Subscribe
Sign in
GoodUI Evidence — People's A/B test results that show which UI ideas work best
14
Replies
@jlinowski I don't know if you could trust them outright, but this looks like a great starting point. Especially when trying to convince a boss/client to simply *try* some A/B testing. Often the biggest hurdle I find when A/B testing is getting the buy-in I need to try it in the first place.
@clarklab @davidiwanow Agreed that the data is still thin. I wonder what would be the reaction if for each idea there were 10, 20, 30? tests? It's a live project of course and we'll be adding more tests for sure.
@clarklab @jlinowski How can anyone argue against A/B testing? There's no losing to be had--you'll always be optimizing. Are people actually saying they don't want that?
@mbmusgrove @jlinowski budget, timelines, staff/skills, arrogance, etc. I see it so, so often. It's insane.
@jlinowski Thanks for submitting this. I've been a fan of the Good UI newsletter since I signed up a couple of years ago!
Interesting enough idea to crowd source the data, I think it's a great platform for inspiration of what you can test... less so taking the numbers without doing the test yourself....
Seems odd that having social proof decreases conversion rate. Any insights on the methodology of the tests?
@topdesignme The test you are referring to (https://www.goodui.org/evidence/...) is just one dealing with Social Proof. I'm sure once we start adding other ones, we should see some positive ones as well. The -% in Test003 could be a result from negative comments.
Crowdsourcing of A/B testing. It's all about the scalability of tests and # of submissions. It would be interesting for you guys to collaborate with other A/B testing tools in the market and aggregate anonymous information. Rock on!
@will_lam I'm sure we'll share a mobile test if someone decides to submit one as such. :)
Jakub Linowski