You are about to read a practical guide of the different email marketing A/B tests you can start running today, which will help you to improve your user engagement rate, email click through rate, and generate more revenue.
Table of Contents
- Why Would You Want to Run A/B Tests?
- A/B Tests to Get You Started
- Email Examples
- Keep Track of Your A/B Tests
- How Do You Know if Your A/B Tests Were Successful?
- Closing Thoughts
Why Would You Want to Run A/B Tests?
You might be generating more than 30% of your ecommerce/DTC business annual revenue through email marketing already (that is the goal for all of our clients), or maybe you are not there yet. A great way to get to that sweet spot is to constantly run A/B tests in different areas of your email campaigns, including copy, email timing, email content, and the types of emails you send to your customers.
Our goal is not to generate a 10% bump in orders right away, that would truly be unrealistic. We want to make a list of all the items we can A/B test on and then get started with our tests, run at least one or various of them at a time, wait for results and choose the winner. Then do this over and over again.
By the way, running A/B tests is a great way to build a stockpile of the many things that have worked in the past and those that didn’t. That way If at any point your email marketing performance drops you can quickly try out some of the things that were successful in previous tests.
A/B Tests to Get You Started
You would be impressed by how important your copy is for the success of your email marketing campaigns. A great subject line can double your open rate and a clear convincing call to action can at the same time double your click through rate.
- Subject line
- Preview text
- Body copy
Think about your entire copy as an orchestra where every element complements each other in order to reach that one goal (which is usually getting people to click through). You must first cause curiosity through a great subject line + preview text (that will get them to click on the email). Tell an engaging cohesive story to make them interested, focus on handling objections and reducing friction by showing them proof that what you are saying is true (this will get them to read the email) and finally close with a clear action-oriented CTA which tells them beforehand what will happen once they click there.
Here are some of subject line examples:
Focus on what the product does for your customers:
“Get a beautiful and healthy face with our all natural revitalizing serum”
You can also start with a promise and then handle objections:
“How to get an athlete like figure even if you are busy or have a bad diet”
“Cook like a chef with our new cookware lineup, even if you are just learning to cook”
When writing copy don’t focus on what your product does, focus on what it can do for your customers, and then clearly handle all the objections they could come up with. Your copy will convert if you can clearly and succinctly communicate those 2 points.
People open their emails at different times of week and at different hours, your target audience is unique and while you may find online different studies on “when is the best time to send your email marketing campaigns”, please take those with a grain of salt and test both the days and hours that work best for your business.
- Day of the week: Start by sending emails 1-2 times during weekdays and 1 email during the weekend, and monitor their performance. Send your emails the same weekdays and the same day during the weekend, do that over a month, and then switch by sending your emails a different weekday and a different weekend day to see if your metrics improve.
- Hour of the day: Start by testing emails either in the morning, afternoon or at night. Send your emails at exactly the same time for a few weeks or a full month, then switch and test with another hour of the day. Keep the time that works the best, sometimes it is not what you expected.
There is something really important to have in mind, some brands may want to send an email a specific day of the week, for example an update email every single Monday. Customers get used to it and come to expect that Monday weekly email, if this is the case for your brand and your weekly email is performing very well, then there is no need to test with another day of the week.
Number of Emails
Email burnout can happen to any brand, if you are sending an email every single day of the week your emails will most definitely end up in your customers spam folder, or simply being ignored. If you don’t know the correct cadence for your emails, simply start with 2-3 emails per week and monitor your open rates, click through rates and orders. Then switch over to 4 emails per week and monitor the same set of metrics over a month, if you see that your metrics start to decay, switch back to 2-3 emails per week. You can do this with up to 5 emails per week.
Please have in mind that you should not do this kind of test during holiday seasons, or whenever you are running a promotional campaign, since it will skew the results.
Types of Emails
Many brands think that email marketing is all about promotion, so all they do is send sales emails, one after the other. This rarely works since email marketing is about building a relationship with your customers, always think about adding value to that relationship over time.
It is okay to send sales emails as long as you also send other types of emails. You can start with educational emails on how to properly use your product, you can also introduce your referrals program (if you have one) to your most loyal customers. Try sending an email without any images, simply include your brand’s logo and get the CEO to write as if they were speaking to a friend. You can also tell your customers about that great community program your brand is currently involved in, that new partnership that you just signed and how it will benefit them, or simply thank them for being your customers.
Customer Thank You email from Aftershokz
Brand Partnership email from Unsplash
Product How To – Educational email from Dossier
Referral Program email from Scott’s Flights
Keep Track of Your A/B Tests:
This one might seem daunting but is actually pretty simple. Create a spreadsheet and include the following columns:
- Type of test (copy, timing, number of emails, or type of email)
- Test description
- Metrics to improve
- Start date
- End date
- Additional comment
This is just an easy way to get started, so feel free to customize it to your liking.
How Do You Know if Your A/B Test Was Successful?
Your top 2 metrics will most likely be Click-Through Rate and Orders/Revenue, unless you have another goal like number of referrals or the number of loyalty program signups.
Bear with me in the following part, successful tests are measured on something called statistical relevance. Let’s say you are A/B testing subject lines, and you send the test emails only to a very small amount of people, an audience so small that it is impossible to conclude with 100% certainty that changing the headline of the email was what caused the second email to perform better than the first one.
You want to run tests with hundreds if not thousands of email users, the bigger the size the better, since you will have a greater degree of certainty that whatever it was that you A/B tested was the actual cause of the improvement.
What do we recommend to our clients? Run A/B tests on a monthly basis, always and I mean always have at least one A/B test running, why? Continuous small gains lead to huge wins over several months, you might increase your click-through rate by 0.5% with one test, but stacked over time it can grow to a double digit gain for your business.
If you want to learn more on how to improve your email marketing and achieve that sweet spot of 30% of your annual business revenue coming from email, read our blog post Email Marketing an Introduction to Email Flows.
If you are interested in learning more about email marketing metrics and how to use them to your advantage read our blog post Key Email KPIs for Ecommerce & DTC Brands.
Thank you for reading.