Synthetic users

Synthetic users are AI-generated stand-ins for real people: a language model given a persona profile and asked to react to a page, a message or a product as that person would.

Why it matters

They give fast, cheap, consistent feedback, including on designs before they're built. They're useful for catching likely objections and comparing versions, but they aren't real customers and can't predict conversion rates.

How to get it right

• Ground personas in your real audience, as specifically as you can.
• Look for objections several personas share, not single quotes.
• Keep the panel and scoring fixed when comparing versions.
• Treat the output as hypotheses and check big decisions with real buyers.

Common mistakes

• Treating synthetic feedback as user research.
• Changing the prompt between runs, then comparing the results.
• Asking leading questions and reading the agreement as validation.