Digital twins

Describe your customer. Interview people who match.

Upload a persona PDF or paste an ICP brief. We match agents from the panel, brief them on your brand, and ask your questions. You get themes, sentiment, verbatims, and a PDF.

Both happen on a 30-minute call. No account, no card. · Enterprise? Book a scoping call →

Or sign up with a work email and run it yourself →

How it works

Three steps.

  1. 01

    Describe the customer

    Paste an ICP, upload a persona doc, or write it in plain words. Role, industry, and geography all help.

  2. 02

    We match the panel

    Lewsearch pulls people from the simulation who fit that description. Beliefs and memory come with them. You review the roster before anything runs.

  3. 03

    Interview and export

    Ask up to five open-ended questions. Read themes and quotes. Download the PDF.

Fielded a study and want simulated rows matched to your exact cells? See Sample boosting.

Example results

What a run looks like.

Same panel, published receipts

Twins are matched from the same respondents we score in public. The methodology page is the receipt.

See how respondents are built

How we score the panel

650,000 census-grounded Americans with beliefs and persistent memory. Lewis, our proprietary model, runs them in a live simulation. Ask up to 10,000 at a time.

As shipped: about 10 pp on 1,521 real polls, 50 states. Calibrated benchmark: 7.47 pp on 404 non-electoral questions. Strict held-out: 9.97 pp.

Open-ended quotes and focus-group dialogue are qualitative and are not covered by these numbers. Election-margin questions are published but not claimed.

How we measure →

Run your first panel study free. Or score us on a study you already ran.