a/b testing is so important.  In many cases, it’s the difference between being a successful Affiliate Marketer and one who’s just so-so.  That’s why doing it correctly is of utmost importance.  An article from Neil Patel of Quicksprout.com, provides some very salient advice on getting the most from your testing.

1.    Make sure you have the proper understanding of what your data is telling you.  If you see a 23% jump in clicks, does not mean you see a 23% jump in conversions.  Of course, the increase is positive, just make sure you see it for what it is.  If you want to know if conversions have increased, make sure you test for that.

2.    We are preachers of making sure that there is enough data to make a conclusion based on it.  But, to be honest I never really had a consistent number of what that is.  Neil has much more experience than I, so when he says that he makes sure that he has at least 100 conversions and that the test runs at least two weeks, I’ll go with that.

3.    Don’t have unrealistic expectations of your changes.  Unless your testing photo and copy changes, don’t expect a color change to bring you huge results.

4.    Make sure you are actually testing something.  Recently I had a client that wanted to test two exit boxes with just a one word difference.  A very minor word.  That is not a big enough change to test.  You will see a difference, but it will reflect other marketing factors such as times, overall web visits and more.

5.    Don’t test too many variables.  It’s best if you test one at a time.  If you test more than one, you won’t know which revision was the cause of the data change.

6.    4a and 5a.  What you should be testing lies somewhere between numbers 4 and 5.  Don’t test too little and don’t test too many variables.

Proper a/b testing is imperative to your success.  Make sure you’re doing it right to get the most from yours.