- Contest Factory
- February 26, 2019
Split Testing Basics
At its core, split testing is a simple concept that can yield very powerful results when leveraged properly. It’s a technique that allows you to test out and compare different versions of something — be it a website, an ad, a call to action or similar element — to determine which version ultimately performs the best for your target audience. By comparing the results generated by the variations, you can see what people respond to and what they don’t — thus giving a significant boost to your knowledge of what works best to earn more conversions.
But many people don’t realize that the effectiveness of split testing goes far beyond simple images or font choices. You can split test nearly anything about your ad copy, landing pages, and other collateral, and with the following key tips put into practice, your split testing is sure to be a success.
The Art of Split Testing: Things to Focus On
Before any split test begins, you must first begin with a reason. Are you dealing with a situation where people are adding items to their shopping cart but are eventually abandoning them? Are people not scrolling all the way down to the bottom of your landing page to reach your offer? Is one particular page on your website shattering your bounce rate and you’re not sure why?
Begin with a reason to conduct the test in the first place, then move on to your hypothesis. What is happening, and what difference do you hope a change will make? This will give you an adequate metric to weigh your success against, and it will also help prevent you from making random changes, instead giving the entire affair a larger purpose for which to strive.
Split Testing Ads
As stated, split testing is all about dividing your audience into two separate-yet-similar groups who are then shown materials with subtle variations. If you were split testing a PPC (pay-per-click) ad, for example, for the best results you would want both ads to be identical, and test variations about their deployment.
You might decide to show the ad to two different audience types, experiment with two different placements, or choose different delivery optimizations. The key point to remember is that you should only test one variable at a time. Do not change both the copy of your ad and the ad placement at the same time. Pick one or the other and run the test. Once you’ve collected your results from one test, test additional variables in separate, isolated instances.
Landing pages are another popular source for split testing, because by design they’re more conversion-centric than most of the other materials you’ll be creating. You might choose to split test two different versions of a landing page for the same offer — one at 1,000 words and one at 500 words. Or, you might consider a test with several H2 headers versus one with very few. You could even test out the frequency, placement, and tone of certain calls-to-action — the choice is yours, but this level of experimentation can help take an already strong page and make it even more effective.
Emails are also a popular target for split testers everywhere, because they too are very conversion-driven materials. Once you’ve settled on email copy that you’re satisfied with, consider testing out variables like the subject line, included images, and the CTAs. You can even experiment with the tone of those CTAs — one might be very direct and could legitimately ask for the purchase, while another might be more general and friendly, featuring copy such as “contact us today” rather than “click here to buy.”
In conclusion, split testing is a tried and true practice that uses the science of trial and error experimentation to yield hard data that shines light on what kinds of things your customers respond to the most, and thus allows your email, ad, website, or anything in your business to operate at a more effective level.