5 Email A/B Split Test Ideas You Haven’t Tried
You’re probably sitting there thinking, “A/B testing my emails is hard!” or “Where do I even start with A/B testing my email campaigns?” The truth is, most online businesses could spend a lot more time testing their email marketing campaigns, and this means you can stand out from the crowd by being one of the […]
You’re probably sitting there thinking, “A/B testing my emails is hard!” or “Where do I even start with A/B testing my email campaigns?”
The truth is, most online businesses could spend a lot more time testing their email marketing campaigns, and this means you can stand out from the crowd by being one of the companies that does.
If you’re looking for ideas on what to test, here are five things you can experiment with — they’re likely things you have never tried.
Let’s break each idea down and throw in a few examples so you can learn to harness the full power of A/B testing for your email marketing campaigns.
1. Split Your Primary CTA Into Deeper Links
It’s always a good idea to focus on a single call-to-action, as this will generally increase your email marketing conversions. However, there are times when you can drive click-throughs and conversions even higher by splitting your primary call-to-action into parts.
Take this example from Indeed.com:
As a reader, I find this email very engaging — it not only reduces the number of clicks I have to make, but also makes it perfectly clear what the goal of this campaign is and how I can reach it.
Amazon is no stranger to this concept, either. When you purchase a book on your Kindle, Amazon will generally send you an email a little like this:
…however, they also test a variation that looks like this:
As you can see, they are clearly breaking up the call-to-action into a series of “deep links” — links that take the customer further down the funnel. When you click “Rate this book” in the first email, you are taken to a page where you select the rating (out of five) and, optionally, leave a comment.
In the latter email, a click and a step are removed so you have to do less. This is a fine example of the KISS principle at work: keep it simple, stupid. By making it extremely clear what you’re getting yourself into when engaging with this email, Amazon nails the split CTA.
A final example of splitting out the primary CTA into deeper links comes from Badoo with their social “rate another profile” update emails:
So, next time you’re setting up an A/B test, ask yourself: are you able to break out your CTA into more detailed links? Can you help your customers “skip a step”? If so, test this for an extra boost in conversions.
2. Repeat Your CTA In Your Signature Or Postscript (P.S.)
Even if you can’t split out your CTA, you should certainly try repeating it!
Including too many links is overwhelming — but having just two or even three links pointing to the same ultimate goal generally leads to a lift in conversions rather than a drop-off.
This message from KISSmetrics is a classic example of a long-form email that ultimately ends with a simple call-to-action — in this case, to join a webinar. You will note the repetition of the CTA in the postscript. A postscript (or P.S.) is a magical device, as readers are visually drawn to them, even when skimming through an email. If you want to drive a point home, repeating it in the P.S. really helps.
You can see that repeating the call-to-action isn’t too difficult. The most common ways to frame the call-to-action in the P.S. are:
- Offering a discount or bonus. E.g., Remember, the first 10 people to click through get an extra 10% off!
- Adding a personal touch. E.g., I can’t wait to see these five tips help transform your email conversions. Sign up and let’s talk.
- Creating a sense of urgency. E.g., Don’t forget that this offer ends in 12 hours.
All of these can be effective, but one thing is for sure: you should test whether including a CTA a second time, as part of a postscript or footer, lifts your conversions.
3. Test The Variety & Order Of Your Secondary Links
Lots of campaigns, particularly those that use standard templates, contain a series of secondary links located in the header area. These are designed to give the customer options or to take them down different paths to the same goal (e.g., purchase).
Here’s an example from Net-a-Porter:
Here’s a second example from Amazon:
Aside from testing whether removing these links entirely increases conversions, you should definitely test the order and contents of these links.
In the Amazon example, there are a number of interesting links in the header, including a link to personal recommendations, a link to sign up for Amazon Prime, a link for free UK delivery, and a series of three links for “MP3,” “Deals of the Week,” and “See all departments.”
There is a lot going on there. What if you were to change the order of these links? Should men see the same links as women? Should people under 30 see the same as people over 30? Is including a link to Amazon Prime cannibalising the Prime conversions?
These sorts of tests are becoming easier and easier to build, as you can use advanced templating languages to include or exclude content based on various parameters. Even having two basic variations – one with links A, B and C, and another with links D, E and F — will tell you a great deal about what works!
This doesn’t just apply to complex templates. In this blog update email from KISSmetrics, there is a link to sign up in the footer.
We do a similar thing in the Vero blog update update emails:
What if this was a link to our email marketing course instead of our homepage? Or a separate landing page entirely? What if it was different for different customer segments?
A/B testing the order and format of every piece of content and every link in your campaigns allows you to make informed decisions on these questions!
4. Ask Questions
“Questions are worth a thousand opens.” Perhaps that should become a new email marketing maxim.
The truth is, asking questions is really powerful. This is particularly true when it comes to subject lines. Rather than writing:
Find out if your blog is growing fast enough
…you might write:
Is your blog growing fast enough?
It’s a subjective difference, but the question implies a call-to-action. Litmus pulled together an awesome email marketing infographic that also emphasizes the power of questions.
This doesn’t just hold up in email subject lines but is a great way to approach new content. Flightfox doubled their conversion rates on this lifecycle email marketing campaign using a question-and-answer format.
They understood that customers are looking to be educated. By finding the most common customer support questions, Flightfox was able to pull together a campaign that truly answers customers’ questions and focuses on them.
Next time you send a campaign, run a simple A/B test on your subject line or body content that uses questions — you may be surprised by the results!
5. Go Mobile With Responsive Templates
As mobile devices take over the world, it is more important than ever to ensure you maximize your conversions for recipients reading their emails on mobile devices.
Up to 65% of emails email opens already occur on mobile devices. That’s a huge percentage — so you should definitely track what devices your customers are using. It is worth conducting an A/B test with a responsive template to see the impact.
Litmus recently blogged about the success Deckers had with the email marketing campaigns for their Tsubo and Ahnu clothing brands. Using responsive templates, they were able to increase click-through rates by 10%. That’s a nice bump simply from using a little extra CSS to style the campaign. Here’s an example they published from Tsubo:
Here’s what you should do next:
- Grab your next email marketing campaign (perhaps you’ve already got one in draft ready to go?) and set up an A/B test. Keep it simple and keep it small, if you’d like, but use one of the tips above, and see how you do!
- Make sure you’re correctly tracking your email marketing conversions.
See you in the comments!
Opinions expressed in this article are those of the guest author and not necessarily MarTech. Staff authors are listed here.