The Most Active and Friendliest
Affiliate Marketing Community Online!

“Propeller”/  MyBid

A Beginners Guide To Split Testing | 3 Keys To Doing It Successfully

Stevens

New Member
affiliate
Landing pages, emails, and ads are all things you should be split testing if you want to maximize your affiliate marketing results, but if you rush into it without knowing what you’re doing you’ll end up doing more harm than good. So make sure you read this whole post, to get some key pointers to split testing successfully.

The first key to split testing correctly is to only test one variable at a time.

So for example, if you are trying to optimize your landing page you are going to start off by only changing one thing like a headline or an image.

You are going to split test like this because it allows you to know exactly what changes are improving your landing page, and which ones are hurting it. If you were to test out multiple variables at a time you wouldn’t really know what changes were making it convert better, or worse, which would make it extremely hard to get to your ideal landing page.

Think of it like this, let’s say you change your headline and your main image, and the split test increases your optin rate by 1%. Now you might think this is a good thing, but what you might be missing is the fact that maybe your new Headline is actually worse than your previous one, but your new image was a ton better, so you still ended up getting a higher optin rate.

But if you had been split testing correctly you would have been able to know that your new headline wasn’t doing that well, and you would be able to test a different headline out that could take your split test from a one percent increase in conversions to a 5% increase in conversions.

By A.B. testing correctly you will be able to quickly optimize your landing pages ads and emails, but if you’ve ab tested all the main elements on your landing page or ad and you’re still not getting the conversion rate you need, you might want to consider multi variant testing, which is a different style of split testing.

With multi variant testing as the name says, you will be testing multiple variables at a time. I recommend doing this once a month, so you can test if a completely different layout of your landing page is going to convert better than your current one. With A.B testing, you may have found the perfect landing page, for one lay out, but there might be a completely different layout that will increase your conversions even more, so make sure you use multi variant testing every so often.

The second key to split testing correctly is to only make a decision on which variant is the best when you have enough data.

A lot of people who are just getting into affiliate marketing will start split testing their landing pages when they aren’t getting a lot of traffic, which is fine, but they run into problems when they start making decisions on which variant is the best too soon.

If you are only getting let’s say five leads per day, you are going to need to run that split test for at least two or maybe even three weeks, to ensure that the version you pick is actually the best one. If you make a decision to early you could end up picking the wrong one.

For example, if you were only getting 100 landing page visitors per day, and you ran a split test for one day, and one version had a 30% optin rate, and the other one had a 50% opt in rate, you might assume that the second landing page is the winner, when in reality the other page was going to end up out performing the original one, but you cut it off too soon and now you are stuck with a less ideal landing page, and you don’t even know it.

It basically all comes down to the law of averages, you need to run enough traffic to your split tests to ensure that your opt in rate is actually what you think it is. At the start of the split test you’ll see your optin rate jumped from 10% to 20% back down to 10% and up to 30%, but as you run more traffic to it you’ll slowly get to the average optin rate of let’s say 20% for this example, and only once you get to that point are you going to want to make a decision.

I already kind of touched on this last key when I talked about multi variable testing, so I’ll keep it short, but you need to make sure that you aren’t giving up after one test.

Just because you make a change and it increases your conversions by a significant amount, let’s say 5%, doesn’t mean that there aren’t more changes that you could make that can increase your conversions even more. You could change the headline and notice of 5% increase in conversions, but maybe if you test another headline you’ll find that that one converts even better. You should always be testing and improving your landing page, because unless your conversion rate is 100%, there’s room for improvement.

That’s it, those are the three keys to split testing correctly! If you found this post helpful at all, I’d appreciate an upvote, thanks!
 
MI
Back