If you’re looking to improve your website’s conversion rate, AB testing is one of the best ways to do it – if your site is attracting more than 100,000 sessions per month. If it attracts less, it’s likely to take too long to achieve statistically significant results – so following proven guidelines and best practices may be the best approach.
If you’re still keen to get into AB Testing, before jumping in, you need a few things in place to make sure your tests work and will deliver results that will optimise conversions. In this blog post, we’ll walk you through some key steps that will help you prepare your site for AB testing.
Solve Technical Issues
One of the most important things to do before you start AB testing is to make sure that your site doesn’t have any technical issues. If your site is slow, has broken links, or isn’t mobile-friendly, you’re probably not going to get the most out of your tests just yet. Countless studies continue to show that slow site speeds and error prone sites, kill conversion rates. So there’s no point optimising your checkout experience if you really need to spend time and effort resolving fundamental technical issues. Spend some time fixing any technical issues on your site before you start testing!
Pinpoint which funnel and metric you’re trying to Optimise
Another important thing to do before you start AB testing is to make sure that you’ve identified your key user journeys, know which metric you’re trying to improve and why. There are a variety of methods used to identify key user journeys, but they fundamentally come down to finding the most travelled journey of your most valuable customers towards the most important goals that impact your company’s success. Although tempting, it’s best to avoid covering every scenario. If you need help identifying your key user journeys we can help.
Make sure your analytics data is reliable.
It’s also important to make sure that your Google Analytics data is accurate. If your performance data is skewed by irrelevant traffic (non-humans, customers outside of your serviceable area, etc…), or erroneous events or a flawed goal setup, you’re not going to have a reliable baseline to start from and you won’t be be able to trust the results of your tests. Take some time to check your Google Analytics set-up and make sure that it’s configured correctly.
Take care of fundamental UX/UI & Conversion best practices.
Once you’ve taken care of the technical, identified key user journeys and have truthful, reliable data, you’ll need to make sure that your site is following core UX and UI design guidelines with your current design. In other words, making sure that the things you are doing are being done correctly.
After you’ve made sure that your site is using fundamental conversion guidelines correctly, it’s time to look at deploying advanced UX and conversion opportunities – or in other words, adding proven elements and features that you’re not doing.
There’s no point spending time, effort and money on tests if the problem you’re trying to optimise or feature you want to add is already covered and published as a guideline by leading UX and Conversion research bodies like the Baymard Institute, MECLABS or Conversion Kings.
If your site’s UX, UI and Conversion tactics are violating fundamental, well known, evidence-based best practices – it may be best to go ahead and fix them before designing and building your tests.
The caveat here is that if your changes are on pages with high volumes of traffic, in a conversion journey, and likely to have a significant impact on a customer experience or brand, you may still want to test them. This is where it’s best to review your strategy and backlog with an experienced digital conversion strategist.
Now you’re ready to start testing.
Once you’ve done all of these things, you’ll be ready to start AB testing. By following these steps, you can make sure that your tests are accurate and that you’re getting the most out of them.
What other steps do you think are important to prepare for AB testing? Let us know in the comments!
