We are asked all the time about the best day/time to send an email campaign, why a campaign got the results it did, if certain elements should be included in a blast, etc. We are happy to give guidance based on our experience, but when it comes down to it – our opinion doesn’t matter. Only the opinion of your audience matters.
How to find out what your audience thinks? Test, test, and test some more. A/B testing is great for figuring out which part of your campaign your target audience responds best to.
You might think your email campaign looks great, the copy is exactly what you want to say, there’s an obvious call-to-action, and the timing works best for your schedule. So if your results aren’t what you were expecting, you’re left scratching your head … perhaps you change a few things around and send another blast. But that’s really just another shot in the dark.
Get into your audience’s head by doing an A/B test. The best way to measure the results is to change one thing per campaign. It can be hard to know what to change, especially when you’re sure your design is perfect as it is, so here are a few ideas of some differences you can test:
- Headline – Change the wording, the color, the size, or break it up into a heading and sub-heading.
- Images – Use a different image (test images with and without people), remove an image, add one, or put them in a different places.
- Call-to-Action – If you have a “click here” type button, add another one higher up in the design, or make it bigger, change to a contrasting color, or change the wording – make it more descriptive.
- Text – Break up the text into smaller paragraphs, use bullet points, change the font style/size, or change the text to focus on a different aspect of your service/product.
- Timing – Send at different times of the day or different days of the week.
- Layout – Change a multi-column design to a one-column layout or vice versa.
- Subject Line – Change the wording or add the person’s first name.
- Add an element – Put in a couple testimonials, PS at the bottom, or personalization.
For it to be a true A/B split test, only one element should change for a single campaign – that way, when you see the difference in response, you know exactly what to keep for the next one. If you change more than one thing, you won’t know what worked, and that defeats the whole purpose of testing.
And be sure when you send the second (and third, fourth, etc) campaign, do another A/B split to test something else!