A B Testing and the Art of Improvement
Perfection is an ideal. A concept in the mind with no grounding in the physical reality in which we live. Akin to the perfect circle or parallel lines. There are no such entities anywhere in the known universe, save for the human imagination. Yet we know about them and strive towards them. There are mathematicians at this very moment trying to calculate the next decimal in p. And why? Why do we strive for an unreachable ideal? Because we are human beings, and the one thing we value most is progress.
There’s never been a single great idea in human history that someone, a generation or two after its inception, didn’t look at and say: “You know, if we just did it this way, it would work a whole lot better.” You’re website is no different. You may be happy with your layout, your images might just be immaculate, and you could have a superbly customized WordPress theme but there’s always something that you could tweak to make it just a little bit more effective.
That, my friends, is what A/B testing is all about.
Split Decisions
A/B testing, also known as split testing, is a method of examining your website to see which variables positively or negatively affect conversion rates, and to what degree. You basically make two different variations of your web page, and see which one works better at achieving its intended goal.
It’s basically a complicated system of guess and check. You make an assumption about how your web page is performing, such as: “perhaps my images aren’t effective at converting my traffic.” Then you test that assumption by making a different version of the page. In the case of the assumption above, you would try using different images, or changing their placement.
Then you would split the traffic between both variations of the page, and measure the relevant metrics for each. Once you’ve got your measurements, you compare the two sets and determine which version performed better for the desired outcome.
Sounds super simple and easy, right? Well, hold on a tick. We haven’t gotten to the technical part yet.
How Can I Conduct an A/B Test?
There are a variety of testing tools that you can use to aid you in your testing efforts. Some notable names are:
- Google Website Optimizer
- Visual Website Optimizer
- Vanity
- Unbounce
Take a look at each option, or do some research and find an alternative of your own. See what the best fit for you is, and act accordingly. Each tool has its own set of requisites and directives, but in general they all have intuitive visual interfaces, and easy to understand guided directions. Depending on your tool, you may have to have a little knowledge about HTML, JavaScript, or CSS.
Of course, if you’re a hyper intelligent web developer, you might be able to develop your own solution. However, if that’s the case, you probably already know plenty about the subject, and should actually be writing your own articles. If this describes you, leave me a comment and let me know if I missed anything.
There are two different ways to conduct A/B testing. You can either replace a single element on your test page, or you can redirect to another page entirely. Obviously the former is a lot easier to do. You just have to create an alternate version of this element.
What Sort of On-Page Elements Should be Split Tested?
There are numerous things on a landing page that can affect conversion rates, and it’s a good idea to test them all in sequence. In addition to images, there are many different variable that can be A/B tested. Here are a few to get you started:
- Headlines
- Textual content
- Testimonials
- Calls to Action
- Call to Action Buttons
- In text links
- Security badge placement
This is far from a comprehensive list, so look around your website for little tidbits of inspiration.
Really, you should examine your site’s metrics to determine which variables need testing. The whole process hangs its hat on the scientific method. Make a hypothesis and go about testing it in a thorough sequential manner. If you put the work in, and continuously sharpen your web pages edges, you’ll eventually wind up with conversions far beyond your initial expectations.
That’s the entire point of A/B testing, to approach the ideal, and never stop in your efforts to reach it. Happy testing!
Zack Rutherford is a freelance copywriter. He contributes web content and especially snappy articles to TemplateMonster. Combat sports enthusiast and poetic soul, Zack endeavors to create beauty through syntax, sentence structure, and the liberal use of hyperbole. Follow him on Twitter (@zack_rutherford) or Google+ (or visit his website (Zackrex.com) to read all of his innermost thoughts and unfounded opinions.