Why an Increasing Number of Conversion Professionals Prefer A/B Testing
A couple of years ago, several key employees at Webati were privileged to attend the first ever conversion conference that was headed up by Tim Ash. This was a gathering of some of the world’s most renown conversion experts. All of us were able to derive inspiration from the seminars we attended, but one of our most surprising findings was just how many of these professionals preferred to use A/B split testing over multi variate split testing. These experts weren’t just enthusiastic about it. They were avid advocates.
We were curious why so many of them preferred A/B testing.
Size Does Not Matter
Rationally, you would think that testing a larger amount of variations would be better than just a few at a time, right? As the famed Lance Loveday so bluntly put it, “Not if all the variations you test suck.”
Fans of multivariate split testing often brag that they are able to test hundreds and thousands of variations for a single webpage. The important question is whether or not these tests are actually valuable and meaningful. Bryan Eisenberg, who is often thought of as the godfather of SEO, is not a fan of multivariate split testing. It is his contention that it wears people out on testing and does not deliver the results you are looking for.
In his mind, experimenting with thousands of combinations for headlines, images, and so on is not a valuable marketing strategy. He likens it to playing the lottery.
Again, Lance Loveday believes that A/B split testing is the antidote to this random testing strategy. It is his belief that A/B testing is capable of “harnessing the power of large changes.” A/B split testing is concerned with looking at the larger picture overall, rather than just tweaking small details, like a font or color.
How do you define large changes as it relates to your business or targeted demographic? The answer is to start by asking the right questions. For example, what is your business’s ultimate goal? Who is your business’s targeted demographic? What course of action do you want visitors to your site to take? Asking such questions will allow you to develop a hypothesis about how to woo your target audience, and then develop handcraft methods of determining whether or not these hypotheses are valid.
In this manner of thinking, you could liken A/B split testing to playing a game of poker, rather than the lottery. Each hand is played deliberately.
How to Think In a Holistic Manner
When it comes to combinatorial thinking, humans just aren’t good at. Let’s say you took six of the best employees here at Webati. You gave the six of us three examples of headlines, subheadings, images, and calls to action. Even with all six of us working together, it’s going to be impossible for us to imagine all of the different combinations and to evaluate them effectively. On the other hand, if you were to give us three, carefully considered and planned webpages, we would be able to look at them instantly and evaluate their effectiveness.
A/B split testing encourages one to be innovative and daring with individual components and the whole page. Different layouts and content structures can be evaluated for their effectiveness. Think of it this way. Layout A might depict its story through a series of images, while Layout B uses text, subheadings, and bullet points. Evaluating the effectiveness of the comparativeness would be next to impossible with multivariate testing.
Another advantage the A/B split testing has to offer is that it will enable you to experiment with different page paths. For example, Layout A offers the pitch and call to action on one single page. Layout B offers a two step call conversion process. The results that A/B testing produces in this matter can be invaluable. It lets you know which option is more effective for your business’s target audience.
What About Those Who Prefer Multivariate Testing?
In its own right, multivariate testing is a powerful tool. It can be used as an effective follow up to A/B split testing. In certain cases, the combinatorial power of multivariate split testing can be used to test a plethora of mutually compatible possibilities and determine the interaction of each of these elements. This is something that can’t be done with A/B split testing.
At Webati, we understand the value of both kinds of testing. A/B split testing is not necessarily better than multivariate. It is a matter of knowing when to use each type of testing. This is why so many SEO experts are proud to proclaim their affection for A/B testing.
When you choose to engage Webati’s services, we will utilize both types of testing to ensure that your page is optimized for your target audience. It is not the sophistication of the mathematics that matters. It is the sophistication of your marketing strategy.