How to set up a marketing experiment comparing two variants

How to set up a marketing experiment comparing two variants

The New York Times ran a crossword puzzle with the clue “a marketing experiment comparing two variants” but for those who want to dig a little deeper than the answer, what is an A/B Test and why would I use one?

Why have two variants?

Example of a sign up form A/B split test where one test has no progress bar and the other one does

What’s even the point of having two experiments at the same time? Why not just launch a website and call it a day?

A/B Testing allows marketers and project managers to really understand what makes users connect with the product.

A really simple example could be a sign-up form which shows a progress bar as you press “next” vs a sign-up form with no progress bar. You hypothesise that adding the progress bar adds clutter and a little more confusion to the page, so you do what’s called a split test or multi-variant test.

After running the tests for a while, it turns out that the version with the progress bar gives people assurance that they’re almost there and results in more users joining your website. So you make the change perminant.

So how can I use it on my website?

The Simple A/B Test User Dashboard showing a test and Statistical significance

There are various tools to help you optimise your website and Simple A/B Test is one of them.

Getting started is really easy and involves placing one line of code in your website’s header.

That’s the hard part done!

Now you can sign in to the dashboard and set up your marketing experiment comparing your two variants. You’ll have to pick out which element you want to test, such as the sign-up button or a call to action (CTA) and you’ll then be able to test out different ideas and see which works best.

Half of your users will see one test and the other half will see the other result. The line of code you entered earlier will track which performs better, and you’ll then know which change to keep.

That’s it?

Pretty much! It’s not a complex process, but a 1% increase in sales on 1,000 visitors is still 10 additional sales.

That’s the beauty of split testing. All sorts of websites use this method of testing and make big product decisions based on the results. Now you can too!

Want to increase your sales?