Once you’ve effectively planned and identified goals for your A/B testing emails, it’s time to start executing your tests. While running an A/B test can be fun, exciting, insightful, and rewarding, you can’t click send then immediately start analyzing the results. Ensure that your schedule provides enough time for the A/B test to produce reliable results—a minimum of three days is often a good benchmark before determining a winner.

Once you’ve allowed for the test to run its course, you can then analyze your A/B test results and optimize your tests and email marketing to better reach your prospects and customers.

1. Schedule your A/B test then let it run its course

During that testing period, avoid making changes to your test elements. It’s fun to watch A/B testing and you may want to change elements so that there is more behavior around what you are testing, but don’t do it. You will not have reliable results in the end.

The actual timing of when you send your tests is also important. If you’re testing two different subject lines, you want to send the two test emails to the test groups at the same time on the same day. If you send Subject Line 1 to Test Group A on Monday at 10:30am and Subject Line 2 to Test Group B on Tuesday at 7:15pm, the timing of the email will more than likely cause different results, so you won’t know which subject line actually performed better.

With that, it is possible for the element you’re changing in your tests to be day of the week or time of the day. If that’s the case, you’ll want to send the same email (subject line, body copy, layout, imagery, etc.) at the specified days and times. If you are testing to see which day of the week is better to send emails, you’ll want to send the same email to users at the same time of day on the specified days you’re testing. For example, send your email to Test Group A at 9:30am on Monday then send the Test Group B email at 9:30am on Wednesday to test if Monday or Wednesday is a better day to send your emails. Once you narrow that down, you might then test Wednesday at 9:30am versus Wednesday at 6:30pm to focus in on the best time to send emails on Wednesdays.

2. Analyze your A/B test results

While identifying your A/B test goals, you should also determine what a “win” means. Do you just need a simple majority to declare a winner or does one test variation need to win by 25% to be deemed the winner?

Sometimes an A/B test will give inconclusive results. For example, you may have a 49% reaction for Test A and a 51% reaction for Test B. Is that 2% difference enough for you to say that Test B was definitively better than Test A? That’s up to you to determine but deciding those requirements before testing will allow you to avoid hindsight bias. If your results are close, you can always run another A/B test, and this time, change the element being tested to see if you get more conclusive results.

3. Optimize your email marketing with A/B tests

After you have the winning test declared, ensure that when you send the actual email (after testing) that you are sending the correct “winning” email. Additionally, one A/B test might not be enough to fully optimize your emails, so continue to run unique A/B tests to experiment across different elements in order to hone in on your optimal email. And, remember, just because your A/B test showed a clear winner this time, doesn’t mean you shouldn’t test different elements and variations in the future—marketing is always evolving and so should you!