5 min read

You’ve probably seen this before. You’ve spent hours cleaning your data, writing killer content and designing a stunning email campaign…only to have no clue as to its impact.

Without measurement, marketing is essentially a wasted effort. That’s why it’s so important to make the most of the analytics tools available to you. Open rates, click-throughs and bounced emails are all crucial data…but there’s one step you can take to get even better insights. Enter: A/B testing.

What is A/B testing?

A/B testing involves distributing two different versions of the same content to different audiences. In an email marketing context, this could be the same campaign sent to two different segments.

You can A/B test a huge number of variables, from email copy to design, subject lines – even something as minor as colours of call to action buttons.

A reputable email client, such as Campaign Monitor or Mailchimp, will offer this service for you. It will measure the effectiveness of both versions of your content, then automate it to send the most successful type in future campaigns.

Source: ours friends at Campaign Monitor

How can A/B testing help me?

By giving you key insights into your campaigns’ impact, A/B testing can help with:

  • Increasing open rates
  • Increasing click-through rates
  • Increasing conversions
  • Measuring different customer segments’ preferences
  • Determining the best time to send a campaign.

What type of variables can I A/B test on my email campaigns?

Depending on the data you want, you can measure any of the following with A/B testing.

Content
Perhaps the most obvious, you can test the effectiveness of subject lines and the copy used for calls to action. With the latter, you can even measure how people react to the visuals, for example colours on backgrounds of buttons.

The guys at BuzzFeed do a cracking job of this in their newsletters.

Personalisation
From subject lines to the pretext, there are ample opportunities to personalise your copy. You can test to see what kind of personalisation works best, which will also give you a better idea of how readers are viewing your email.

Recent study from Return Path shows that the optimal length is around 61-70 characters.

Timing
Perhaps you want to know if weekdays or weekends perform better. This will vary enormously depending on the industry, as well as other factors like time zones. You can automate emails to send at certain times based on the best converting weekdays/times.

Design
Your audience may be more visuals-led, for example if you’re selling clothing, or more text-led, for example a medical newsletter. You can trial different email designs to see which ones garner the most engagement from users.

Here’s a good visual example from The Wolf in Sheep’s Clothing.

Source: really good emails

Some Dos and Don’ts for A/B testing

DO:

  • Test a number of different audiences/segments (A staggering 39% of brands don’t test their emails or segmented emails)
  • Focus on the important elements like calls to action, subject lines and visuals
  • Always test two versions at the same time.

DON’T:

  • Test with tiny data sets – you’ll need a large database for conclusive results.
  • Test more than one variable at a time, as this can skew results
  • Go with your gut. Only make campaign decisions based on conclusive data.

Remember, you should be confident in your mail client’s testing tools before you start. With time and effort, you could optimise your campaigns even further.

Let the testing begin…