Landing Page A/B Testing – Adding Method to the Madness

By now most marketers, business owners and web developers understand the importance of A/B testing as part of an ongoing process to improve any website business. The better your website performs and the easier it is for your audience to use, the better your site’s performance will be. That means more sales, more leads, more sign ups, what ever your goal or goals may be.

While it’s common for online businesses to be well aware and engaged with A/B testing, it’s surprising how many are going about it in a less than methodical way. Though by far one of the most surprising mistakes is taking the dramatic variation route. Producing extremely different landing pages for use in a single test. Sometimes using entirely different layouts, images, or even colour pallets. In essence, dropping test visitors on to sites which look almost completely different in order to gauge their reaction.Now, this approach has some merit perhaps if you’re planning a major site overhaul and you’re conducting a small overall design test before moving ahead with full scale design work, but it’s far too commonly being used as part of ongoing A/B testing methods. Here’s the issue. If you make a dramatic change to a page for the sake of an A/B test, you may see some minor positive results, not ground breaking but enough for you to consider changing the page. If you have made a number of major changes for this test, you have no way of knowing conclusively, which changes contributed the most to the KPI increase.It's very possible and actually more than likely, that one or two of the changes made contributions to the increased performance. It’s also more than likely that those changes would have done the same or perhaps better job of increasing performance if other amended factors of the previous page had remained.It’s heavily rumoured that Google experimented with over 500 different shades of blue before finally settling on the colour for their sponsored ad text links. The key to testing is to consistently and constantly test minor changes even if it only yields a minor impact. This way, you can continue to build on the site and learnings you have instead of gambling with big changes. It also makes for a more consistent user experience while you continue to innovate your site a piece at a time.·Modify the size or position of your call to action buttons.·Change the colour of major call to actions associated with your primary KPIs.·Reduce header sizes to allow more page body to display upon opening.·Edit the order of your main navigation etc.But do these small items one test at a time until you find a winning element. Incorporate it into your default live site and move on to the next item systematically and set this plan out loosely before you even start. With the view in mind that there is no finish line when it comes to growing and improving.