Getting your email marketing right takes a bit of doing and we like to think we’re getting it nailed. So it was great to have this reinforced with some international recognition in SitePoint’s latest book on creating stunning HTML emails. An email reminder to an event back in March 2009 (we like to think we’ve evolved further since then) was one of 15 Gorgeous and Effective HTML Emails featured.
The article outlines key principles behind successful emails, listing things such as having clean layout and a clear call to action. The idea being that if you follow these guidelines your emails will be appealing, and people will respond more to them.
In our experience (and that of loads of other companies) it’s not just about making things look and sound great to get the best results, you need to go one step further and it’s all about testing – testing everything you can.
Many of you have probably heard the term A/B testing or split testing. You can do it with websites, emails, phone calls and even the homeless. In the case of email, the most basic way to A/B test something is following these steps:
- Identify a goal you want to improve on, e.g. increasing replies or clicks on a link
- Make sure you can measure the result by using things like promotion codes, different link destinations or cool marketing software that does it all for you
- Send half your customers the original email
- Send the other half a slightly modified email
- Check which email version improved your goal the most
Just remember, while you might be tempted to make lots of changes to improve your goal faster, it’ll be harder to determine what particular change did the trick.
Here’s what we did
We decided to A/B test something that’s just a small thing and often overlooked – the email subject line. We wanted to see if a company name at the start of the subject line encouraged more people visit our website.
Version A (red) had no company name in the subject line, and Version B (blue) had “Xero -“ added to the start of the subject line.
Having the company name in the subject line saw more people open the email, but the difference was so small it’s unlikely it meant anything. However, by not having a company name in the subject line meant we received around 70% more clicks through to the website.
What does this mean?
There’s a bit of guesswork in reading the results. Sometimes the answer is clear – a bigger button might attract the reader’s attention more. In this case it may be that the email felt more personal and human without a company name at the start, and people were more receptive to the contents of the email. Or it could’ve been sheer luck, and by sending the email again to a few thousand more people we’ll see the new data start to average things out to show a smaller gain.
The important lesson to take from all this is that by continually testing and measuring everything, even the very small things, you can discover your own best practices in everything you design. On their own these percentage improvements may seem tiny and disconnected, but compounded over time they can have a huge impact on the success of your business.
Things you can start testing now:
- Call to action
- Page copy
- Arrangement of email elements
- Timing – try sending emails at different hours or on different days