Personas are a way to get insights and ultimately deliver value. Who are we serving? What are the scenarios they are in when using our products? What are their pains and gains? Understanding this provides guidance and helps us determine whether to accept or reject certain Product Backlog items. The more we learn, the better we understand our users. We have to ask the right questions to uncover that information. But when the bridge to our users is too big to frequently talk to them, how can we learn?
These days, we have the magic of creating our own GPTs. These GPTs can be used to practice and learn. Based on our book, Solving for Value: A Journey of Stupidity and Ambition, I've created an Airport Persona Interviewer as a demonstration of this. It can be useful for:
- Practicing asking better questions
- Connecting the dots across the boards faster
- Find your own blind spots
- Prepare yourself for upcoming user events and meetings
- Workshops and training
- Etc
What’s the Airport Persona Interviewer?
It’s a custom GPT designed to help you generate realistic user personas by “interviewing” them in their own setting, like an airport lounge. Think of it like a guided improv session between you and a hypothetical user. The GPT plays the persona. You play the interviewer. And the back-and-forth uncovers behaviors, motivations, pain points, and even little characteristics that make the persona feel real. Instead of writing out generic demographic templates, you’re having a conversation. It allows you to practice your questioning skills, which, in my experience, we rarely take the time for. The more we validate in real life and uncover more information, the better we can tweak our personas and our GPTs. Here's our initial example Persona:
Why an airport? Because that is the narrative in the book. Different types of people make use of the same type of machines. Are they experiencing it the same way? Probably not. That makes it a great narrative setting to uncover not just what someone does, but why they do it.
How Creating a Persona GPT Works
You start the GPT and give it context. For example, in this case, I uploaded this persona, the others we use, some additional data files, and asked the LLM to analyse them. Then, I told the LLM to ask me clarifying questions it would need answered before starting.
It creates a persona based on that input—age, role, goals, frustrations, the works.
You interview the persona. Ask anything. What’s your day like? What tools do you use? What stresses you out? How do you feel about money/kids/data privacy/whatever’s relevant? The GPT answers in character, revealing insights along the way.
You can tweak, dig deeper, or ask for a summary. And now you’ve got a working persona that actually feels alive. At the end of the conversation, you can ask for feedback on how to ask better questions.
Why This Complements the Traditional Persona Sheet
In my experience, we revisit our personas too infrequently. Most personas end up as PowerPoint ghosts: once created, never used. But personas built through dialogue stick in your brain. You remember that your user gets anxious checking their bank account, or that they trust advice from Reddit more than official sources. That type of feedback is rarely found in surveys. Additionally, you can quickly explore different types of users. Want to see how a cautious CFO might react vs. a tech-obsessed early adopter? Run two interviews. Inspect and adapt on the fly.
How Accurate is It?
It’s not meant to replace real research and critical thinking. If you have access to user data or actual interviews, that’s awesome, and you should definitely use it. However, if you're in the early stages, working on a new idea, or simply trying to get unstuck, tools like this help you quickly consider the human side of your product. They’re great for:
Startup ideation
Story mapping
Marketing experiments
UX sessions
Internal alignment (“Here’s who we think we’re building for, let’s pressure test it.”)
How to Make the Most of It
Be specific in your prompt. The more detail you give upfront, the better the persona will be. Industry, goals, pain points, even rough psychographics and characteristics help.
Go beyond surface-level questions. Ask about frustrations, workarounds, fears, and moments of delight.
Roleplay with purpose. If you're designing onboarding, ask the persona how they’d react to your flow. If you're pricing a feature, ask how much they'd pay.
Save the best outputs. Export, annotate, and share with your team.
GPTs Make Personas a Team Effort
With tools like the Airport Persona Interviewer, you can quickly build your personas. And you don’t need days of workshopping to make it collaborative. Fire up the GPT in a meeting, run a live interview, and watch your team explode with ideas. Don't forget that it's not a replacement, and you cannot trust it blindly. Nothing beats real-life information, interaction, observations, and validation.