Comments on Testing Incentives: The Best Way to Pay

Kirsten Robinson — Tue, 02 Dec 2008 15:31:12 GMT

Mary,

I don’t know about all countries outside the U.S., but I did run into this problem working with remote usability participants in Germany. Apparently compensating them (even the small amounts for a usability study, and even in non-cash equivalents) would lead to complicated tax issues, so we were unable to do it. We finally determined that it would be OK to send each participant a Christmas gift (our study was in the fall) of the new software release that they had helped test.

—Kirsten

www.dynamicdiagrams.com

bird lavonne — Wed, 28 May 2008 09:46:15 GMT

it is sometimes difficult to make sure only true participants get the incentive.
—————————————————-
bird.lavonne

Didn’t do much with them to test. I got a check yesterday from them for $38 bucks.

I will be using them more now that I have gotten paid.

check it out yourself
www.PaidWebSurfing.com http://www.PaidWebSurfing.com

Mary Hettinger — Thu, 07 Jun 2007 19:10:50 GMT

Can anyone tell me if offering incentives to increase survey response rate is prohibited by law in countries outside the U.S.? Our research firm is telling us that this is illegal practice in all countries outside the U.S., but I find that hard to believe.

Thank you.

scott evangelou — Thu, 22 Mar 2007 08:41:38 GMT

Great article, Cliff.

Our rule of thumb is to compensate test participants around NZ$50 per hour in the form of a food voucher; for example: a gift certificate to a restaurant or grocery store. Food is always appreciated and, well, most people need it :)

David C Dunkle — Fri, 16 Feb 2007 20:28:11 GMT

Another practical topic to consider: the security of your incentive method.

Cash and gift cards have a level on anonymity that goes with them from test to use. Once those incentive items disappear from your lab, it’s hard to know who really has them, or who really spent them. Of course, it’s not difficult to make sure your participants do get paid, however, it is sometimes difficult to make sure only true participants get the incentive.

Specifically, I was once leading a team when one of our members was quietly stealing gift cheques meant for participant compensation. Because of the ease in “repurposing” gift cheques, and also due to the laxity in our bookkeeping, this continued for over a year, and added up to several thousand dollars.

Surely, a stricter accounting policy would have helped, but personalized checks would have also eliminated the problem.

Jon Nakasone — Tue, 06 Feb 2007 19:41:01 GMT

With the cost of gas these days, it can be difficult getting anyone into their cars for anything other than cash. I have no issues with this, but the problem has always been with finding the appropriate compensation amount for tests of varying sunstance. For instance, there are times when a usability test may last 90-minutes and dive deeply into multiple tasks, etc. Other times, we need to quickly test users (30-minutes) on a single (yet mission-critical) bit of interaction. In both cases users had to travel the same distance and expend the same amount of fuel. The only variant was the amount of time spent with actual test partcipation. Is it fair to compensate the 30-minute test subject the same as the 90-minute test subject?

My experience has been that compensating test subjects for varying lengths of tests is akin to buying a sofa set. A loveseat, though fully 1/3 smaller than a couch, is just a wee bit cheaper. This is because it’s the ends of each piece (round-trip transportation) that are costly to make, not the cushion space between them.

Mary Deaton — Tue, 06 Feb 2007 17:33:47 GMT

But how much? I tell my clients, who pay the incentive, that the amount must show respect for the participant’s time and indicate how much we value their opinion. And different users need different incentives. Highly-paid IT professionals need a larger incentive than students because their time is already paid for at a high rate and I am competing with their hourly rate. Students, of course, need the incentive more, but usually are happy to get whatever is offered. I have also found that the amount of the incentive has a direct correlation to the number of drop-outs and no-shows.

I try to give cash or an Amazon certificate this is posted to participants each day at the end of testing; immediate gratification is especially welcomed by students, retirees, and people in average income jobs.

Rebekah Sedaca — Tue, 06 Feb 2007 16:27:48 GMT

I have run unto some issues with using any incentive with government employees. Due to some recent scandals in our area with local and state officials, there is reluctancy to except anything from an outside entity. Luckily, participants have been willing to participate sans compensation. Did you receive any response from your survey that had tactics for compensating government employees?

I appreciate this article and will look into some of the online gift certificates referenced for some upcoming remote testing. Thanks!

Michael Beavers — Fri, 02 Feb 2007 13:25:48 GMT

I think that cash is really the best incentive for users in a lab setting—hands down ($75-$100 per lab hour, depending on the market). However, there is an intangible compensation with many respondent populations that I’m not sure how to get around: the need to share an opinion, be heard, and interact with people.

Regardless of the city or the screening demographic or behavior, one thing I’ve never been able to shake is the professional respondent. These are the people for whom part of their livlihood depends on being a research subject, or a primary outlet for social interaction. We also see that many respondents aren’t “heard” in other aspects of their lives, and it can be quite sad. I realize this post is not entirely related to your article, Cliff, but could make a good story idea if any researchers on B&A have found a good way around it—or if it is really necessary to find a way around it.