How It Works Use Cases Accuracy Get in Touch

by Thursday Strategy

Human panels
+ AI twin modeling

Rigorous human insights meet individually mapped AI twins — giving you faster, more flexible, and deeply trustworthy consumer research.

Watercolor illustration of diverse people

Built from a gap in the
research landscape.

Synthetic respondents have a clear value proposition — speed, cost savings, flexibility. But existing platforms sacrifice the depth and authenticity that comes from real human insight. We believe synthetic responses need to be deeply rooted in rigorous human data to earn your trust.

Human Respondents

  • Authentic & nuanced
  • Inspiring depth
  • Slower timelines
  • Higher cost
  • Respondent fatigue
  • Harder with niche audiences

Synthetic Respondents

  • Faster results
  • Lower cost
  • More flexible
  • No research fatigue
  • Difficult to trust
  • Flat & uninspiring

one+1 gives you the
best of both.

Real human depth, with the speed and flexibility of AI — validated at every step.

Depth Accuracy Humanity + Speed Efficiency Flexibility

How it works

01

People answer questions

One-time study or ongoing engagement. Your respondents and data, or ours. We design rigorous, longitudinal human panels for depth, accuracy, and humanity.

02

We generate AI twins

For each respondent, we build an individually mapped AI twin using our proprietary modeling approach — combining LLM, graph model, traditional ML, and vector-based techniques.

03

You get instant insights

Ask one question or several. Receive a research synthesis with reported confidence and accuracy rates. Flexible access to your audience, whenever you need it.

How we know it's working

Twin-level accuracy

We measure accuracy at the individual AI twin level, testing predictions against actual human responses and boosting through proprietary prompt tuning.

Prediction confidence

We measure semantic similarity to gauge confidence on every individual prediction. If we have low signal on a topic, we flag it — no black boxes.

Statistical rigor

Because we create AI twins 1:1 with individual respondents, standard significance testing and confidence intervals remain valid — reported alongside accuracy context.

How it can work for you

Pilot-ready

Get more from existing quant studies

Ask the questions you didn't, or couldn't, ask in your actual study — unblinded concepts, confidential stimuli, questions that got cut — and get predicted responses for each individual respondent.

Pilot-ready

Supercharge proprietary panels

Build robust twins of your panelists over time so you have the flexibility to get feedback and insights whenever you need, without exhausting your respondents.

Pilot-ready

Niche audiences without fatigue

Create an ongoing panel of hard-to-recruit audiences for more flexibility and reliability — no more worrying about respondent burnout.

Launching Q2 2026

Gen-pop on demand

Query a representative sample of US adults with the flexibility to ask the exact question on your mind. Think Pew Research meets on-demand access.

Ready to see one+1
in action?

We're currently onboarding pilot partners. Let's explore what one+1 can do with your data.

Get in Touch