Split Testing Email Elements

Digital Marketing - Study Notes:

Benefits of split testing

Why should you conduct split testing?

• Increased open rates: Split testing can help you with your open rates. It can help you increase them, because you’re evolving, you’re growing, you’re figuring out what works and what doesn’t.
• More targeted and relevant content: It helps you determine a more targeted and relevant content mix.
• Increased subscriber engagement: This is all about getting the people who subscribe on your list to actually engage with your content, to do something with it.
• More data-driven marketing: As with any data-driven initiative, your actions are going to be based on data, and they’re not necessarily on assumption. Your gut can only get you so far. Data will be the thing that drives your actions, will be the thing that helps you figure things out.

That’s where split testing can be very useful as a marketer. You get to figure out what works, what doesn’t, and you’re making subtle tweaks, small changes, one at a time to help you figure out what works and what may not work so well.

Gleaning insights

How do you glean insights from split testing?

In this particular example, you’ve got Variation A, Variation B, and you can look at conversions. And again, when you think about testing, this again is how you figure it out. You have to look at your metrics, look at audience, and look at whether the conversions were at a higher level with A or with B, and you just go from there. And you split it up.

A lot of testing that takes place these days, in addition to email marketers doing their testing, is via social media and apps. I see a lot of testing on things like Twitter or Facebook where you’ll have friends who are on Facebook or Twitter, and they see a bit of a different interface. They see a subtle interface change, and that design change is clearly A/B testing. There’s a lot of split testing going on in that social media space. And the same thing goes with email. You’re going to be able to split it up. And that isn’t to say that you’re going to favor Group A or Group B. You’re just going to be using the various splits as a way of figuring out what works the best.

And then eventually it’s going to let you hone in so that everybody’s getting the same message based on measurable data and measurable results.