The benefits of split testing
Measuring is about knowing how things really are, instead of guessing. But it's also, possibly even more beneficially, about finding out what really works, instead of guessing. Read on to find out more about the benefits of split testing.
It’s one thing to know what your lead conversion rate is, on average, from month to month. But it’s another thing entirely to know what works to boost your lead conversion rate.
I’m still realising how I can use measures to track, test and tune my own business. One thing I measure is the sign up rate to my membership program. It’s simply the percentage of my email newsletter subscribers who sign up to join, when I send a monthly update of the next members-only event. Yes, it’s good to know this. But let me tell you how I managed to *double* it with five minutes of work.
It happened when I opened an email in my inbox which was an invitation to sign up for a free webcast. I’d seen plenty of these before, but looked at this one differently that day. The thought came to me: I wonder if this company uses this layout because they know it works best? Their layout was quite different to the layout I used for my webcast invitation emails, but our target audience was exactly the same.
Theirs was structured with headings throughout and not at all personalised. Mine, however, was not structured and more conversational, and I personalised it by greeting my subscribers by their first name. My approach was supposed to be best practice, or so it was for those I learned it from. But I figured, why not find out for sure?
Want more articles like this? Check out the measuring success section.
I used what’s called split testing, which meant splitting my subscriber list into two randomly chosen groups, and sending my original email layout to one group, and a new one based on the structured layout to the other group. Then I looked at the sign up rates for each group.
"If you want better results, pick very specific potential improvements and split test them."
The new structured layout got more than twice the sign up rate of my original layout. More than twice! Needless to say, that’s now the layout I use all the time, for webcast email invitations.
The point is not about how to lay out your emails! That will vary depending on your business type, and more importantly, your target audience. And nor does this idea of split testing apply just to email marketing.
The point is this: if you want better results, pick very specific potential improvements and split test them. Pick the improvement that tracks the best, then test new ideas against that one. Keep tracking, testing and tuning until you get the results you’re happy with.
What can you do now?
Think of one result you’d like to improve in your business. Set up a measure to track it. Then come up with 1 or 2 ideas for improving it. Now test those ideas using the split test concept.
Have you used split testing effectively? Were the results surprising? Let us know!