Simulatte / Glossary

The vocabulary of synthetic consumer research.

Plain definitions for the terms Simulatte uses — what a persona pool is, what a verdict contains, where the "real" in a synthetic population comes from, and how faithfulness is measured. If a term shows up in a report and you want the precise meaning, it's here.

See how the method works →
11
CORE TERMS
4
DECISION FAMILIES
2
EVIDENCE BODIES
Traceable
EVERY VERDICT
DEFINITIONS

The terms, plainly.

Every conclusion Simulatte returns is traceable to its population, evidence, and reasoning. These are the words that carry that chain.

Synthetic consumer research
Pre-testing a business decision against a simulated population that behaves like a real market, before that market ever sees the decision. It returns a verdict with the reasoning underneath, in minutes rather than the weeks a survey or focus group takes.
Persona pool
A population of behaviourally distinct simulated buyers, grounded in real data, that a decision is run against. The personas are deep simulations of real people — not real individuals and not a generic chatbot.also: persona population
Decision test
A single question — a price, a claim, a piece of creative, a name, or a survey — run against a persona pool to get a verdict. Simulatte groups decision tests into four families: Creative, Brand & Risk, Commercial & Investment, and Custom Survey.
Verdict
The output of a decision test: the population's response, plus the behavioural drivers underneath it and the segment that disagrees — not a score with no explanation. Every verdict is traceable to its population, evidence, and reasoning.
Grounding
How the "real" gets into a synthetic population. Simulatte grounds a persona pool in one of several modes — the open web, a brand's own CRM, real voice interviews, or a defined ICP — by distilling real buyer voice into structured behaviour: triggers, objections, trust anchors, and price sensitivity. Signal, tagged — not scraped noise.
The fidelity ladder
The principle that a population's faithfulness scales with how close its input sits to a real, specific human. Broad web grounding sits lower; CRM data and real voice interviews sit higher. The closer the input, the more the population behaves like the market it stands in for.
Qualia
Simulatte's adaptive AI-moderated voice interview system. It conducts voice interviews with real people, hundreds at once, following up on what was actually said — the highest-fidelity way to ground a persona pool. In beta.
Digital twin of a customer
A simulated buyer built to behave like a specific real segment or individual, used to test how that person would react to a decision. A digital twin is a simulation of a real person, not the person themselves.
Calibration
Checking and tuning a synthetic population against real-world ground truth. Simulatte reports two things separately, because conflating them is how research gets oversold: whether the population's distribution matches reality (benchmarked against national surveys), and whether its behaviour predicts real outcomes (blind sealed-outcome replay).
The segment that disagrees
The minority of the population whose response runs counter to the majority verdict. Surfacing it is a core part of a Simulatte verdict: you see not only what wins, but who it loses and why.
Prediction infrastructure for human systems
Simulatte's broader category: reusable infrastructure for predicting how populations of people will behave, so a decision can be tested against a market before the market sees it.

Now run one against your own decision.

Bring a price, a claim, or a piece of creative. You'll have a verdict — with the reasoning underneath — in about a minute.