In late April in Amsterdam, I met up with Optimizely CEO Dan Siroker to talk about Optimizely’s arrival in Europe. Fresh off a $28M round of fundraising announced last month, Optimizely is bringing its A/B testing web application to Europe, having just opened up its European office in Amsterdam – which means they are actually in Europe (as opposed to the UK/I). The service is fairly simple: allow A/B Testing – the ability to track site visitors interaction on a given page by changing one or two details and splitting the traffic between them for a given period of time – for anyone, without needing a developer (that’s the kicker).
“everytime I wanted to change something, I had to call a developer,” says Dan.
Designed after Siroker’s experience working on the Obama campaign site, the service requires little-to-no technical knowledge, just a Google Analytics-style implementation of a one-line piece of JS code in the of your website. Optimizely then adjusts the way the visitor views your site (so every visitor still sees the same webpage, no redirects) on the fly, based on “experiments” that you design. On the Optimizely web app, the site loads a visual editor – a click-to-edit version of your site that allows for easy change of titles, images, etc. – where you can design experiments (for example, testing the title of an article on the homepage to see which generates more clicks) and track clicks and other actions.
Siroker says that big clients like the New York Magazine are already using Optimizely to optimize their headlines – letting initial visitors see one of two headline possibilities, and then using permanently the title that gets the higher clickthrough rate. These kinds of tactics are great ways to constantly improve your site.
When Siroker explained the service to me, I instantly signed up and wanted to use it, and quickly discovered that, because I wasn’t doing event tracking with Google Analytics (i.e: tracking who clicks where, how users interact with each page, etc.), I was a little in the dark on what to A/B test. So, at the end of April, I began tracking user interaction on the site – once I have an idea of what users are clicking on – or not clicking on – I will begin to massively A/B test.
A/B Testing is essential if you want to take your site from 70% efficiency to 95% efficiency – each A/B test that improves conversions by 1-2% is important, and they add up as you test more and more. Siroker says that the company A/B tests nearly everything on the site – he even offered me one of two business cards, which had different URLs on them, in order to track which business card resulted in more follow-up visits to the site.
One thing’s for sure – site owners have no excuse not to be A/B Testing. It doesn’t take a genius – trust me, I’m A/B Testing this article right now!
Leave a Reply
You must be logged in to post a comment.