Products · Research

Simulate the decision before you make it.

Research Lab turns a question into a study and a study into a report — read against synthetic populations of thousands of calibrated Brazilians, in minutes.

Subscribe →Start free →
§ 01The instrument

Studies, not surveys.

Describe what you want to test — a price, a claim, a concept. The Lab assembles a synthetic population — research personas calibrated to real Brazilians — runs it through the cognitive engine, and returns purchase intent, perception and switching risk. Not a deck you wait weeks for. A report you read the same afternoon.

IIdentity
IIMemory
IIIEmotion
IVSocial masks
VBeliefs & values
VIDecision engines
VIIEvolutionary narrative
§ 02Calibration

Calibrated against ground truth.

Every population is raked to the official 2022 Census marginals — sex, age, income, education, region. Each cohort ships with a calibration panel showing how close the synthetic population sits to the real one, marginal by marginal.

Calibration · Census 2022 · IBGEillustrative
SexΔ 0.4pp
Age bandΔ 1.2pp
IncomeΔ 1.8pp
EducationΔ 1.5pp
RegionΔ 0.9pp
§ 03Disclosure

Bands, not points.

We don't send point estimates. Every behavioral estimate comes with an uncertainty band (±MAE) and a provenance card naming its calibration sources. False precision is worse than an honest range — so we never hide it.

See how we measure →
§ 04Inside the Lab

What you actually work with.

Study wizard

Type → audience → protocol.

Start from a blank study or a CPG, retail or fintech template, define your stimulus and questions, and run.

Calibration panel

Census 2022 marginals.

Sex, age, income, education and region, scored with ρ and MAE against the official anchors — green only when it earns it.

Uncertainty bands

P10 · P50 · P90.

Every estimate carries its range, not a single misleading number.

Audit card

Model · seed · hash.

Each report ships with a provenance card and honesty flags — green, amber or red.

Ready-made populations

Banking & beyond.

Pre-built populations for major Brazilian banks, plus general, CPG, streaming and longevity panels.

Lab · Cascade

Watch it spread.

Define an event and see how a reaction propagates cluster to cluster, day by day.

§ 05Who it's for

For the teams that make the call.

Marketing

Test campaigns and messages.

Pre-test concepts, claims and creatives against simulated Brazilian audiences — before spending on media.

Product

Validate features and concepts.

See how real segments would react to a feature or positioning before you build or launch.

Strategy

Explore new segments.

Investigate demand, pricing and fit in a region or cohort you haven't entered yet — without a months-long study.

Agencies

Run concept tests at speed.

Build repeatable tests for multiple clients, each grounded in how Brazilian consumers actually decide.

§ 06Plans

Start free. Scale when it pays off.

Semente
IBGE · ±9pp
Free
  • General BR population (2022 Census)
  • ~2 studies / month
  • Basic export
Estudo
±3pp · solo
R$ 290 /mo · R$ 2.900/yr
  • Segmented populations
  • ~8 studies / month
  • Full export + audit cards
Hidratada
±3pp
R$ 890 /mo · R$ 8.900/yr
  • Segmented populations
  • Unlimited studies
  • Lab · Cascade
  • Full export + audit cards
Enterprise
custom
Talk to us
  • Your own ground-truth
  • API · SSO · SLA
  • Hydration with your data

Prices in BRL. Annual = 2 months free. Recurring billing via Stripe — Pix, boleto or card.

Simulate before you decide.

Early access →
By invitation only · Cohort 03 open