Synthetic User Research

Synthetic user research is the use of AI-generated personas to simulate user behavior, surface product assumptions, and generate qualitative insights — without recruiting, scheduling, or compensating real research participants.

It's not a replacement for all user research. It's a way to dramatically increase the volume of research moments so that the expensive, high-quality work you do with real humans is better informed and more focused.

The Problem With Traditional UX Research

User research is underused in most product teams — not because people don't value it, but because it's slow and expensive.

• Recruiting 5 participants for a usability study takes 2–3 weeks
• Moderated research requires a trained researcher and costs €3,000–€8,000 per round
• Survey results take days to compile and are plagued by response bias
• By the time insights arrive, the team has already made the decision

The result: teams do less research than they should, at the wrong moments, and rely too heavily on instinct between research rounds.

What Synthetic User Research Adds

Synthetic research doesn't replace usability studies or ethnographic interviews. It fills the gaps:

Before you recruit real users. Generate hypotheses. What do you think users will say? Where do you expect friction? Run the scenario with synthetic personas first — you'll ask better questions in real research.

When speed matters more than depth. A team shipping every two weeks can't run a moderated research round each cycle. Synthetic research gives you a fast, cheap signal to inform the next sprint.

For volume at the top of the funnel. Test 15 versions of a concept in the time it would take to recruit 5 participants. Use synthetic research for breadth; real research for depth.

When you can't access the right users. If your target is a German manufacturing CTO or a US hospital procurement director, recruiting 8 of them is nearly impossible. A well-calibrated synthetic persona can approximate their perspective.

How to Build Useful Synthetic Personas

The quality of synthetic research depends almost entirely on persona quality. Generic personas produce generic insights.

A useful synthetic persona has:

• A specific role and context. Not "B2B buyer" but "Head of Procurement at a 300-person German Mittelstand company, responsible for software vendor evaluation, reports to the CFO."
• A point of view. What do they already believe about this problem space? What have they tried before? What made them skeptical?
• Behavioral constraints. How do they make decisions? Who do they consult? What would make them say no?

In Minds, you build these personas by writing a detailed profile and optionally uploading supporting material — interview transcripts, CRM notes, customer reviews, support tickets. The more grounding data you provide, the more accurate the simulation.

The Research Workflow

Step 1: Define the research question. What decision are you trying to make? "Does this feature resonate with our ICP?" is a research question. "Is our product good?" is not.

Step 2: Create your persona set. Build 3–5 synthetic personas that represent your key user types. Include edge cases — the skeptic, the power user, the non-technical buyer.

Step 3: Run structured interviews. Ask each persona the same questions you'd ask in real research. Walk them through a scenario. Show them a prototype description and ask for their reaction.

Step 4: Compare and synthesize. Where do personas agree? Where do they diverge? Divergence is often the most interesting finding — it tells you where you need more segmentation in your product strategy.

Step 5: Bring findings to real research. Use synthetic insights to build a sharper hypothesis for your next real research round. You'll get more value from a 45-minute user interview when you already know which questions matter.

Limitations

Synthetic research reflects what you put in. If your personas are based on assumptions rather than real behavioral data, they'll surface your assumptions back at you, not genuine user behavior.

It can't replace behavioral observation. What people say and what people do are different. Synthetic research can approximate the former; it can't replicate the latter. Don't use it to replace usability testing.

Edge cases may be underrepresented. Synthetic personas tend toward the modal behavior of their segment. Outliers, power users, and unexpected usage patterns require real users.

Getting Started

Minds is purpose-built for this workflow. Create a group of 5–8 synthetic user personas, then use Panels to run structured research sessions — asking each persona the same questions simultaneously and reviewing their responses side by side.

Most product teams start with their ICP and 2–3 adjacent user types, run a panel session on their most pressing product question, and use the output to sharpen their next real research round.

Start your first synthetic research panel →