UX: Personas Are Habitual Liars

Chris Feix
3 min readSep 1, 2022

Wait, wait, wait… I’m not telling you to stop doing personas. I’m telling you to stop doing personas until you can do them right. Here’s why.

(If you want to skip to the personas part, scroll down to “Personas Are Liars”.)

Desired Paths is the concept that shows the difference between a designed route and a desired one.

Storytime. Some time around 1999 or 2000 I saw Don Norman speak about something called “Desired Paths” and my head exploded. From that day on, I went from designer to UXer. Obsessed with everything UX.

Unfortunately the world at large wasn’t ready for the timelines and budgets associated with UX Research. I spent a lot of time losing that battle. Eventually I perfected the argument, selling it as “insurance for success”. One of the very first things I would do is create a binder including (in this order):

  • Competitor Research (now Competitive Analysis)
  • Content Map (this is now known as Content Model)
  • User Profiles (now Personas)
  • User Cases (now User Stories)
  • Process Maps (now User Flows)
  • Blueprints (now Wireframes)
  • Idea Boards (now Style Guides)
Example content model: Johnathan Kahn, A List Apart

When I look back at this it’s easy to scoff at the simplicity (and all the necessary missing chunks). But back then it was revolutionary for most clients.

Personas Are Liars

One of those critical missing chunks was User Interviews. “We don’t need that. It’s a waste of money. We know our customers.” Later I would combat this with “Ah, but you don’t know your Users.”

Without a proper interview, your personas are worse than useless. They are dangerous. What do I mean by “proper”?

Are you interviewing Customers or Users?

Which is it? Answer: it should be both. First get to know the customer, then get to know the user. I’ll be writing a lot of articles on this in the coming weeks — starting with an article I’ll probably title “Users Are Liars”.

How Many Personas Did You Create?

My next mistake was creating one or two generic personas. Lumping all customers into a couple personas is also very dangerous. We’re basing design decisions on a marketing-style persona, not a use case. Now I don’t bother with personas until I need to solve a specific user problem. (ex: Users are dropping out of the checkout process after entering their zip code.)

I bundle task-related Journey Maps with the Persona

Conclusion: Personas are useful tools. They are wonderful artifacts as well. But they can only help you if they are accurate and purposeful. For me that means task-related. For you it might be content-specific, or feature-specific. Just remember they can be costly or even dangerous if they are misused or miscreated.



Chris Feix

Entrepre-holic. UX & Business Strategist. Coined several obscure memes.