Testing Techniques Part 11: A/B Testing

If there was a fly on your conference room wall, would it ever overhear a conversation like this? “Cool. Let’s do it this way.” “But wait, what if this way is better?” “Well, we have to pick one. So, which do we go with?” (Conversations like this happen a lot here at We ❤ Health Literacy Headquarters.)

Fortunately, in this latest installment of testing techniques, we have a solution for you! It’s called A/B testing, and we really ❤ it. With A/B testing, you put 2 versions of a message or design out into the world — that’s the “A” and the “B” — and gather data about which version is more effective or impactful.

Emails are prime A/B testing material — you can test different subject lines, calls to action, or designs. Let’s say, for example, you want to know which type of email subject line will get more readers to open your e-newsletter.

First, pick the types of subject lines you want to try out. Maybe the first takes a question-style approach and the second is a bit more clickbait-y, like so:

  • Option A: Does someone in your family have allergies or asthma?
  • Option B: 5 ways to help a loved one with allergies or asthma

When you have an A and a B, you’re ready to test them out. Using your email delivery system, send a random sampling of half your subscribers the email with subject line A. Send the other half of your list the same email, but with subject line B.

Then, look at the analytics and you’ll see which subject line led to a higher open rate. Pretty cool, right? The more times you do this, the better sense you’ll get for which type of subject line readers prefer. After that, as they say — lather, rinse, repeat.

Note that it’s best to do A/B testing with large numbers of participants — like your organization’s e-marketing list — to make sure you get a meaningful response.

And one last thing: with A/B testing, it’s important to eliminate variables. For example, if you’re trying to gauge the click rate with a new layout of your e-newsletter, use the same content in both options. Otherwise, your findings could be muddled. Did users respond to the layout or to a different content approach? You’ll never know.

The bottom line: Don’t know which way to go? Use A/B testing to explore 2 ideas at once.

Tweet about it: Don’t know which way to go? A/B testing to the rescue, says @CommunicateHlth: http://bit.ly/2o9MGLE #HealthLit


Browse recent posts

Do you heart health literacy? We sure do! Sign up to get practical health literacy tips and tricks — delivered to your inbox every week.