Site Optimisation for survival

Retailers have been having a hard time recently (and not just: Woolworths, Zavvi, Adams, USC, Morgan and Wittards), with stores not having the sales they once had. For many of them their eCommerce offering has been a major revenue stream. Digital has come of-age and is paying its way now.

However, things don't stop there. All but the most ignorant online retailers have realised that just driving traffic to a page isn't enough and they are now looking at ways to optimise their sites. Yes, you can use analytics tools such as Google Analytics or one of the more professional packages to tell you how they get there and what people are looking at. But it won't tell you how to make your content (headlines, promotion material, images, calls-to-action, etc.) more effective.

The conversion of a vistor to a buyer is a complex mixture of art and science. Its therfore the combination of the right information and correct enticement to get the potential buyer to add a product to the shopping cart (and enter their credit card details after that).

But what combination works best for your site and how do you find this out?

A/B Split Testing:
Given two alternatives, users will usually state a preference. Showing the same web page but with two alternative pieces of content can also give you two different outcomes. Measuring which pieces of content makes your users buy more product is a simple but powerful bit of analysis and this is the basis of split or A/B testing. A lot of sites now use this method of site optimisation, from trialling a new homepage layout or design through to dynamically changing the supporting messaging during the checkout process.


Multivariate Testing:
This takes A/B testing one stage further and tests combination of page components or elements and measures the effect they have. Basically it ignores the rules of 'Ceteris Paribus' (the principle that all other things must be equal to be able to test a specific condition).
Note: You can take this process further and also automate the segmentation of your different users according to factors such as

• Geographic region
• Day of week / time of day
• Referencing site
• Etc.

Its quite obvious to me that execs are going to look to their online retail operations to work harder and smarter in the future. By making some simple changes, carrying out analysis and applying this to the evolution of your own site, you can quickly de-mystify the issues and get your site converting more.