As many marketers know, A/B testing is an invaluable tool to understand your target audience. By analyzing your own marketing campaign data with A/B testing, you can run your campaigns in a truly results-focused way.

Only One Item Should Be In Play

It’s vital that the A/B test is simple. You are only testing one element of data: the “A” as the control (no change), and the “B” as the variable. There are not other factors in play. When everything else is the same, and one variant performs better, we can say, without a doubt, that the element we changed was the cause.

learn to a-b test

E-mail campaigns are a great place to start when looking to run A/B tests to improve your e-mail metrics. Why? Because e-mail metrics are straight-forward, and the changes needed to improve your e-mail campaign results are clear after testing.

How to Run a Simple Email A/B Test

Step 1: Identify

To begin an e-mail campaign test, you must first identify which e-mail metric you want to improve. There are three primary metrics that tend to mesh well with your likely KPIs:

  • Open rate

  • CTR

  • CTR to Opens

It’s easy to find industry benchmarks for each of these metrics with a simple Google search if you want some guidance on where you should be. Here’s a great reference point on email marketing metrics.

Each of these metrics gives you specific insight into how your audience interacts with your e-mails. Your e-mail open rate shows whether or not your audience recognizes your sender address, is compelled by your subject line, and/or views your e-mail communications as valuable. Assuming you only have one link in your email, that of your CTA, your CTR should indicate that your audience understood, located, and was compelled by your CTA. Finally, your CTR to open rate gives you a better understanding of your overall engagment rate as it relates to Opens.

Step 2: Choose

Once you’ve analyzed your metrics, it’s time to identify which metric you would like to test. For example, if you feel you have engaging content but have a low open rate, you would run your A/B testing to improve your open rate. There are several ways to do this, but the most important is to only test a single factor at a time.

Step 3: Implement

Try this: to improve your open rate, send your next e-mail campaign at a 50/50 split, send the “A” group an e-mail from your regular sender address, and send the “B” group an e-mail from a specific member of your company (CEO, CMO, etc.). Run the same exact A/B test until you reach statistical significance – think at least 100 emails.

Hubspot’s A/B testing sends the first 100 emails with version A, then tabulates the best performing variant and sends that to the rest of your list.

Step 4: Analyze & Update

Analyze your learnings from your A/B tests, record the results and determine if you’ve found any significant findings. If you run the above example A/B test with regards to Open Rate, did your “B” group open rate beat out your “A” group overall? If so, then you’ve identified a way to make your data work for you, and discovered a change to your e-mail marketing campaigns that can bring you better results.

learn to a-b test

A/B testing can be used on any marketing channel, but it’s incredible easy to implement on an email campaign. Once you get used to running A/B tests there, you should add in website pages, app interstitials, PPC variants and really mine the data for performance.


I have worked with many marketers over the years and you would be surprised how many are not running A/B tests as a matter of process (and how many marketers have never run one at all). They are very easy to run and will allow us to optimize marketing efforts around the best performing content (stretching a tight budget, making resources work smarter not harder and pushing results into boastable returns).

As marketers, we must understand the metrics we are collecting and what they mean – pulling KPIs isn’t fun so let’s remember that we aren’t doing it as a useless exercise but to inform our marketing spend/efforts.

data-driven-web-design