A/B Testing

Wait, why does this blog post look different?

This post is part of the ongoing series I've imaginatively entitled ‘Glossary' where I define popular marketing terms and ideas.

Click here to see all the terms in one place.

Definition of A/B testing

A/B testing (usually called ‘Split Testing') is when you'd test version A of, say, a web page, against version B, and see which gets a better results.

It's also widely used in online advertising, like Facebook and Google Adwords.

Why bother?

A/B testing is a great way to optimise and improve on an offer, advert or call to action.

In fact, if you're scaling your business (in the ‘Child' or ‘Teenage' phases) then it can be the fastest way to improve profits without spending any extra income.

Who is it right for?

A lot of marketers think that they should always be split testing, but spit testing in the ‘infant' stage (0 – £100k in revenue per year) is usually a waste of time and resources.

Split testing is great to scale and optimise – but you need plenty of data (and clicks/sales) to gauge what is actually working.

If you're making fewer than 1 sale a day, then my advice is to forget about split testing for now.

What is the one thing I need to know?

“Test screams not whispers”. 

I can't remember who said this, but it's brilliant – basically, it means don't test whether the colour red on the ‘buy now' button increases sales – test something with a big impact, like the headline, or the price.

Once you've finished testing all your big stuff, then you can move onto the small stuff.

Where can I learn more?

Optimizely is a web app that will make split testing on your website super-easy. They've got a great guide that goes through the basics.

Or Neil Patel has done one of his trademark ‘Beginners guide to' articles on Split Testing on his personal blog.