Writing seductive copy for a show that’s lagging behind at the box office is a bit of an art in itself. Maybe you’re not sure what to stick in the subject line of an email. Or perhaps you’re wondering whether your target audience are more motivated by a free glass of wine or a special offer on the tickets. Well, wonder no more.
A/B testing, at its simplest, lets you compare the results from two different versions of a marketing campaign that you send to a small proportion of your total list. The idea is that you can then take the one that performs best (either by getting opened or clicked more) and then send it out to everyone else.
Let’s say you’re emailing out a campaign about a particular show. You’ve already created your segmentations to find a group of people who are likely to be really interested in the act – they’ve come to similar performances before, they’re around the right age range, they’re active customers, and you haven’t emailed them in the past two weeks. (This takes about 5 minutes in Spektrix by the way – if it’d take you longer on your system, you should probably give us a call!) You’ve got this set of 5000 people, but you’ve only got one shot at emailing them with something that’s going to grab their attention in the midst of a busy inbox.
So we’ll choose 5% of our list and send them an email with the subject line “BOOK NOW for a GROUND-BREAKING show.”
And we’ll also send another 5% of recipients an email with the more restrained subject, “Booking now open for Spektrix: The Musical.”
We can then tell Spektrix to wait for a specified period of time before sending out whichever version got the highest open rate to the remaining 90% of the list.
In this case, “Booking now open for Spektrix: The Musical” was 1.5 times more likely to be opened than our original email, so that’s the one that will get sent out.
OK fine, but this all sounds like a bit of a hassle – who really wants to write two versions of every email? Ignoring the fact that sometimes an A/B test can be as simple as changing one sentence or a subject line, there are other compelling reasons to run A/B tests as part of your regular marketing strategy:
1. Expect the unexpected
Sometimes the results won’t be what you expect. I’ve recently found that in some cases, a more boring subject line has performed better than its exciting counterpart. It’s counterintuitive, but I’ve got the data (and the bookings) to prove it. Now I’m not saying that you should deliberately send out cures for insomnia, but it just goes to show that it’s worth testing before going with what seems intuitive.
2. Get to know your recipients
It’s important to remember that there are people on the end of those email addresses! By running A/B tests, you’re getting to know your audience (even if it’s in a small way).
3. Times and tastes change
If you’re sending out emails regularly, it’s easy to get into a routine. But it’s important to remember that what works now may not always be so successful. As customers get accustomed to them, strategies that once were innovative and attention-grabbing become gradually less successful. A/B testing is an easy way to keep your finger on the pulse.
4. Challenge yourself
A/B testing forces you to be more creative when coming up with promotional material or strategies – and that can only be a good thing.
A word of warning, though: there are some things which you’re going to want to be wary of before you jump into the world of A/B testing...
- Be patient
- Make sure you wait long enough before deciding on which version to send out to the full list. Otherwise a few early recipients could skew your results in the wrong direction. So if you’ve got a campaign that needs to go out RIGHT NOW, consider saving the testing for another time.
- It’s not a magic bullet
- A/B testing isn’t going to revolutionise your marketing overnight – it’s about making small changes and a process of gradual improvement. Of course, that’s not to say that you won’t hit lucky and see a 1000% improvement…
- Test one thing at a time
- In addition to testing email subject lines, you can test the “from” name and email address, the template and the content. But don’t go changing five distinct things about your email at once, otherwise you’ll be stuck wondering what caused the increased (or decreased) response. Instead, just change one thing at a time. The joy of A/B testing is that there’s no pressure to get it right from the very first attempt.
It goes without saying that everything I’ve mentioned is baked right in to Spektrix. Get in touch – we’d be delighted to give you a tour.