HackerOne users: Testing against this community violates our program's Terms of Service and will result in your bounty being denied.

How to do multivariate testing on different versions of forum elements?

Hey all!

So I'm looking to do some multivariate testing on banner ads on my site (Google Experiments or Optimizely). But I can't really give an "alternate" page for visitors to go to since what I'm actually testing is the template page itself that shows on all the pages across the site.

I'm not a programmer, so I'm hoping you all will be able to help me out!

I'm currently using the Pockets plugin — http://vanillaforums.org/addon/pockets-plugin — to put messages throughout the site, including the banner ads that I ultimately want to test different versions of.

Is there a way, either through directly editing a new version of the master template itself and offering that as a variation or is there a way to use the Pockets plugin in cahoots with a multivariate tester?



  • Hey thanks x00! It looks like that's for getting more precise on the analytics side of things? How do you see this tying into multivariate testing specifically?

  • ads will/could have identifiers like id="AdNumber1" this could be used to track clicks on such ads. You an also categorise within goggle analytics in order to present A/B testing more easily.

    grep is your friend.

  • What type of ads are you using third party/your own, images/text?

    grep is your friend.

  • Okay, so what you're saying is I could tag the first ad variation "AdNumber1", then put it up live on the site, let it run for say a week, then switch it out with the second ad variation "AdNumber2" and let that run for another week, etc.? So I can't actually use any of those solutions like Google Experiments or Optimizely that serve different versions of full pages to visitors?

    (And to answer your next question, I design my own banner ads to use as lead magnets to get people onto the email list. I'm just looking to optimize the ads.)

    Hm, so it doesn't look like there's an easy fix to this dilemma ...

  • I found the solution! I'm going with Optimizely since it allows for domain-wide changes to templates. Google Experiments require actual separate pages to test. Thanks for your help, x00!

Sign In or Register to comment.