Your survey might be representative… but of what?

“We need a representative sample…” These are words we at IOM hear daily. But how many people really understand what a representative sample actually means? And more importantly, how can you verify that the research you are paying for is truly representative of the types of people whose views you want to hear?

Traditional panels have done a great job of persuading us that they can deliver a representative sample of the wider population. But does the simple fact that your sample is made up of people who have gone put of their way to sign up for a survey panel promising to pay them £1 for every survey they complete, not, by very definition, make them different from the wider population?

At Institute of Mums, we take a different approach to recruiting research respondents. Over the past couple of years, we have worked hard to build up a large social media presence that we can tap into and leverage for research purposes. We don’t typically incentivise our respondents (we find we don’t need to) and instead, use a combination of psychographic segmentation techniques and purposive targeting to gain access to parents who would not normally take part in market research.

We find this approach works well. When a client comes to us looking for a survey of 1,000 pregnant women, we are not constrained by the number of women who have previously opted in to our panel, who happen to be pregnant. Instead, we treat each survey as an opportunity to source new respondents, making use of our 80,000+ strong Facebook fan base to help us source fresh respondents who exactly meet our clients’ brief. (Thanks to the advent of so-called Facebook dark posts, we’re not restricted to our own fan base either).

This approach has lots of benefits over traditional survey panels, and we believe, leads to more representative results. A simple example, but how likely is it that a GP with kids who spends half her day filling out forms will sign up to answer surveys for the promise of £1 a pop? On the other hand, if a survey happens to pop up on her newsfeed while she is browsing Facebook in the evening, then might she be tempted to answer a short survey that is directly relevant to her? Quite possibly!

If you want your sample to be representative, then you must pay attention to how survey companies recruit their respondents. If their main source of traffic comes through a link on a website that promotes answering online surveys as a viable source of income, then it’s probably not the panel for you.

Don’t rely on the survey panel to tell you how representative their panel is. Instead, try to figure it out for yourself by looking at how they recruit their respondents, whether they have to bribe incentivise their respondents to take part in their surveys, and of course, how experienced their researchers are in understanding their target market and optimising their surveys to minimise survey fatigue and respondent dropout.

All survey panels providers will offer you a representative sample. The question is, representative of what?

At IOM Research, we work hard to deliver the premium our clients expect from us. And we try to be as transparent as possible in this process, because we have nothing to hide. We live and breathe surveys, and have worked hard to find a way to make ours as representative as possible, actively sourcing respondents through social media as and when we need them. As such, we guarantee that the results you receive will be of the highest quality possible – and as representative of the wider population as we believe it is possible to be.