Demo
Product · Synthetic populations

Build the populations
your decisions will encounter.

Imagine All The People builds coherent synthetic populations from statistical data, social structures, and behavioral models.

Each individual has a situation, constraints, preferences, and reasoning that are consistent with the population they belong to.

See the method
France · Mobilité longue distance1 247 382 individus
  • Urban63 %
  • Peri-urban22 %
  • Rural15 %
  • Nadia#01 204 77134 · Lille
  • Claire#04 812 33741 · Tours
  • Lian#08 470 21638 · Lyon
  • 14structures comportementales
  • 38variables structurantes
Sources: INSEE · Eurostat · client data
Changing scale

A population is more than
a handful of personas.

  • Starting point

    Real-world data

    Demographics, geography, equipment, consumption, mobility.

  • Aggregates

    Structures

    Households, income, occupations, habits, constraints, preferences.

  • Individuals

    Individuals

    Each characteristic remains consistent with the others.

  • Result

    Coherent population

    Together, the individuals reproduce the expected collective structures.

Coherence is not added after generation. It is a construction constraint.

Under the hood

10 million individuals.
Not 10 million invented stories.

01Observe

France · ages 18–75

  • Urban63 %
  • Peri-urban22 %
  • Rural15 %
  • Age
  • Geography
  • PCS
  • Household structure

Reference distributions define the population constraints.

02Structure

  • Age × income
  • Geography × mobility
  • Household × consumption
  • Occupation × availability
  • Equipment × usage

Variables are not sampled independently. Their relationships are preserved.

03Instantiate

  • Claire41 · ToursManager · 2 children
  • Samuel45 · CréteilTechnician · single
  • Inès23 · MarseilleWorking student

Each individual is instantiated under the population constraints, not written separately.

04Control

  • DimensionTarget populationSynthetic population
  • Age18–34 · 35–54 · 55–75aligned distributions
  • Geographyurban · peri-urban · ruralaligned distributions
  • PCS8 reference categoriesaligned distributions
  • Household structure5 configurationsaligned distributions

Control means comparing distributions, dimension by dimension.

From the population to the individual

Behind every point,
someone can answer.

1 247 382 individus

Age

  • 18–3426 %
  • 35–5438 %
  • 55–7536 %

Geography

  • Urban63 %
  • Peri-urban22 %
  • Rural15 %

Household structure

  • Single34 %
  • Couple29 %
  • With children37 %

PCS

  • Managers / professionals21 %
  • Intermediate occupations26 %
  • Employees30 %
  • Manual workers23 %

SelectionIndividual #04 812 337

Portrait of Claire, synthetic individual

Claire

Individual #04 812 337

41 · Tours · manager · two children

Situation
Relatively flexible working hours.
Constraint
Needs to coordinate travel with family life.
Price sensitivity
Accepts a premium when the benefit is predictable.
Attitude to uncertainty
High.
What costs me most is not necessarily the ticket. It is not knowing in advance how much I will pay.
Two different objects

Generated persona ≠ synthetic population.

Claire was not created to represent a segment: she is an instance of the population. That is what separates a generated persona from a synthetic population.

  • ObjectA plausible characterA coherent population
  • ConstructionGenerationInstantiation under constraints
  • ScaleA few personasUp to several million individuals
  • Collective coherenceNot guaranteedControlled
  • DistributionsNot guaranteedCalibrated
  • IndividualsIllustrativeInterviewable
  • UseInspirationSimulation

The plausibility of an individual does not guarantee the coherence of a population.

A reusable infrastructure

Build once.
Interview again and again.

Long-distance travelers · France

1 247 382 individuals

  • New pricing policy
  • New loyalty offer
  • Change in opening hours
  • New app
  • Environmental communication

Same population. Five decisions. Multiple simulations.

Starting from the real world

Start from the real world.
Add what is specific to you.

Reference populations

  • France
  • United Kingdom
  • Italy
  • Brazil
  • United States

Available demographic, socioeconomic, geographic, and behavioral data.

Augmented populations

  • Customers
  • Equipment
  • Retailers visited
  • Products owned
  • Purchasing behavior
  • Usage
  • Business segments
  • Proprietary variables

Your business structures enrich the reference population.

Client data enriches the population. It does not need to identify the people within it.

Calibration

A synthetic population
must be controllable.

  1. SourcesStatistical data and reference data.
  2. ConstraintsStructural relationships preserved during generation.
  3. CalibrationComparison of synthetic distributions with expected distributions.
  4. ValidationComparison of simulated results with real-world outcomes when comparable data exists.
Discover our methodology
Your turn

Which population
will encounter your decision ?

  • Customers of a retailer
  • Residents of a territory
  • Travelers
  • Employees
  • Citizens

Free trial, no credit card required. Your data remains confidential.

Your next decision

Which decision do you want to explore?

Describe your need. We can point you to the right level of support.

Talk about your project →

Which population
will your decision meet?

Build it, then question it.

Request a demo