Your customers have already written their own persona. Here's how to find it.

Welcome to Prompt, Tinker, Innovate —my AI playground. Each edition gives you a hands-on experiment that shows how AI can sharpen your thinking, streamline your process, and power up your creative work.

This week’s playground: Building a persona from the voice of real people

Let’s be honest: most customer personas are fiction.

They’re created in a conference room, based on demographic assumptions and decorated with a stock photo ("Meet 'Marketing Mary,' she's 35 and loves yoga!"). The result? Made-up names, arbitrary demographics, and stock-photo backstories. These documents check a box, then collect digital dust.

When a persona is built from real customer language—their frustrations, their goals, their "aha!" moments—it transforms from a caricature into a powerful decision-making tool. It helps you write copy that resonates, build features that solve actual problems, and create strategies that stick.

The new playbook

Smart teams are flipping the script. They’re feeding AI raw customer data—support emails, product reviews, interview transcripts—and letting the model do what humans struggle with: synthesizing insight from noise.

• Marketing teams are pasting in G2 reviews to extract sticky copy and recurring value props.

• UX researchers are analyzing interviews to capture emotional nuance and real friction points.

• Community leads are mining Discord chats to hear what users actually care about—unfiltered.

AI becomes your research assistant: tireless, unbiased, and fluent in pattern recognition, able to surface the core human story hidden within.

Your AI experiment: Try this prompt

Your goal is to turn qualitative data into a high-fidelity persona. You're not asking the AI to invent anything; you're asking it to synthesize reality.

👉 Time to tinker:

1. Gather your raw material: Collect 5–10 reviews, a few support chats, user interview notes, or Reddit comments. Real words from real people. (Remember to anonymize all data.)

2. Copy and paste the entire prompt below into ChatGPT, Claude, or Gemini and insert your data where noted.

📝 Prompt: “You are an expert market researcher and UX strategist. Based strictly on the raw feedback provided below, synthesize a single, cohesive customer persona. Go beyond demographics—focus on psychographics, motivations, pain points, and behavior.

Create the persona profile using the following structure:

• Persona Name: Give the persona a descriptive title that reflects their core behavior or mindset (e.g., "The Cautious Optimizer," "The Overwhelmed Achiever"), not a generic name.

• Summary: Write a short paragraph that captures the essence of this person and their relationship with the product/service.

• Direct Quotes: Pull 3-5 powerful, representative quotes directly from the source text that best exemplify this persona.

• Goals: What are they ultimately trying to achieve? (List 2-3).

• Frustrations/Pain Points: What specific obstacles or annoyances are getting in their way? (List 3-4).

• Motivations: What is the underlying "why" driving their behavior and their search for a solution? (List 2-3).

Here is the raw feedback:”

[Paste or upload your collected customer reviews, interview snippets, or feedback here.]

📝 Prompt: Go deeper

Don't stop at the persona. Use these follow-up prompts to turn insight into action:

• For marketing: “Review this web copy through the lens of this persona. What feels off, irrelevant, or generic? What would make it hit harder—emotionally and clearly?”

• For product: "What is the primary 'Job To Be Done' this persona is hiring our product for? Frame it using the 'When I..., I want to..., so I can...' format."

• For strategy: "Now, describe the 'anti-persona.' Based on the same data, who is our product not for? What are their characteristics?"

What did you discover?

This isn’t just a better way to build personas. It’s a shift in mindset. Stop thinking for your customers and start thinking with them. Every review, every complaint, every praise-filled rant—they're giving you the blueprint. AI just helps you hear it.

• Did the AI-generated persona feel more "real" than ones you've created in the past?

• What was the most surprising pain point it pulled from your data?