Product · Dynamic agentsInterview #04 812 337 · simulation excerpt

An answer is only
the start of the interview.

Our agents interview each individual, detect hesitation, contradiction, or friction, then decide what to ask next.

They do not run through a fixed questionnaire. Each interview evolves based on the answers received.

Understand how it works
Portrait of Claire, synthetic individual
Claire#04 812 337

Question askedWould this new pricing change your habits?

  1. Claire

    « Probably not. But I find it rather unfair. »

  2. Agent

    What feels unfair to you: the price itself, or the way it is applied?

  3. Claire

    « It is not really the amount… »

Friction detectedPerceived fairness
The principle

One question.
Then every interview follows its own path.

The question is the same. What follows depends on each answer.

Question asked to everyone

“Would this new pricing change your habits?”

Portrait de Claire
Claire#04 812 337

“Probably not. But I find it rather unfair.”

  1. Unfairness
  2. “What feels unfair to you?”
  3. Lack of consultation

Perceived fairness

Portrait de Samuel
Samuel#04 812 908

“No. It would not change much for me.”

  1. Low impact
  2. “Why would it make little difference to you?”
  3. Occasional use

Low exposure

Portrait de Inès
Inès#04 813 244

“It mainly depends on the amount.”

  1. Price
  2. “At what amount would it become a problem?”
  3. Comparison with an alternative

Price sensitivity

The next question is not written in advance.

Changing scale

100 000 individus.
100,000 different interviews.

One shared opening question. Then as many trajectories as there are individuals interviewed, all conducted simultaneously.

The same process, repeated for every individual in the population.

  • 7–12follow-ups per interview
  • 20–30on complex cases
  • 1 000+parallel processes
Two different methods

A questionnaire asks.
An agent investigates.

Path
QuestionnaireDefined before the interview
Dynamic agentEvolves with the answers
Next question
QuestionnairePredetermined
Dynamic agentDetermined by what was just said
Frictions
QuestionnaireCollected if anticipated
Dynamic agentDetected during the exchange
Contradictions
QuestionnaireRarely explored
Dynamic agentCan trigger a follow-up
Depth
QuestionnaireLimited by the protocol
Dynamic agentPursued until the reasoning is clear
Going deeper

The first answer
was not the right answer.

74 %No

“Would you accept your bank branch closing?”

Stated response · simulation example
The interview continues · 6 follow-ups
  1. Why?
  2. “I still need it.”
  3. In what situations?
  4. “Actually, I rarely go there…”
  5. So what bothers you about it?
What the interview reveals
“It is not really the branch I would miss. It is the feeling that our area is being abandoned yet again.”

Emerging frictionRegional neglect

Stated responseUnderlying reason

What we had not anticipated

Discover what you did not know you needed to ask.

An interview that follows the answers eventually moves beyond the initial scope. The real issue behind a decision is not always the one written into the protocol.

  1. Planned questionPrice
  2. Contradiction followed upStated intent ≠ behavior
  3. Emerging topicTrust
Case · Crisis & reputation

00:30 — the crisis breaks.
01:30 — you know why.

Within an hour, Imagine All The People turns thousands of reactions into usable understanding.

  1. 00:30

    The alert lands

    The new announcement triggers a wave of negative reactions.

  2. 00:38

    The question is framed

    “What is really driving the rejection?”

  3. 00:44

    The interviews begin

    5 400

    people are interviewed simultaneously.

    and 5,397 other conversations, all different.

  4. 01:14

    The underlying reasons emerge

    • Perceived fairness
    • Trust
    • Price
    • Lack of consultation
    • Fear of change
  5. 01:30

    You know
    where the risk lies.

“It is not really the price. It is the way it is being imposed on us.”Individual interview · “long-standing customers” typology
  • 3frictions majeures
  • 4scenarios explored
  • 5 400individuals interviewed

The team can now test what calms the situation, what makes it worse, and what could turn it around.

From insight
to conversation.

A conclusion is only useful if you can understand what produced it. Every result can be linked back to the individual interviews, follow-ups, and reasoning that made it emerge.

Discover Explainable AI
Your turn

Which decision
do you want to explore ?

  • Test an offer
  • Understand opposition
  • Explore a friction point
  • Comparing two messages

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.

Question a population,
interview by interview.

One interview, then a hundred thousand.

Request a demo