How do you prevent fraudulent respondents?
This is a very important question. All of marketing research is based on having the sample that we require. If we want to survey cancer patients, they need to be cancer patients! And with more and more surveys paying respondents, and more studies being conducted on line, fraud is a very significant issue. Anyone can get 100 email addresses in a few minutes and take a survey many times.
The most important action that we take to prevent fraud is how we recruit our panelists. We meet them in person! We go to patient events like walks, conferences, and education symposia, explain what market research is, and invite them to join our panel. No one just walks in off the street to the Hemophilia, Fabry, ALS, etc. conference and pretends to have the condition! At least we’ve never seen that in the hundreds of events we’ve attended. We had an incident recently where it appeared to our client that two hemophilia caregivers, both in Baltimore, were the same person, trying to participate twice in a study. But we had met them both in person when they joined our panel, we knew they were related (it is a genetic disease), but they were absolutely not the same person. We wouldn’t have been so sure if we hadn’t met them at a hemophilia patient event in Maryland.
Those that we recruit in person often refer other patients to us. They’ll tell their support groups, closed facebook groups, and relatives (for genetic diseases). We’ve found these referrals to also be as authentic as those we recruit in person.
But some people do sign up on line, and we need to make sure they aren’t frauds. First, we run some automatic checks through our database provider, Q One Tech, for such things as IP country match, browser language match, etc. We manually look through the information provided – what is the email, address, phone, and other information provided? Is it a real address? We find the best way to validate a panelist is to call them – after even just a few minutes it becomes clear if a person has the condition they say they do. In fact, when there is a fraud, usually the phone number they provide doesn’t go through. When that happens, we don’t accept them.
We also use unique links for each respondent in our studies, so one respondent can’t just pass that link around to others.
Finally, when our panelists complete a study, we pay them by check, not gift card. They need to give us an actual address to get paid. This is one final verification – it is much easier to get 100 gmail addresses than more than one real address to which their check is sent.
So please, don’t just assume respondents from all sources are real. Check them out. I would never say we couldn’t get a fraudulent respondent, anything could happen, but we take a lot of steps to prevent it.
