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.