Skip to content
Xplorence
Echoby Xplorence
Synthetic panel

Test ideas with a virtual audience

Echo is a synthetic panel: virtual respondents with profiles based on statistics and research. AI answers from each profile’s perspective, so you can compare ideas and group responses.

This is a simulation, not a survey of people. Validate important findings with real buyers.

credit towards a live survey within 60 days
One questionnaireVirtual profilesCompare responses

Synthetic panel diagram. Profiles are virtual; AI generates the responses.

What you get
·Simulated answers by question and available audience groups
·Buyer profiles in the modelled audience — groups with different responses
·A short summary of which options to test with real buyers
·A report, a shareable results page for colleagues and a data export

What people use Echo for

For initial idea selection and preparing questions for customer research.

Test an idea before investing

Compare ideas and identify what to test with buyers before investing.

Explore pricing and payment options

Explore pricing and payment hypotheses, then test them with your target audience.

Shortlist the best of many options

Compare names, descriptions and advertising messages to select options for further testing. The model cannot taste or experience the product.

Understand the category buyer

Form hypotheses about customer needs and choices for further research.

Products

Three products — which to pick

Sonar
a survey of real people

you need a measured figure for a big decision

Timing, price and scope
Who answers
real people — exactly the ones you need
How many questions
your full questionnaire
When you get the answer
the timing is agreed before payment
What it costs
calculated for your task — a quote before payment
Chosen when
you need a measured figure for a big decision
Open Sonar →
Pulse
your question in the country’s shared survey

you need to track the market quarter by quarter

Timing, price and scope
Who answers
real people — a nationwide sample, 18+
How many questions
from one question
When you get the answer
once a quarter
What it costs
quoted
Chosen when
you need to track the market quarter by quarter
Open Pulse →
Echoyou are here
synthetic panel — virtual respondents

you need to select hypotheses to test

Timing, price and scope
Who answers
an AI model, not survey participants
How many questions
your full questionnaire
When you get the answer
within hours of approval
What it costs
quoted for your task
Chosen when
you need to select hypotheses to test
Guarantees

What we guarantee

Echo → Sonar

Echo’s cost is credited towards a live survey within 60 days.

Guarantee terms

If you run a live Sonar survey after Echo, the full cost of the calculation is credited towards it — within 60 days.

Where Echo is weak, we flag it ourselves — before payment, with a “verify live” chip on each question.

Verifiable interviews

In-person interviews are recorded. No recording, no charge.

Guarantee terms

Every in-person interview comes with an audio recording and a quality passport. An interview without a recording is not counted and not billed.

Interview quality is checked before inclusion in the research results.

Transparent results

Pulse reports show respondent counts and margins of error.

Guarantee terms

In every Pulse report a figure comes with the number of respondents and its margin of error; the index follows the European Commission methodology.

The same calculation as in the workspace: one set of maths for report and screen, a margin for every cut.

How to read Echo results

Echo supports initial hypothesis exploration. Reliability depends on the topic and audience; there is no universal accuracy promise for every task.

How to use the results
  • ·The model uses the audience characteristics specified for the simulation. This does not make its responses a representative survey.
  • ·Model responses are simulated estimates, not observations of buyers. They help formulate questions for research.
  • ·Echo results should not be used as measured market share or a sales forecast without further validation.
  • ·For launch, pricing or investment decisions, we recommend validating hypotheses through research with real buyers.
How we check it

We compare model and survey response distributions on specific validation tasks. This identifies limitations on those tasks and does not guarantee accuracy for a new product or audience.

Echo is most accurate about people themselves and about a category it already knows; it is weaker in an unfamiliar category with no history. We send the detailed accuracy table by question type and the checking methodology on request.

Where Echo is weaker
Where the model has little to lean on: narrow B2B audiences, brand-new categories nobody talks about yet, sensitive topics and small segments. We flag such answers in the report and propose checking them with a live survey — rather than presenting them as a measured figure.
AI, data and the rules
AI use is disclosed under the ICC/ESOMAR Code (2025 revision). No real individuals are profiled. Your questionnaires and results are never used to train models. Data is stored in Uzbekistan. Full ESOMAR disclosure →

Frequently asked

Tell us the task — we will show what Echo answers

In plain words: which product, which decision and by when. We name the quote and the deadline before the start.

contact@xplorence.com