A/B testing, or split testing, is the process of testing two or more variations of something to see which version works best. In email marketing, this usually means testing different subject lines, different images, different layouts or even different copy. If your database is large enough, some email marketing tools will let you send the test to a percentage of the database first, and then send the winning version to the rest of your database. This all sounds great, but if your software doesn’t handle the process automatically, it could be extra work. So why is A/B testing your email campaigns important, regardless of any extra effort?
Small changes make big differences
Changing the wording of a subject line slightly or choosing a different colour for your call-to-action buttons might not seem like a lot, and it probably isn’t much, but these little changes can make an enormous impact to your email opens, clicks, and ultimately, your website conversions.
Let’s put a few metrics out there. Let’s assume:
- your open rate is 20%;
- your click through rate is 3%; and,
- your website conversion rate is 5%.
If you are emailing 1,000 customers that means you are getting 0.3 conversions and would need to send out 3 emails to get a conversion. It also means if your database grows to 10,000 you would make about 3 conversions off each email.
However, if you can increase your open rate to 25% you are now getting closer to 4 conversions off 10,000 subscribers. If you can also increase your click through rate to 4%, you are now up to 5 conversions.
As your database grows, those small percentage increases can start to add up to big dollars. Every test you do should help to increase one or more of those metrics so you can make more sales
Plain text or HTML email as a test example
MarTech companies are constantly researching the effects of changes in emails, so it’s tempting not to test with your own audience and take their results as gospel. For example, HubSpot have surveyed users multiple times over the years on whether they prefer HTML or plain text emails. The results typically come in with a majority indicating a HTML preference.
However, HubSpot notes that in every one of their tests, the simpler the emails are, the better they perform. Every image that gets removed improves metrics. This doesn’t mean going all the way to plain text is always best, but it does mean keeping emails simple and smaller in size should, get you better results.
A quick Google search for “Are plain text or HTML emails more effective” yields countless results suggesting the same thing.
Put it to the test with A/B testing
As we like to do, we put it to the test with our audience using the A/B Test functions in SharpSpring. I created two emails, exactly the same copy and subject lines. One was a simple HTML email containing basic formatting for emphasis (bold, coloured links etc), and a HTML footer.
The other was a bigger HTML email with am image header, a preview image of the offer, the same text links and a download button.
We decided to run the test against 2,000 records in our database with about 1,000 people receiving each version. What was the outcome?
Here are the results from our A/B test:
One was clearly a winner, and it was not the simple one.
Based on the data that the internet suggests is ‘correct’, I would expect the simpler email to perform better. However, the longer, larger HTML email got almost 3% more opens, half the unsubscribes and 1.6% more clicks/deliveries or 5% more clicks/opens. On our test of 1,000 people, as you can see in the screenshot above, that is an extra 18 clicks. It doesn’t take much and we are starting to see more conversions just by more effectively engaging our database.
Does this mean that we always get the same result? No, we’ve done tests on emails with different goals and got different outcomes.
So why is A/B testing important for your email campaigns?
Every database is different, every market is different, and every goal is different. Data from tests done by others is great and can be provide a lot of insights that can save time and money. However, if we relied solely on the data collected by others, we could be missing out on opportunities that are relevant to our database.
After the results of the A/B test above, we created another test and got a similar outcome. We didn’t expect to see a better result for HTML emails, but we got it multiple times when our goal was to encourage readers to download our free resources.
This does make sense with our buyer persona and the offer in these emails. We are a B2B business, and a large percentage of our ideal customers see our emails on a desktop email client where internet speeds have less impact and where the HTML may be rendered more accurately. The offer is a free download, and while that can be accessed on a phone, we have noted our database does tend to take these offers up more on a desktop.
By the way, if you don’t know your ideal buyer, you can grab our free template and learn how to build your own ideal buyer persona here.
However, for our regular shorter email blasts we find plain text emails work better with our database, market and goals for these emails.
So why is A/B testing important for your email campaigns? You could be missing out on opportunities!
Most MarTech tools, like SharpSpring, can run split A/B tests with minimal effort, so if you are spending money on one of these tools, you want to get the greatest return on your investment. A/B testing helps achieve that by boosting your open rates, click through rates and ultimately your conversions while reducing spam complaints and unsubscribes.
Whether you choose plain text emails or HTML, always be testing.
Want to learn more about the Revenue Growth Automation platform, SharpSpring by Constant Contact?
Interested in learning more about how SharpSpring can streamline your email A/B testing?
At Dovetail Group, we provide simple, yet effective revenue growth services and solutions to make everything work well. We have achieved excellent statistics for our clients and saved them thousands of dollars.
Interested in learning more? Start a Conversation.
