A/B Testing

An Introduction to A/B Testing  Email Campaigns


Email marketing is a tried-and-tested method to retain and expand your customer base. But how is a marketer to choose what type of campaign or elements of a campaign will prove most effective? A/B testing or split testing helps compare two versions of your campaign to see which yields a better conversion. In this blog post, we bring you our top tips for getting you started on A/B testing email campaigns.


Decide upon your "Call To Action"

Your “call to action” is the number one variable you want to optimize for. Consider what the larger goal the campaign ties into. What do you hope this campaign will result in? What ideal response would you want it to elicit? For example, increasing sales of a specific product may be your call to action. Explore what email campaign best accomplishes the same and design your A/B test accordingly (for instance, by including two different images of this product, or using the same image but different text).  

Choose your test elements 

Decide what email campaign elements you want to split test for. Some examples include: subject line, shorter versus longer copy, the main image, number of products, the kinds of products or offers, background colors, font, product images, dimensions of elements (like the header, body, images, buttons), details of the click-through button such as size, color and text, etc. You could also test for broader concepts such as appealing to intellect versus emotion, using a more formal tone versus an informal tone in the copy, a soft sell (outlining the advantages of a product) versus a hard sell (a direct call to “Buy now”), etc.

Choose your audience

The larger the sample size you test for, the more accurate your results will be.  If you have a large enough list size, it may not be a bad idea to segment your customers by various metrics— demographic or socio-economic features, geographic location, paid versus organic users and so on. Make sure everyone gets the email at the same time to remove time as a variable from the equation and ensure it doesn’t skew your results.

The Verdict

Along with the call for action, decide what metrics you want to want to measure and analyze.  Upon completing your campaign, measure each of the campaign variations to see which did better on these metrics. Typical examples include: open rate (percentage of recipients who opened your email compared to the total recipients), the click-through rate (percentage of recipients who clicked on any link within your email), the conversion rate (the percentage of recipients who take the desired action) and the average revenue per email sent.   You may not find the perfect campaign straight away. But through the process of experimenting with a few different split tests, you can know what works and what doesn’t. Eventually, you will discover the best combination of elements that will yield your desired outcome.