Skip to content
Contact
Contact research

One quality layer for every respondent, wherever they come from. What it checks, and what it never does.

Quality layer preview with demo data: respondents pass five stations, some get a second chance, some are stopped.

How we protect data quality

Fraud prevention we build ourselves.

Protecting your data from fraud is a core part of what we build. Five stations run before a respondent reaches your survey, and we keep working on them.

  • Built in-house

    We design, build and tune our checks ourselves, so we know the reason behind every decision.

  • Before every survey

    Every respondent passes five stations before being forwarded to a survey, on every route into our panel.

  • Signals, not documents

    We recognise real people by device and behaviour signals, not by collecting documents.

  • One person, one participation

    We block repeat entries into the same survey, and duplicate participations are not rewarded.

  • In the respondent's language

    The profile and our checks run in the respondent's language.

  • Measured and tested

    We measure our checks against what research clients report back on completed interviews, and test them with our own automated test agents in a closed environment, never in real surveys.

No check catches everything. That is why we keep measuring and improving them.

The five stations

What we check, and what happens if it fails.

StationWe checkIf it fails
01EntryConsent, language and the minimum age of 18.The respondent leaves before any survey.
02DeviceAutomation markers and device signals. No name, e-mail or phone number.Stopped at the door. No survey is shown.
03ProfileAge, year of birth and earlier answers have to agree.A second chance for an honest slip, a clear exit for contradictions.
04Open textOpen answers for effort and sense.No redirect to a survey.
05MatchFit with the survey's audience. Repeat entries into the same survey are blocked.Other surveys, or none today.

What we never do

Three things that stay off the table.

  • Reward duplicate entries into the same survey.
  • Mix synthetic answers into real data.
  • Let anyone skip a check that others pass.

For your review

The documents behind the checks.

We follow the ICC/ESOMAR International Code. How we apply it is set out in our research ethics.

Research ethics

Available on request

  • Answers to the ESOMAR 37 questions on request.
  • Panel profile and feasibility on request.
Ask for them

“Honest mistakes get a second chance. Patterns of fraud do not.”

The rule behind every station.

Ask us about our checks