Your Last-Minute A/B Testing Resolutions

Your Last-Minute A/B Testing Resolutions

Best PracticesQuick Tip

It’s the end of January, which means you might have already given up on your New Year’s resolution. Don’t sweat it. We’ve come up with a few new resolutions for you that we promise will be easier to stick with than giving up coffee, or sugar, or Netflix. In fact, we’re not asking you to give up anything–this is all about implementing new testing strategies in the new year!

Top Testing Resolutions for 2016

  • Let the data speak for itself. We were pretty bummed that the AFL-CIO audience didn’t respond very well to our Harry Potter quiz email, and instead preferred a very standard, totally Harry Potter-free reactivation email. But the numbers don’t lie. If the people want emails that are straightforward and to the point, that’s what they’ll get. At least, that’s what they’ll get for the AFL-CIO reactivation list–which brings me to our next resolution…

giphyOur staff after learning that the Harry Potter email had a .1% action rate

  • Remember that your list is unique. There are a million articles about testing best practices, the secret to writing the best headlines, the most shared articles on Facebook, whatever. And there’s something to be said for trying those tips out. But ultimately, success is all about what works for your list, not what worked for Buzzfeed. Maybe you have a list of older folks who respond best to larger email text, or maybe your audience loves gifs. Figure out what works for your audience, and then test based on that.
  • Keep track. Make a Google Doc of tests that you’re eager to try out (may we suggest ShareProgress as a platform for trying them?), then track your results.
  • Make a friend. Not just any friend–a testing friend. Find someone else in your office to help you brainstorm new ideas, track what’s working in that handy Google Doc you just made, talk over the results, and test some more. If you get stuck in the “coming up with ideas” stage, we even included some testing ideas to get you started.


  • Test everywhere! One big insight in our 2015 round-up was that the same campaign can perform very differently across different social media platforms. Try a variety of different tests on a variety of social channels for the same campaign to figure out where your audience responds the best.

And as an extra bonus, we’re arming your new testing swagger with a few experiment ideas. If you try them out, let us know–we’d love to share the results!

Testing ideas

  • Exclamation points in Facebook headlines — Does adding an exclamation point increase excitement and social sharing?
  • “Sign This Petition” before the petition name — We’ve found that this trick has performed well across the board at increasing actions. But we know from the resolutions above that every tip isn’t for every list, so try it out for yourself!
  • Facebook images with the campaign description text — Does it help to have the text from your campaign included on the banner image you’re using?
  • Something totally ridiculous — So our Harry Potter quiz didn’t work. Doesn’t mean yours won’t! I would love to read about a totally wacky idea that also got people engaged.

So go forth, into an internet that’s just waiting to be A/B tested–and let us know what works, what doesn’t, and what you learned along the way.


Written By

Anna Schmitz