The problem with perfect personas.

Do you even know what a ‘perfect’ persona is and what problems it should help you solve?

Do you know the perfect way to do personas? If you had unlimited time and budget, what would you do to create them? What would they look like? Think about that for a sec.

You now have perfect personas. Are you absolutely sure they will work?

You should be, right? There are no excuses to fall back on. These are perfect.

But I bet you feel a little weird in the pit of your stomach. I bet you’re not entirely sure.

Why? How can that be? Many reasons. But I’m going to start with the two big ones:

1. “There’s something bigger than Phil.” There are forces at work in your organization (or your client’s) that can strike like lightning or tsunamis at any time. These tend to come from the Castle of Executive Mysteries. All of a sudden something huge (or small) changes at some point (usually many points) during a project. Your perfect personas are left charred with smoke coming out of their well-documented ears or spluttering with underpants full of sand on the shore of some new island far from where they were born. There is a way to control the weather and I’m going to get to that in future posts.

2. You spent some happy moments dreaming up perfect personas. Did you also dream up answers to: "How will we know the personas worked? What does that even mean?” And if I ask you that question now, what would you say? Maybe ‘they built empathy for our users,’ or ‘they brought data about users to life,’ or ‘they helped us make product or marketing decisions.’ Um, how do you know? Did you measure the impact? How?

The first step isn't a plan to measure effectiveness. The first step is to figure out which problems you need solving much more specifically than "we need to be more user-centered." Even way back in our 2005 Persona Lifecycle book, John Pruitt and I highly recommended defining the goals for your persona effort itself. We recommended a four-column spreadsheet to list:

* Problem we are trying to solve by using personas
* Examples of that problem in our org/project today
* How those examples SHOULD look when we fix the problem with personas
* How we will measure whether the personas solved this problem

Do you do this? Probably not. We are all so busy arguing about the perfect way to do do personas that we don’t talk about what concrete changes we think perfect personas would make in our strategy, design, and dev processes. But we can.

What problems have you tried to solve with personas?