If you’re not running a split test right now, you’re probably losing out on conversions.
“Wow there, steady on Matt! We’re all friends here!”
OK, fair enough, point taken. But truth be told, I’d be doing you no favours at all by sugar coating it up. And that’s exactly why I wrote this post today.
I’m going to show you why you should make A/B split testing a top priority, how you can use split tests to increase your sales and, more importantly, how to set those tests up (don’t worry, that bits a doddle! Even I can do it).
To begin with, I’m going to rewind a couple of steps.
What is an A/B split test anyway?
Want to increase purchases or email sign ups? Want to reduce your bounce rates? There’s a million different things you can measure your sites performance by. Which metric you use will be down to your individual website.
An A/B split test is something which can be used to test and compare different onpage tweaks which you think might improve it.
Essentially, an A/B test is a controlled test whereby you have 2 different versions of a page. One version is the original (the control) and the other is a copy of that same page, but with one element tweaked.
When a visitor lands on that page, they are randomly served with either version A (the control) or version B (the tweaked version). Since users are served at random, you should expect to find that roughly half of your visitors see version A and the rest see version B. I say roughly, because it’s served at random, so it can never be an exact science.
The result is that you end up with a conversion rate of each version of the page. You can then compare the 2 to see which measures up the best. Since the data is from users who have arrived on your site at the same time, on the same day, via the same sources etc, it’s completely fair to weigh the 2 up against each other. Rather than the alternative which would be to run one version of the page for 1 week, then swap it for the 2nd version for 1 week so that you can compare the stats. Doing it that way would introduce other variables beyond your control and possibly skew the data. A/B split tests remove that risk. And that’s why split tests rule!
How To Set Up An A/B Split Test
Now here comes the good news. A/B testing is a complete doddle to set up. You won’t need to know any coding or possess any real technical skills. It’s also completely free to set up. There are a lot of paid tools out there that can help you run your split tests, and those are the ones you’ll hear most tutorials talking about using. But what they won’t tell you is that you can do this completely free in Google Analytics. And I’m going to show you how, in 8 easy steps.
Step 1) Define your test. Decide what metric you want to try to improve, and define one element on the page which you will change in order to (hopefully) improve this metric. For Example…Let’s take my email sign ups. At the moment, I have a form in the top right of my sidebar of my home page, that you guys can all fill out to get my latest posts. Let’s say for example, that I wanted to see which colour gets me more sign ups – red or blue. To find out, I would set up an A/B split test to test the theory. Game on…
Step 2) Create an exact duplicate of the page you want to test, but with one element tweaked. Going back to my example, I would create a duplicate version of my home page, but with the colour of the signup form changed. So I have to home pages uploaded which are almost identical, but with 1 difference between them. When designing an A/B split test, you should only ever test one element at a time. If I had changed both the colour and the placing of the form in one go for example, it would have made it impossible for me to pinpoint whether it was the colour or the positioning which contributed the growth.
Step 3) Jump into Google Analytics and set up a goal to monitor your conversion rates.
Step 4) Still within Analytics, head to “Content” and then “experiments”. Now enter the URL of the page you wish to experiment with (i.e. the original “control” version of the page):
Step 5) Tell Google Analytics what you are trying to improve:
Remember – In the “Objective for this experiment” you can either select a goal (in this example case, I would select my “email sign up” goal), or alternatively you can select from one of the onsite metrics that you might be trying to improve – i.e. bounce rate, time on side etc.
Step 6) Fill out the URLs to tell Google Analytics where it can find the duplicate versions of the page you want to test. So in my example, I would enter mattsbackpack.co.uk and mattsbackpack.co.uk/index2. You can test more than 2 different versions of the page at a time, but as I mentioned earlier, you only want to be changing one element per A/B test.
Step 7) Finally, all you need to do is copy and paste a few lines of code into the header of the 2 pages:
Remember to make sure that your Google Analytics tracking code is on both versions of the page, otherwise GA wont be able to track the test properly.
Step 8) Hit start, and then sit back and relax for a couple of weeks while Analytics gathers some lovely data for you.
And that’s your A/B split test set up! Told you it was a doddle.
Once the test has been running for a couple of weeks (or however long you choose to run it for), you will be able to easily jump into your Analytics and compare the stats from the test to see which version of the page performed best against the metrics that you were testing.
Continually carrying out A/B split tests allows you to pinpoint elements of the page which can be tweaked in order to improve your conversion rates.
It’s something that is used extensively in squeeze page design for lead generation, and of course, ecommerce. But in reality, it is one of the most underused and most overlooked tools in any digital marketers arsenal.
So if you’re not already playing with A/B split tests, then hopefully this article has inspired you to have a dabble? As always, if you get stuck or have any questions, just give me a shout.
Rumour has it that clicking one of these shiney buttons makes magical dancing pixies pop out of your USB port and shower you in riches and lucky 4 leaf clovers. Of course, that might all be bullshit, but are you willing to take that chance? Don't miss out...
Don't worry, I've got much better things to do than send you spam. Neither of us need that.