How to go about A/B and usability testing and why it’s so important for your website
/ 0
Squeezing the last drops of performance from an ecommerce website is vital in making a success of an online retail business. In a world where competition is fierce and margins are tight, being able to increase the conversion rate even by a fraction can make a significant difference to your bottom line. Imagine an ecommerce site that takes £50,000 per month in sales – even just a 1% improvement in conversion rate can make a £500 per month difference. So testing things that may make a difference here is a potentially valuable exercise.
Any sort of site testing is your friend. Whether it be A/B testing of product pages, usability testing across the site, or conversion rate optimisation of your shopping cart using your website analytics system.
Google Analytics (by some distance the most widely implemented analytics platform) includes native tools that enable webmasters to set up tests across different landing pages to accurately and consistently analyse which ones perform best.
Known as Google Content Experiments, the tool enables you to set up multiple different versions of a web page, serve the differing versions to randomly selected visitors and analyse the performance results.
The key to a successful experiment is to test carefully, changing just one variable at a time (for example, making the ‘buy now’ button red instead of green) so that you can isolate the impact of individual changes on performance and make informed decisions accordingly.
Other website analytics platforms have similar tools and there are also stand alone applications and even agencies that focus on this factor in particular – an indication of how vital testing is to the success of ecommerce sites.
Most tools around this area allow you to do more than just straightforward A/B testing. Certainly the Google tool (which has the benefit of being free), enables you to test multiple versions of a page simultaneously. They also enable you to test various calls to action. For example most of the time with an ecommerce site you will probably want to test experiments that improve sales, but you may also want to test potential improvements to your ‘newsletter sign-up’ page, your ‘contact us’ page, a competition entry page or indeed, others.
In summary, testing is probably an avenue much underused by webmasters – typically we’re happy to spend significant amounts of money driving traffic to a website, but less likely to spend time and money on testing to optimise the return on this investment once it’s up and running. But, it’s an angle that anyone looking for competitive advantage in a crowded online marketspace should allocate time to.
