The research does not address the more important aspect — how does an incentive change the responses (or does it change the responses?). Just getting more cooperation is not the goal — the goal is to get a representative sample of the population you are interested in giving accurate/true answers. The third MarketingCharts article suggests the larger samples were no more accurate.

The level and topic of the survey can drive the need for incentives, I’ve found. I’ve also seen that the majority of survey questions are not well-written and need to drive better, actionable insights. Once the survey format and content is really optimized in a way that encourages deeper responses from the audience, incentives tend to be less required.

Back in my day there were consumer panels where a participant wrote down all their purchases. What they got for this ridiculous activity was a coffee pot or some such thing.

Every quarter we would get a presentation from the research provider (Nielsen?). And our people would pay rapt attention to the behavior these people indicated.

I always had a problem with this. “Are we really going to base our marketing decisions on people who write down every one of their purchases for a coffee pot? In no way is this a typical consumer.”

I see incentives the same way. The data shows that you can increase the response rate by about 50 percent if you give an incentive. But most importantly, it doesn’t indicate if those additional participants reflect the same population in terms of behavior and answers as those without the incentive. If they are different, then the incentives skew the data. If they are the same, then you don’t need the incentive.

I’m not a fan of professional respondents and, therefore increasingly, I’m not a fan of surveys. If the object is to get a reward the incentive is to speed through the survey. So while it improves response rates I think it degrades overall quality. In terms of what’s most effective — cash or gift cards.

I need to assert that respondents overwhelmingly want to convey their true attitudes and behavior. This idea of respondents not speaking the truth troubles me because I’ve spent a career managing the process of bringing insights to clients with solid marketing research. At the same time, I know that if you present questions that offer grids asking respondents to rate 25 items and do it over and over again in one survey, fatigue, frustration and annoyance will take over and respondents will go the path of least resistance. If you ask consumers to estimate how often in the past year they’ve done something that’s hard/impossible to estimate, they’ll just come up with whatever pops in their head, wishing they could be accurate but convinced if someone cared the question wouldn’t be so ridiculous. No incentive will overcome the barrier of a poorly designed survey. (I can show you a video with real live respondents talking about their concerns and frustrations with surveys.) I started my marketing research career as a Marketing Research Field Director, responsible… Read more »

Our work consistently shows consumers prefer mobile surveys over responding by phone, and those with an incentive see higher response rates than surveys without. More importantly, for surveys with an incentive included, brand-sponsored incentives, such as a discount on a next purchase or a chance to win a high-value gift card outperform general sweepstakes.

In countering survey fatigue, we advise brands on starting with Net Promoter Score (NPS) as the basis for the survey, then branching to just a few additional questions for those respondents indicating they are Detractors or Passives. The key is keeping the survey concise. Another important consideration is that of avoiding over-surveying customers, such as not inviting their feedback after every brand interaction.

We’ve curated our thoughts around NPS surveys here.

I don’t have an issue with incentives for survey respondents so long as they are modest. A $5 gift card really feels like you’re just compensating them for their time. I worry less about whether accepting the incentive somehow skews the results as I worry about the selection process for the set of respondents. If you’ve followed a good process to select your target audience, the small inventive shouldn’t have a tangible impact on the response quality. Given that everyone seems to be bombarded with survey requests after every single purchase you make (when was the last time you got a receipt at a store that didn’t have a customer survey link on it?!) a little incentive goes a long way to towards gaining valuable feedback you might otherwise not receive.