Once a term that no company really quite understood 5 years ago, has now become commonplace. It seems like online survey panels are popping up everywhere in the market research world. Not only are market research companies developing their own in-house panels to rent out for a quick buck, but companies are developing their own customer panels to tap into for their own research needs. Panels are everywhere and you may begin to wonder, how do I choose one? Drive Research is here to help.
Understanding search traffic around online surveys and panel surveys paints a clear picture of the market. You'll see tons of searches around "get paid to take surveys", "paid surveys", etc. which tells you some percentage of the people signing up for these online panels are simply money grabbers. These money grabbers sign up for 10+ online survey panels across the country and take as many surveys as possible. As you can imagine, the credibility of these respondents comes into question. Learning about what qualifies and disqualifies participants from surveys, allows them to game the system a bit. "Ate at McDonald's in the past 7 days? Sure, absolutely, even though I haven't been there in over a year."
When searching for your next online survey panel company be wary of online panel companies who fit the following profile:
Have no quality control plan in place - ask your panel company about the make-up of the panel, how they are recruiting, what is done to screen qualified participants. Even go as far as asking for their quality control plan for their panel. Panels that are made up of a grab bag of respondents will result in your outcomes being significantly skewed. The only thing worse than doing no market research at all, is getting bad market research that sends you down the wrong road.
Do not regularly scrub their panel database - be wary of those who market the "number" of people on their panel. Typically these panel companies care more about quantity than quality. Those who care about numbers will not regularly scrub out bad participants or those who game the system because they fear they'd have to take a remove a significant amount of people. "You're telling me 3,000 of our 10,000 panel members failed our quality control test? Leave them in anyway, 10,000 sounds better than 7,000." With research I'd rather have 300 representative and quality completes than a grab bag of 600+ completes. In that scenario, more than half of the responses will swing the findings in a questionable direction. You'd be surprised to learn many panel companies do none of this. I still receive invites from a panel company I signed up for in 2006 on a Yahoo mail account that I haven't checked since 2008.
Do not consider weighting data - one of the limitations of panel surveys is that it is a convenience sample not a random sample. It's a major difference. You cannot apply a margin of error to a convenience sample. To account for these abnormalities in data, the analysts should consider weighting the data to make it representative of the population. For example if you know 20% of the market is 65+ in age but your panel survey yields 4% of completes from that age group, you need to weight the data to account for the difference. Without weighting by market, age group, or other known data points it's like getting a hand-count of a 3rd grade class on whether they like Coca-Cola or Pepsi and claiming the hand-count is representative of the U.S. population.
Drive Research is an online survey company located in Syracuse, NY. Our market research company designs surveys, programs surveys, hosts surveys, finds participants, manages fieldwork, collects data, analyzes results (and most importantly) provides our clients the insights, strategy, and next steps using a data-driven approach. Questions about how Drive Research can help with your next online survey project? We work with a number of qualified and credible online survey vendors across the country. Email us at firstname.lastname@example.org.