(This post was originally posted on www.EcomTricks.com by Daniel)
AB Testing - Test big!
I've often had people ask me, 'Where do we start with AB testing?' and 'How should I go about testing?'. These are fantastic strategic AB testing questions and the answer will vary depending upon the business that you work for - every situation is different.
One thing which is consistent is the logical approach to AB testing.
If you want to get to the optimal version of your website or website quicker, it's best to test in extremes and 'test big'. Experimenters do this when AB testing to find their optimal version quicker than they would have if they performed small, iterative tweaks with their AB Tests.
An example of this logic is to imagine your page you wish to optimise in the centre of a grid. You can change the page in any way you like and this will take the page in a new area of this grid. Your aim is to move towards the optimal version of the page, but you don't know where that is.
When most people conduct AB Tests, they test in small steps. They make a couple of small wins, then get distracted and move to another test on a different area of their site, forgetting their strategic approach on the page altogether. Even if the AB tester did stay on track, testing small takes a long time to find the optimal:
Best practice suggests you should 'test big' and tests opposites to ensure that you reach your optimal version sooner than you would have through small minor tweaks:
Conducting AB Testing using this approach ensures that you reach your optimal version of the page quicker than you would if you were ‘testing small’.
Optimizely, VWO and Maxymiser are some great AB testing services you can take a look at to get started.
Key Takeaway
If you want to find your optimal version sooner than normal, test big. Test extreme versions of your page and go from there to really get the most from your AB testing.
Over to you!
As always, it’s your turn – what are your thoughts on testing big?
