How to A/B test your email marketing

As an email marketer, you’ve probably had a meeting with your boss where you’ve heard “Let’s A/B test it.” So, what exactly is A/B testing? How can you do it right in email marketing? How can you measure it?

A/B testing is a smart way to optimize your results. Here are some tips to better understand the process and slowly but surely improve your email marketing metrics.

What is A/B testing in email marketing?

In email marketing, A/B testing, also known as split testing, is the process of showing two versions of the same email to different segments of your customer base and tracking the results. The best way to ensure accuracy and fairness in your results is to make the segments similar for appropriate comparison. For example, splitting your list randomly in half is a perfectly reasonable way to conduct an A/B test.

The importance of A/B testing

In email marketing, you only get one shot to get things right. Once you press send, that’s it. The email goes into your subscriber’s inbox, and you can’t change it. Unlike website development, where you can make constant changes and improvements to your user’s experience, email is out of your hands once it goes live. For this reason, A/B testing is one of the best ways to ensure you’re learning from your email campaigns and making improvements for future ones.

Additionally, email campaigns typically vary a great deal from one to the next. The subject lines, copy, images, calls to action and other elements usually change from campaign to campaign, making comparison here apples to oranges. If you create some variation within the individual campaigns, however, you can truly learn what performs better by isolating single variables.

The different elements to A/B test in email marketing

An email is much more complex than what meets the eye. Dozens of inputs go into creating a single email, from the subject line to where you put the unsubscribe button and everything in between.

In order to properly A/B test your work, you should isolate individual variables so as to not get biased results. For example, if you want to see if a particular image performs better than an alternative one, don’t change the text above the image. If you do, you’ll never truly be able to tell which element the reader was responding to.

So, which variables should you test? The list below is not exhaustive, but it should give you a start. Once you do, you’ll have a great foundation to make incremental improvements to your email program.

Subject line

The simplest and perhaps most important element to test is your subject line. More than anything else, it determines how many people open the email and, therefore, how many eyeballs you get on your message. You can’t sell anything if nobody opens it.

You need to test the subject lines on almost every email you send. Many Email Service Providers (ESPs) have the ability to send test messages to a smaller group (say, 10 percent of your subscribers) an hour early. You can then send the email to the rest of your audience based on which subject line the smaller group opened more.

Be wary of finding a formula and sticking to it for too long. A formula on a subject line works…until it doesn’t. If you’re always testing and iterating and getting better, then your subjects won’t get stale for your subscribers.

There are infinite ways to split and segment your list, but if it’s your first time building a calendar, don’t overdo it. Stick with your core business, and once you start to see some results, then you can get more complicated and perform some segmentation.

Preheader

The preheader (also called the preview) is that short text you see next to the subject line before you’ve opened the email. The preheader acts as an extra shot at a subject line. It can give your subscribers a little more insight into what’s inside the email.

The most common technique for using a preheader is to include a high-level summary of what’s in the email. Another method is to use the first five to 10 words of the email as the preheader. If you’ve settled on a subject line, try some different preheaders and see if they improve your open rates.

Headlines, images, body text, links

Now, we’re in the body of the email, and we’re testing for effectiveness after a subscriber has opened it. Typically, an email contains a headline, image, body text and links. You have two options for testing the inside of an email:

The first approach is to isolate one variable and change it. For example, your headline in Test Email A could be “Memorial Day Sale” and your headline in Test Email B could be “30 Percent off for Memorial Day.”

Alternatively, you can create a completely different email. Since the subject line will be the same, you can test the effectiveness of a different style of email. If you haven’t experimented with a plain-text email that appears to be from the CEO of your company, try that style against your usual marketing email templates.

Audience

Everyone believes they know more about their business’s audience than anyone else, but the reality is that your audience is constantly evolving and changing. If you immediately place a customer into a particular segment after their first order, you are restricting their opportunity to grow and see other products your business offers.

For example, if a customer’s first order is well below your usual average order value, don’t automatically put this customer into a low-spender segment. In this scenario, you should send a test email with higher-priced goods to your low-spenders and see how it performs.

What you may find is that a customer’s first interaction with your brand is not indicative of future behavior. Perhaps they were just testing the waters with your brand, spending a little bit in the first order with plans to spend a lot more if the experience was positive. Be flexible in testing your audience, and you may find that it’s less predictable than you imagined.

Time of day/day of week

Do you have a set cadence for your emails? For example, you might always send newsletters on Monday, a promotion on Wednesday and a product feature on Friday. If so, that’s good; you have some consistency.

Now, change it up! Send your newsletter on Sunday to half of your list and the rest on Monday. Send the Wednesday promotion at 6 a.m. instead of 9 a.m. Not only will you learn about how to optimize your email schedule, you will surprise (and hopefully delight) your audience.

How to measure success

Of course, testing your email campaigns is only effective if you know how to measure whether or not the tests are working. Each variable above corresponds to at least one metric that either directly or indirectly leads to more sales.

When you’re testing, there are three main metrics to follow, depending on your business goals: open rate, click rate and conversion rate.

• Open rate refers to the number of emails that recipients opened out of the total sent. To track this, you might measure which of the subject lines you’re testing led to more opens.
• Click rate tracks either the number of clicks per emails sent or the number of clicks per emails opened. Either way, here your goal is to measure which elements drove the most clicks in your send.
• The conversion rate is the number of orders you ultimately received based on the number of clicks. You want to determine what drove customers to actually buy your product or service.

If your goal is to increase eyeballs into your brand, optimize for open rate. If you’d like more visitors on your website, test for the number of clicks. If you ultimately need to add revenue, focus on conversion rate.

Even a small test can lead to incremental growth, whether that means a few more eyeballs opening an email or a few more orders that get you to your monthly goals.

Start testing now!

The next time someone in your company tells you to A/B test an email, now you’ll have the tools at your disposal to succeed. Don’t let your emails get stale. The idea that you can “set and forget” your email program is a common misconception that leads to diminishing returns.

Instead, continue testing and iterating and find new pockets of opportunity, one variable at a time. Try a flashy subject line tomorrow against your usual formula, and you may be surprised by the results.

This piece was originally published on BuiltIn.