We all know what we think makes an effective website design.
The trouble is, you and I aren’t the customers and we can’t possibly know for sure what works the best across the online audience. We can also waste a lot of time arguing within the design team, or with what analytics guys calls the HiPPO (Highest Paid Person’s Opinion) about what works. But it’s the customer who counts as they vote with their mice…
Step forward A/B and multivariate testing, which are increasing in popularity to squeeze higher conversion rate out of a site. Established, paid tools from suppliers such as Optimost, Omniture (Adobe Product Now), Test & Target, Maximizer have mainly been the preserve of large organization. The profile of these tools has been increased by the free Google Website Optimizer.
A/B testing is where most people start. It’s great for testing major changes such as different versions of a landing page, or a key element of a page such as a headline or a call to action. It can offer huge improvements.
Many large site owners have now graduated to multivariate testing, where you statistically define the best converting version of a webpage or pages in real-time. Recently at on Unicus website, we did a test with 192 elements and found two major changes that increased conversion by 10 percent.
So structured testing is well worth a look if you’re not using it already, but remember it’s a specialist technique requiring an understanding of experience and statistical validity.