This is the fourth post in a series on how to make your email marketing strategy more effective. Last week’s post showed you how to use email segmentation to send the right message at the right time. This week, I show you why you should be split testing your emails regularly.
“Split testing? Isn’t that for websites?” I hear you cry.
Yes, split testing is commonly done on websites and landing pages. You should split test your emails too. After all, how do you know if something works unless you test it?
In case you have no idea what I’m talking about, let me answer a basic question first…
What is Split Testing?
Split testing is also referred to as A/B testing, when you’re testing 2 variables, and multivariate testing, when you’re testing 3 or more variables.
And it’s exactly what it sounds like. You split your audience (or a portion of your audience) and show each section a variation of whatever it is you’re testing.
As I mentioned earlier, most people think of websites and landing pages when they hear the term split testing. They think of buttons, headlines, colors, images and the like.
Your emails include these components too. Which is why you should conduct split tests on your emails.
Here are 5 ways you can use split testing to improve your email open and conversion rates.
1. Split Test Your Subject Lines
Your subject line needs to grab attention, so your subscribers will open and read your email. This makes it the perfect place to start your split testing efforts.
It’s also one of the easier parts of your email to test. Many ESPs and CRMs have subject line split testing built into their platforms.
For example, Drip allows you to test multiple subject lines when you create your email.
The results are displayed once people start opening your emails. You can see how two of my past subject lines did here:
I recommend split testing short subject lines against longer ones, controversial subject lines against more standard or “boring” ones and so on, to see what increases your open rate.
Yes, this does mean you have to come up with more than one subject line. It’s not as hard as it may seem. Check out my post on writing a good subject line here.
Wondering how much difference a subject line can make?
President Obama’s fundraising team did rigorous split testing before and during the 2012 campaign to see which of their subject lines were most effective. Their main indicator was how much money was made off each email, based on the subject line.
The email marketing team would test as many as 18 subject lines on small groups before sending the one that got the best results to their main list.
Their winner?
“I will be outspent” brought in more than $2 million.
They also found that a given subject line only worked a few times before they had to find another angle. Yet another reason to keep split testing.
If you want to learn more about what Obama’s team did and how they did it, check out this article in The Atlantic.
2. Split Test Your Email Copy
Yep, this too.
You can test everything from one or two lines of text to the entire email. Small differences can yield huge results. So can sending entirely different messages.
The simple way to do this is to select a few small segments of your list and send one of them your “control” or original email, then send the others the variations. Whichever one gets the best results is the one you send to your full list.
By the way, this is how major direct response mailing houses test their sales letters. It does work.
Your ESP or CRM may also allow for split testing your email content. It’s usually in the same section as split testing for your subject lines.
Here’s another example from Drip. They let you test your design and copy elements in one function.
3. Split Test the Length of Your Emails
I’ve separated this from the email copy section because I consider this a different type of split testing.
If you subscribe to my emails, you’ll notice sometimes my Thursday emails (with information that goes exclusively to subscribers) are fairly short, other times they’re quite long. I’ve found that my subscribers like longer emails, so I’m happy to accommodate them.
Your subscribers may prefer shorter, more succinct emails. It’s worth testing. You can figure out the results by looking at your analytics over time.
4. Split Test Your Email Design
“Umm… It’s an email…” I hear you say.
Yes, it is.
And many companies are adding more design elements to their emails to attract your attention. For example, I recently got this email from Truekind.
Candidly, this is a LOT of images for an email in my opinion. This design could get your email caught in the spam filter of your subscribers’ ISP. But it is eye-catching and will attract attention if it makes it into their inboxes and onto their preview screens (primarily on desktop).
It’s worth split testing design ideas like this in your emails, like:
- How many images can you use safely in one email?
- What is the ideal image to text ratio?
- Do your subscribers respond better to images or to text?
Other design elements to test include:
- Buttons vs text links
- Header vs no header
- HTML vs plain text
Regarding that last one, most ESPs and CRMs let you give your subscribers the option to see an HTML version of the email online. Giving them this option is a good way to test whether plain text works better for your subscribers.
An email has to jump through a lot of virtual hoops to get to someone’s inbox. Your subscriber’s ISP may filter your emails into the promotional or spam folders based on the design elements.
5. Split Test Your Call to Action
This is the part of the email you really want people to pay attention to, right? Split testing tells you what works best.
I’ve mentioned copy and design earlier in this post. These elements both play into split testing your CTA. Here are a few good tests to run:
- Button vs. text link: Which one draws your subscribers’ eyes and gets them to click through to your offer or content piece?
- Button color: Yes, this matters a lot, and it’s something marketers test constantly on websites. Test it in your emails too.
- CTA Copy: What words to your subscribers respond to? Do calls to action with a lot of power words work well, or do they prefer “Click here to learn more.”
To “learn more” about writing a good email call to action, check out this post.
6. Split Test the Day and Time You Send Your Emails
There is a constant battle among marketers to send their emails at the “perfect time,” so they’ll get read.
Candidly, there is no “perfect time.” And as soon as someone announces there is a “perfect time” everyone jumps on that minute, which results in a flood of emails coming in and yours getting lost in the fray.
The best way to figure out what day and time to send your emails is…
You guessed it, split testing.
Because every audience is different.
If your audience is made up entirely of CEOs, VPs and Chairpersons of major corporations, they spend most of their day in meetings. They’re more likely to check the emails that are filtered to them by their assistants between 5:30 and 8:30 in the morning.
If you cater to busy moms, they spend their day focused on their kids, their jobs and their households. They’re not likely to check their email until after the kids are in bed and they’ve got a moment to themselves.
I have a client who sells to doctors. I write 3 different subject lines for each email they send so they can send the same email 3 times during the same day.
They do this because doctors have a very short window to check their email first thing in the morning, again when the middle of the day, and again late at night and they want to be at or near the top of their subscribers’ inboxes.
You should know enough about your subscribers to make an educated guess as to when they’ll be most responsive to your email landing in their inbox. Start testing around that time.
Your ESP or CRM should show you when the most people open your emails by time of day. It’s usually in the Analytics or Statistics section (whichever title they use).
Now You Understand Why Split Testing is So Important
If you’re not already using one of these split testing ideas, you should be. If you are using one, add a couple more to your overall email marketing strategy.
Your goal is to get your subscribers to open and read your emails, then take the action you want them to take. When you test each element in your emails, you get more information that can help you reach that goal.
If you’re not reaching that goal, it’s possible there is a fundamental problem with your emails. One that’s far deeper than split testing can solve.
Click on the button below to find out if your email marketing strategy suffers from one of 5 fundamental mistakes by taking my quiz, What’s Your #1 Email Mistake?
Next week, I’ll wrap up this series by exploring how email automation can make your life easier.