Guest article provided by: mrwpress.com
Among other advice for new business owners running an eCommerce website, we always recommend running AB tests. But what does that mean? AB tests are basically comparing two different versions of a specific element on your website, like a block of text or a button or even a section’s background. You serve option A and option B of that element randomly to different visitors to your site, and then compare the results once you have a good sample size. This is an immensely effective way to boost your own conversion rate, since you’re not following broad principles but instead acting on real data from your site itself. Take a look at some of our tips for best practices in AB testing.
Only Test One Piece at a Time
One of the biggest temptations when running AB tests is try changing everything at once, and get as much data as possible. But this is a mistake because it blurs the data. If you change both a button and an image in the same test, can you say with 100% certainty which element actually gave a boost to the conversion rate? Maybe button A and image B work best together, but it’s much more difficult to tell that because button A and image A didn’t work together at all. Be patient, and only test one element at a time for the most accurate and impactful data.
Choose a Specific Test Goal
We’ve pressed on conversions being the end goal, and while that’s likely the end goal for your site as a whole, it may not be the direct goal for every specific element you test. Maybe you have a complex product, and your immediate goal for a button on the homepage is to get users interested enough to learn more. Even if it doesn’t lead to a direct conversion, that interest may plant a seed that results in a sale later down the line. Not that there’s anything wrong with using conversions as the test goal of course, but be intentional in your choice for that goal.
Wait for a Good Sample Size
Another time to be patient! It’s easy to make bold assumptions based on just a day or two of data, especially if one option seems to be running away with the results. But schedule a certain timeframe, and let it run its course. Maybe the type of person who browses your sites on the weekends reacts differently than the weekdays (valuable insight in and of itself). Or maybe that early lead was just coincidental. Let the test run its full course to get the most accurate data.
Look for Deeper Data Insights
We hinted at this in the above section, but don’t look simply for conclusions like “option A performed better” and leave it at that. Look at additional factors like the day and even time of day, or the time spent on page and bounce rate between tests. You can learn about your audience and their patterns with this data, as opposed to simply treating them as statistics in your conversion funnel. Sometimes the losing AB test option may having winning insights.
Leave Your Bias Behind
Another common mistake in AB testing is to try morph the data to fit your views. All people have their preferences, and maybe one button tagline grabs you personally much more than the other. But don’t let this skew your view on the data. Try to look at the results objectively, leaving any personal opinions behind, and you’ll glean much more useful insights from your data.
Don’t Stop at One Test
So you’ve completed an AB test, analyzed the data, picked a winner, and made changes to your site. Fantastic! Time to start planning the next AB test. These tests aren’t a one and done improvement, but an ongoing process of boosting your conversion rate and improving your eCommerce site as a whole. You don’t have to test the big call to action button every time, you can change blurbs of text or background images or just about anything else on your site. You may be surprised by the impact certain elements can have on your users.
Pick the Right Tool for the Job
Whether your hosting provider has an AB testing feature built in or you need a WordPress plugin to get the job done, do some research and pick the right tool for the job. You want the tests to be easy to implement so you can run lots of them, and present data that’s important to you so you can rest assured in these important choices.
Web development is one of our services, and we’d be delighted to help you start running AB tests on any platform. Contact us today to get started!
Image source: unsplash.com