3 ms·
I have one data point for you: for the last two weeks I've been collecting user surveys. I want people to participate, but was unsure whether altruism or incen
by patio11 17y ago
I have one data point for you: for the last two weeks I've been collecting user surveys. I want people to participate, but was unsure whether altruism or incentivization would work better. Enter the A/B test: group A got told "Do this and I'll give you an increase in your print quota", group B got told "This will help us improve the product for our users." (Everyone who took the survey got the goodie.)
This has been the most decisive A/B test I've ever run. 2.62% of incentivized users in A convert, versus a mere .73% of non-incentivized users. That's waaaaaay pay 99% confidence that incentivization works.
It is pretty trivial incentivization, too: I'm giving them more of something which is too cheap for me to even describe a way to measure how much it costs and, even more weirdly, essentially nobody is helped by the freebie. (Normal print quota is 15, incentive is +5, number of users printing between 15 and 20 is below miniscule.
- toddh 17y agoYou have to incent repetitive, boring, skillless tasks. Altruism might work if there was some higher purpose involved, but that's hard to imaging for a survey. So you are left for paying for results.
- ryanelkins 17y agoI'm interested in finding out more about what kind of incentives drive people and what kinds of returns businesses can get from them. Does anyone else have any good examples of this kind of behavior in action?