3 A/B testing hacks you can use today

a-b-testing-for-success

The online power players use A/B testing (also called split testing) constantly to fine-tune their web pages to get the best conversion rates. They wouldn’t present virtually any web page to the public without having first tested it to see if it does the job.

Small business owners may think that they don’t need to do this or that A/B testing is too complicated for them to bother with, but that’s not the case today. I will outline three ways you can bring A/B testing to your website. But, before I touch on those three methods, let’s quickly define what we’re talking about when we discuss A/B testing.

As I said above, we are going to describe how to use A/B testing to determine what works best on a website, such as landing pages or calls to action. Therefore, in its simplest form, we would create two web pages that are designed to get the visitor to do something, such as buy an item, push a button, or perhaps stay long enough to read some information.

We would send half of our visitors to “Page A” and half to “Page B.” We would measure what we’re testing for and then compare the results. After some amount of testing, we would eliminate the page that finishes second in our test. We might tweak it again and then test that new version against the winner of the first A/B test.

A/B testing with Google Analytics

The problem with Google Analytics is that it allows you to do so much, that it’s sometimes difficult to know all the tools that are available to you. If you have Google Analytics on your website, you can do A/B testing fairly easily…but you have to find it first! (And you have to know that Google doesn’t call it A/B testing, or even split testing.)

analytics-menu-navigationNavigate to your Google Analytics page, then select Behavior, then select Experiments; that’s where you can set up your A/B test. There are a few “default” experiments you can set up (bounces, pageviews, session duration), but if you select “Create a new objective,” you’ll get an even bigger menu of tests.

There is a little bit of code involved in setting up these tests; you need to put line of code on Page A in your A/B test. If you have a static website, you just paste it into the header, or have someone a bit more HTML savvy than yourself do this. If you’re using a WordPress site, there are plugins that will get this done for you. And that takes us to our next “hack.”

A/B testing with WordPress

As with everything in the WordPress world, there are multiple plugins available to help you accomplish A/B testing. There are also various services you can use to accomplish different A/B tests on your WordPress site.

Further, a service you’re already using might have A/B testing built in. Thrive Leads, for example, lets you test different email-capture pop-ups against each other, as well as against other forms of capturing email addresses.

Brenda Barron has written a good article that gives you the run down on what she believes are the 11 best tools for WordPress A/B testing. If you have a WordPress powered site, you’ll find it very helpful.

Voyeuristic A/B testing

Talking to small business owners, one of the problems many have is that their websites don’t generate enough traffic to return good A/B test results in a timely manner. I get it. But, that doesn’t mean you shouldn’t study and familiarize yourself with A/B test results.

There’s a great website called Behave.org and it houses a huge collection of test results that you can browse. In fact, they typically format these as quizzes, so you have to guess which variation won.

As a wise man once said, there is nothing new under the sun, and if you spend some time exploring Behave.org’s “Tests of the Week” and sign up for its newsletter, you’ll find a lot of A/B testing case studies that deal with issues and questions similar to yours. For example, which of the 15 percent off calls to action below was the better performer? The straight offer, or the “gamified” offer?

standard-or-gamified

The test above measured the results of nearly 12,000 users, so you know that the conclusion is pretty solid. It would take many small business websites a long time to gather that kind of data, so you can really shorten the learning curve if you spend some time studying these test results.

Finally, remember how I kicked this off by referring to the fact that all the big companies do a lot of A/B testing? They do it because when one variation is shown to be the best performer – even if it’s by just a few percentage points – it can translate into far greater profits.

This will also prove to be true in your business!