Learn Tech Skills

A/B Testing

A/B testing compares different versions of the same web page to see which one performs better with users.

In the web world, people love to argue about what goes on their company’s site. Red buttons, blue buttons, more text, less text – no one can ever decide anything! Everyone has their own personal opinion on what will resonate most with the user.

A/B testing allows you to test different versions of the same page – or even parts of the page – to see which one makes more sales or gets more clicks. Some users see version A, some users see version B, and then you track which version your users prefer. It’s a great way to get data to back up hunches.

Mailchimp (who is sending you this very email), allows you to do A/B testing on your email subject lines to see which one readers are more likely to open; it sends a few emails out to see which subject line performs better, then sends the winner to the rest of the list!

A/B testing can be run many different ways. One of the easiest is Google Website Optimizer – it uses JavaScript to show certain content to certain visitors, and then track them as they use the rest of your site.

After a few days of A/B testing it’s easy to look at the results and see which version gave you the most new customers or generated the most traffic. Then you get to do it all over again with new changes!

A/B testing is especially popular with people in the Lean Startup movement, who believe in using all manner of scientific methodology to determine what users want.

Cocktail Party Fact

In 2009, Google’s vice president of search products and user experience, Marissa Mayer, decided to test 41 different shades of blue on a toolbar before settling on the one that performed the best!