Key takeaways
- Fraud shows up in the screener first; design it to catch fraud.
- Combine identity verification with behavioural checks.
- Clean panels continuously, not just before a study.
In B2B studies, a small share of fraudulent or inattentive respondents can distort results badly, because the target sample is already small. Every bad complete matters.
Where bad data comes from
- Respondents who claim a profile to qualify for higher incentives.
- Duplicate accounts and shared panel logins.
- Automated or scripted completes.
- Speeders who skim through the questionnaire.
Layers of defence
- 1
Verify identity and role
Match respondents to professional profiles and company data before they reach the survey.
- 2
Design a smart screener
Avoid obvious qualifying answers. Use knowledge checks that only a real practitioner can answer.
- 3
Add in-survey checks
Use trap questions, consistency checks, and open-text review alongside time-based flags.
- 4
Review after fieldwork
Remove flagged completes and feed the findings back into the panel.
None of these steps is enough alone. Quality comes from stacking several light checks so that no single weakness lets a bad respondent through.
Need experts for your next study?
Tell us your audience and we will scope feasibility.
