AB split testing of your websites page head, images and call-to-action methods can help you improve your conversions. CHRIS THOMAS
By Chris Thomas
When larger companies develop their new websites, they often call on the services of a usability testing company prior to launch. Probably the best known example in Australia is www.usabliltyone.com.au.
I was lucky enough to sit in on a presentation they did at a conference I attended back in February. The case studies showcased were fascinating and demonstrated how designers and developers can sometimes get too close to a project; forgetting how a typical user might interact with their designs.
Unfortunately not all of us can afford the luxury of a new website coupled with usability testing. Luckily there other (and in some cases free) methods you can use to increase the conversion rate of your existing website.
Don’t be scared of the big words(!) but the answers lie in “AB split testing” and/or “multivariate testing”. Certainly this week’s interview Amanda Gome conducted with Jason West from Websalad confirms this.
Having enjoyed many conversations with Jason, I wholeheartedly agree with his maxim that there are two ways you can double the amount of conversions; you can double the amount of traffic to your site, or you can optimise your website to achieve the same result. If you do both, you can retire rich.
AB split testing is a method where you create two (or more) versions of a web page and analyse the conversion results to see which one entices your visitors to convert in greater numbers. You then ditch the lower performing page and keep the good one! If you’re really serious about it, you keep testing. Forever.
Multivariate testing is similar in concept to AB split testing, but instead of serving up two pages, you serve up just the one page and change and test various page elements to hit upon the best combination.
Page elements you’re likely to test usually include:
- Your page heading.
- Your images.
- Your body text copy.
- Your call to action method.
Google explains it all quite well in this video.
Research shows that two elements affect conversion rates the most – your heading and your call to action method. In one case study, a red “buy now” button was compared to a green “buy now” button. Multivariate testing quickly revealed that the green button received more conversions. Red means “Stop!” and green means “Go”.
Surprisingly, images and body text tended to have lower impact on conversions.
Google has a reasonably new (and free) tool called the “website optimiser”. You’ll find it inside your Google AdWords account. You get what you pay for though! You’ll probably need a developer to help you set it up as the coding requirements are quite involved. Plus you’ll need a lot of time and traffic for the tool to accurately report back the results.
Of course there are paid tools available, which are typically more sophisticated than Google’s free website optimiser tool. I guess if you were to weigh it all up, using a paid testing system set up by an experienced operator might prove cheaper than trying to learn it all yourself. They typically also use sophisticated mathematical modelling to allow you to test much higher numbers of variations in a much shorter period of time.
Certainly if your conversion rates are static, at say 2%, then doubling that to 4% (or more) might make the few thousand dollar average spend worth the investment over time.
Chris Thomas heads Reseo a search engine optimisation company which specialises in setting up and maintaining Google AdWords campaigns, Affiliate Programs and Search Engine Optimisation campaigns for a range of corporate clients.
For more Online Sales blogs, click here.