Comments on: iTunes vs. Amazon as Survey Incentive

By: Evan Carroll

Evan Carroll — Sun, 15 Aug 2010 19:18:47 +0000

Fred, Thanks for reporting on this. With the informal research I do as an experience designer, I’ve always wondered what types of incentives encourage a greater response.

I have an anecdotal suspicion that there’s a relationship between the location (Amazon, iTunes, etc) and the amount of the incentive that’s manifested in the number of products that can be purchased.

Consider that $15 from iTunes can purchase 15 items, whereas $15 from Amazon might only provide one (or part of a) book. Considering this, I have to wonder if the incentive cost is fixed at $15, if the iTunes or Amazon card will be more effective. From anecdotal observation, I believe this might be true, but I want to test it at some point.

To your comment, I’ve faced all sorts of issues with purchasing iTunes cards. I use them for small amounts, but I try to avoid them and save the administrative trouble.

By: admin

admin — Wed, 21 Jul 2010 19:44:02 +0000

Kevin, you’re absolutely right – this design doesn’t let us get at these questions. Bob Groves has good work in this area.

Personally, my gut is that I’m thrilled to get 30% off an email solicitation. That’s up near RDD and mail solicits, which email almost always under-performs. Though the question is – was it the incentive or the TDM tactics?

It was shocking to me how difficult it was (administratively) to do iTunes cards. I can’t imagine that survey researchers are a big part of the iTunes gift card budget, but I’m never going to go through that hassle again when compared with the ease of Amazon.

By: Kevin R. Guidry

Kevin R. Guidry — Wed, 21 Jul 2010 18:39:02 +0000

Interesting stuff. But the real question – the one that’s very hard to answer – is if the incentives positively affected the survey. Did they increase the response rate? If so, did that increase have a negative impact on the quality of the data (i.e. did the respondents influenced by the incentives provide accurate responses or did they just fill out the survey quickly and inaccurately to get the incentive)? So there are not only issues related to non-response bias but also data quality. And what about the effect of pre-survey incentives vs. post-survey incentives? So many interesting questions! (Which is why I’m writing a dissertation focusing on survey response.)

I think there was some internal discussion at my research shop about trying to make available Amazon gift codes as survey incentives. I don’t know what happened to that idea but one thing you mention – the relative ease with which Amazon gift codes can be purchased and distributed – was a very important consideration for us.