Tips for Better A/B Split Tests
As an email marketing vendor, we see a lot of email marketing best practices being used…and plenty more neglected due to lack of understanding.
The A/B split test may help optimize your email marketing campaigns, but is often an email best practice misunderstood by email marketers, so here’s a little refresher on how, why and when to do it…
An A/B split test, is comprised of 2 (or more) versions of an email where one thing is done differently. It could be the From name, subject line, headline, or an image. It might be the time of day or day of the week that you send. It might be the email design, the call to action, or the copy. (Just be sure to test only one element at a time, otherwise you won’t know what element was the one that worked better!)
In order to find out what works better in your email marketing program, you compare A to B by sending to portion of your audience (usually between 10% and 20%). Whichever performs better is the one to send to your remaining subscribers.
This is where we’ve seen some confusion on the part of marketers. They plan to do what I call an A/B/C test, comparing three (or more!) versions of an email. But then run out of time and instead of sending the different versions of the emails to only the small portion of their total list, they send out the versions to the whole list: A to one third, B to one third and C to the remaining third.
They might learn that one version performs better than the other, but guess what? They can’t use that information! They might learn that version C performs 300% better than the other two, but this knowledge does them no good because they can’t resend. And this knowledge could not only mean a loss in interest for some subscribers, but a loss in sales!
So what happens here? Lack of time. It’s important to note that these marketers are well intentioned and on the right track. Testing is a crucial part of any email marketing program. But by the time they’ve determined the test emails, a deadline is fast approaching and the time-sensitive emails simply cannot wait. They must be sent. So all are sent. And knowledge gained is useless knowledge. They’ll only learn something they can’t use on that campaign.
As with anything email, it is better to do it right and do it well.