Tuesday 18 February, 2014

Split Testing – The Science of A/B Testing

How do you know if your website is working or not? Many people do it through intuition and reasoning, but successful webmasters figure it out and make their decisions through hard scientific data. We get this data by testing the website and the way to do it is through A/B split testing.

What Is Split Testing?

Split testing means creating two different versions of your website with just one element changed. You run both versions of your site and see which one performs better. Now, you know which element to keep.

For example, let’s say you have a sign-up form on your website. You’re not sure where to put it, so you split test. One version has the form on the left side and the other has it on the right. You run both versions and find that the version with the right-hand sign-up gets more opt-ins. Now you know that the right side works better than the left.

Why Split Test?

It sounds like an awful lot of work but it’s well worth it in the end. The reason is that it yields objective data that’s much more reliable than what you think you know. We often have ideas that we get through our own common sense or common knowledge learned from other site owners.

The problem is that this information isn’t always correct. Plus, there’s a great deal of variation from one niche to another, and even one audience to another. You’ll only know exactly what works for your type of site visitors by testing your site.

What You Can Test

You can split-test absolutely any element of your site and the more split testing you do, the better. In fact, you should run tests on a regular basis to continually tweak your site. You can test anything but a few common areas site owners test include:

Headers – You should test not only your header’s size, font, color and position, but also the actual wording. Write two different versions to see which works better.

Colors – A slight change in color can make a huge difference to your site visitors. Test your overall color scheme, main colors, background colors, text colors – anywhere you have color variation on your site.

Forms – Always test your opt-in or other forms, just like the example above. Like your header, play around with different wordings to discover which is the most compelling to your audience.

Social Media Buttons – You want to put your social media buttons in the best possible place where they’ll get lots of clicks. Experiment and test until you find the sweet spot.

Call to Action – Calls to action like your ‘Buy Now’ button need to be tested. These are vitally important features that are small but make a big difference.

In addition to your website, you can split test email messages and any other kind of marketing you do.

How to Split Test

Most split testing is done using specialized software that allows you to create and run multiple versions of the same website. Google Analytics has a split testing feature and some programs will also analyze the results for you.

Bob Steele

Bob Steele

Bob Steele is an entrepreneur, software developer, marketer, and author living in the Denver metropolitan area. He’s an avid outdoorsman who loves skiing, hiking, fishing, boating, and just plain having fun. His interests include games, space, technology, physics, cooking (well eating actually), economics, business, internationalism, and team sports. With over thirty years of professional consulting experience, Bob has been exposed to many diverse business models and has gained a sensible approach to life. Bob’s company, WaveCentric is focused on commerce, marketing, and entertainment related products.

More Posts - Website

 
Comments Off on Split Testing – The Science of A/B Testing   Posted In: Commerce, Mobile Computing   |    Written By: Bob Steele