Wanna know what separates the winners from the losers in digital marketing? It’s testing. Testing, testing, and more testing. Having an ongoing testing program has been endorsed by everyone from Peep Laja to Albert Einstein. Seriously. Einstein said:
“Compound interest is the eighth wonder of the world. He who understands it, earns it … he who doesn’t … pays it.”
An ongoing testing program leverages compound interest. Here’s an example:
If you start the year with 100 clicks per day (or 100 sign-ups or 100 orders, or whatever) and make just a 2% improvement to your conversion rate every week, by the end of the year, you’ll get 2.8 times the results every day – 280 clicks per day (or 280 sign-ups per day or 280 orders per day).
Don’t believe it? Here’s a table that shows compound interest and compound testing’s awesome progression.
That’s what testing can deliver. That’s why in the battle of Internet marketing, he or she who tests, wins.
For this post, we’re going to focus on testing for email marketing. There are dozens of other things to test in your business, but this is just one blog post, so I’ve got to focus. Besides, it’s actually a key part of success for testing that you only test one element of your funnel at a time. Testing two elements at a time can screw up test results.
Think beyond email subject line tests
The first thing everybody talks about when they talk about testing emails is subject line tests. Subject line tests are awesome, no doubt, but there’s more to test. There are also things to test that are actually more impactful than subject lines.
“Wah?” you’re thinking, “The subject line is the most important part of the email! That’s why everybody tests subject lines.” And you’re right. Completely on target. But you’re not thinking big enough.
You see, testing the elements of individual emails is kind of like testing the leaves and branches of your email program. I want you to think bigger – to think beyond the subject line – to start testing and thinking about the entire structure of your email program.
I want you to think beyond individual emails, and start thinking about the whole arc of how you communicate with your subscribers. Testing individual emails can get you a nice lift. It can maybe even double your click-through rate for that one particular email you’re testing. But testing your entire email program’s structure can double the response of all your email marketing efforts.
That’s why it’s best to test big stuff. As Mike Rhodes, Perry Marshall and Bryan Todd described it in their book, “Ultimate Guide to Google AdWords” – “First test forests, then test trees, then test branches, then test leaves.”
Testing emails in GetResponse
As you know, you can test lots of things in your GetResponse account. It’s pretty easy to do, and if you run into any snags, just pull up that 24/7 chat box and customer service will help you out.
We’ve created several resources on how to set up tests in your GetResponse account:
9 email tests that can yield big results
Now that we’ve laid out all that groundwork, here are a few big tests you can do:
Sending emails once a week? Why not test sending them twice a week? Or switch it up – if you’re worried about list fatigue, test sending half as many emails. You may find your sales drop by only 20%.
I’ve seen small retailers increase their sales by more than 30% by going from mailing once a week to mailing twice a week. Note that they didn’t double their sales, but getting 30% more from your entire email program is a pretty big win.
Often you can add more mailings without incurring a big loss in subscribers. This is one of the best ways to test how much your subscribers like your emails, too. If you increase mailings and see no drop to very little drop in your engagement rates, you’ve got subscribers who like your content.
2) Welcome emails
The welcome email is most important email you’ll ever send to your subscribers. It’s the message they’re most likely to open and read, and it sets the tone for every email after. And yet… way too many email marketers aren’t sending a welcome email.
If you are sending a welcome email, consider testing
- A series of welcome emails
- When you send the welcome email (instantly vs 2 hours after sign-up)
- A different offer in your welcome email (15% off versus a free gift)
This graphic from MarketingLand shows how sending welcome emails affected the lifetime value and engagement rates of subscribers.
3) Which day to send
There are plenty of studies out on which weekday is the best for sending newsletters. All those studies, even our own, are good and useful information. But none of them were tested for your audience.
Take the next step beyond generalized studies and actually test which day of the week is best for your list. Also consider testing which hour or time of day to mail, and test whether our “Time Travel” feature will work for you. It lets you send emails to your subscribers at the same local time for all time zones.
4) Autoresponder timing
Probably not all of you are using autoresponders, but you should be. And if you are, consider testing how many days apart your autoresponders are sent. Just like in test suggestion #1, you may find you can actually get better results by mailing more often.
Segmenting your emails is one of the best ways to jack up results fast. There are dozens of ways to segment a list, including
- Content preferences
- Click behavior
- Order behavior
What works best will depend on your list. But you’ll never know which segmentation strategy works at all until you test it. Check out our advanced segmentation video to get more ideas for this tactic.
These are the most popular ways for email marketers to segment their lists according to a December 2014 study by The Relevancy Group.
6) Sender name
An email’s sender name is more prominent than its subject line in many mobile device email clients. That means the sender name is worth testing. Just put some thought into your sender name tests – you want to keep them to a minimum. It’s important to keep your sender name consistent.
7) Text versus html
Text emails are definitely not dead. This is especially true now that we have the Apple watch and other “wearables” gaining market share. Some of those devices can read emails, but only text emails. Depending on your audience you might find that text emails outperform HTML emails.
Yup – you can send the same email twice. It’s not a tactic that’s talked about a lot, but I have seen a couple of top-tier content publishers resend their emails. One of them resends their emails every single week. Jordie van Rijn estimates you can get a 15-20% response rate by resending emails.
If you haven’t ever tried sending the same message twice, it’s definitely a test worth running… especially for those once or twice a year super-important emails.
9) Opt-in confirmation process
Do you use double opt-in? Do you know its dirty little secret? It’s that you’ll lose about 20-30% of your potential new subscribers to the opt-in process. They’ll enter their email on your opt-in form, but for one reason or another they’ll never confirm their address and actually get on your list.
This sounds like a problem. Actually, it’s an opportunity. If you run a couple of tests of your confirmation page and your confirmation email, you may be able to slash the number of people who fall out of your opt-in process in half. That means growing your list 10-15% faster, just by running a few tests. No other list-building tactics required.
Those are just a few of the big picture tests you can run, but they are some of the best tests to get started with. After you’ve gotten statistically valid results from these tests, then look into testing subject lines and other “leaves”. But first get the big stuff right.
What about you? Are you testing your emails? Even with a few subject line tests? We want to know how it’s working out for you. Give us some feedback in the comments.