eCommerce Marketing Blog

7 Biggest Mistakes in Ecommerce A/B Testing

Posted by Vaida Pakulyte on Jun 14, 2016 5:00:00 PM


As the saying goes, it’s best to measure twice, cut once. When you know where to begin testing, you can make the best use of your time and effort when it comes to analyzing results and making changes. However, when you do something for the first time, there is bound to be a learning curve.

Here are the most common mistakes of running A/B tests on E-commerce websites, with suggestions for how to avoid them.

Running Tests Without Enough Background Information

Naturally, you’re eager to start your A/B testing campaign. But unless you take the time to analyze your website data and make realistic hypotheses, you may be in the dark about what exactly you need to test, and how to analyze your results. Before you start, be sure to spend some time getting to know your website’s problem areas, familiarizing yourself with the analytics and customer shopping trends, an even performing a usability audit to understand person’s perspective.

Running Too Many Tests at The Same Time 

It’s best to run only one A/B test at a time. That way, you can be sure to control for all other variables, and ensure that any results you see are due to the elements you were testing. If you have too many tests going at once, you run the risk of muddling your final outcome.

Wasting Time on Inefficient Testing

Certain elements are really too small to result in any measurable results. These include very minor copy changes (such as changing “a” to “the”) or insignificant design changes (changing a graphic from dark green to light green). Instead, try out more drastic tests to in order to get the more impactful results.

Giving up After Your First Test

It’s okay if your first test doesn’t yield the results you expected. Look over the data and plan your next step. This may involve giving your test some more time, testing different elements on the same page, or moving to a different page and testing something else. As long as you’rlearning about your site and working toward improvement, you’re on the right track!

Ignoring Small Victories

When A/B testing, it helps to keep your expectations in check. There’s a difference between tests with zero results and those that indicate small improvements. Don’t automatically ignore those little wins. For example, adding product recommendations to a popular category might improve clicks by 5% at first. But test for another month, and your results may improve to 10%.

Not Integrating Data With Analytics

Many A/B testing tools have built-in analytics integrations, meaning that your testing data is sent to your analytics provider while the test is happening. This can help enhance your post-test analysis and also provide a back-up evaluation of your test results. This is a good way to keep all your tools on the same page while giving yourself more peace of mind.

Focusing Only on the Design

It’s easy to get caught up in the look and feel of your e-commerce site, but don’t make that your sole focus. Remember to use the data to evaluate the number of real conversions, and you’ll have a better sense as to whether changing that font style really did encourage more sales. 

When making improvements to any of your digital platforms, implementing A/B testing is a smart way to analyze pages, emails, and content by comparing different variations and choosing the most successful version— that is, the one that encourages visitors to become customers.

Ab Testing Ebook


Topics: Ecommerce Marketing, Segmentation, Shopping Cart Abandonment, Email Personalization, ecommerce segmentation, email marketing

Subscribe to Email Updates

Posts by Topic

see all