Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 158 Stack Exchange communities.

What is meta?
Here's how it works:
  1. Any Stack Exchange user can ask a question
  2. The community provides support, votes on ideas, and reports bugs
  3. Your voice helps shape the way Stack Exchange operates

Does Stack Overflow do any optimization (AB testing or MVT)?

If so, what has been tested/discovered? Anything interesting? And using what software?

perhaps to find out how to get more registrations, or more questions, or more voting, etc.

Also, if you do optimize, what is your CTA from both visitors looking at answers, vs. logged in users?

share|improve this question

Over the last few months, we have started to dabble in some split testing. No multivariate at the moment.

We're using mixpanel, which is working out really well.

The only place where we have had time to try some different approaches is on the ask question page for unregistered users.

I'm afraid I don't have time to write it all up at the moment, but I'll definitely put a blog post on my list sometime when we have some great results to share.

Preliminary results show that people dislike fully signing up, which makes our lazy register process attractive. We've also discovered that having the word "login" anywhere near a register form seems to cause people to skip over it.

Our goal was to increase posts and also increase the ratio of fully registered users to unregistered users.

We're currently on round 10, but we might have a new baseline to roll out in the next few days.

As far as metrics, there's a lot to sift through, but just to throw out some numbers: Every registered user that visits the ask question page has about a 70% chance of asking a question. 30% for anonymous users.

share|improve this answer
Interesting...has the "guest" wording in this version resulted in any more completions/lower unregistered ratio? Clarifying what the unregistered part of the form seems important for all the people who end up with accidental unregistered accounts – Ben Brocka Feb 18 '13 at 15:36
Yes, that helped. The version you linked is our new baseline version. We were able to convert about 12% of question askers directly to a full account instead of an unregistered account. Net posts went up about 3%. On the separate issue of accidental unregistereds, I've been waging a quiet battle on that front. Through a combination of welcome emails, top nav bar reminders, and other reminders in various places we've managed to cut the unregistered portion of sign-ups by about 30%. It's very hard to accidentally stay signed up for an unregistered account these days. – Jeremy T Feb 18 '13 at 15:45
test ping ping ping – Kevin Montrose Sep 11 '13 at 17:42

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .