Comparing your company's success against competitors is crucial in evaluating and improving your performance. But not only measuring against other businesses can be a helpful instrument - using A/B testing allows you to test multiple variations of your business against one another. This is why we've compiled this short guide to A/B testing, how to plan and what you can test.

A/B testing - explained

A/B testing, or split testing, describes a way to evaluate your conversions by obtaining data from customer behaviours. The term itself refers to comparing two versions of one thing to then measure their performance. Test results are mostly obtained by trialling the first version with a group and the second one with another and then comparing the results for each of the testing elements. Of course, you are then advised to decide on the option with better performance. A/B testing is a very flexible tool. The great thing about A/B testing is just how flexible it is. If you can change it then you can test it.

Advantages of A/B testing

A/B testing can positively impact the effectiveness of campaigns on eg your web pages and improving your content ultimately help increase revenue. It is a way of collecting fairly reliable data based on empiric methods to streamline your content and take a lot of the guesswork out of marketing.

A good example for a useful application of split testing would be determining how much traffic each of the two marketing campaigns you are deciding between can create. For the testing, you can target one strategy at group X and the other at group Y at the same time. By using key performance indicators/ metrics, you can determine which has the better conversion rate and use this one.

How to undertake an A/B test

First, once you have decided what you want to test, you choose to either have an off-site or an on-site test. You then define the metrics of the multivariate testing (eg the text in an email to the colour of the website design). Then, define the results you wish to obtain. Use your current results as a baseline for this. You then undertake your tests simultaneously to account for timing variations.

A great tool to use for A/B testing is Google Analytics, possibly using their free versions. There also are other tools available online (eg Optimizely or Leadformly).