At Reify, we love patterns for marketing and sales, too. When we work with clients on messaging, we start by developing the persona we’re trying to engage with. Michael shared some great details on how we do this: How to write persona-driven marketing copy that never misses its mark Yes, we used this process to write that

In our process for developing a messaging framework, we are combining two important phases in sequence — Divergence and Convergence. Jeff Veen has a talk called “Designing for Disaster” which I loved. He talks about how important it is for everyone in a product or design review meeting to understand if the meeting is about divergence — finding lots of relevant ideas, or convergence — narrowing down our options to find the best one. (It’s just a few minutes of the talk — check it out.)

When you run out of ideas of what something is, think next about what it isn’t

We were working with a client yesterday, ideating on the personas that they market and sell to. One of the most helpful questions we explored with them was this — who do you not want to sell to?

As with any ideation process, you don’t want to question anything until you’ve collected lots of insights. We gathered a list of anti-personas from the team, and then we talked about them. Exploring these anti-personas, and teasing out the details about what made them a “bad” fit for the company was more useful and enlightening than our initial list of personas. Digging into what something isn’t helps sharpen the edges of what it is.

Michael’s post about persona development contains a great pattern for any ideation process: Generate → Filter → Refine. We can extend that a bit to include this anti-pattern idea:

  • Generate: generate a list of possible options

  • Oppose: generate a list of anti-options, directly opposing your initial list

  • Filter: filter the list and select the best choices

  • Refine: refine the results into the form we’re looking for

What do you think? How do you brainstorm ideas? What patterns have you found most useful? We’d love to hear from you!