There’s an art to telling the story of your brand. Whether you’re trying to convey your message through a brief slogan or a more in-depth communiqué, finding the right combination of words to drive customers to your door can sometimes feel like throwing darts against a board. What you say might stick. But how do you know in advance whether it will work?
This is where art can be informed by science.
A/B testing different message combinations is a fast and inexpensive way to compare approaches and get data to support a business case for evolving your message. Even a failed test—one proving that a new idea doesn’t work—can yield valuable insights that lead to something better.
So, how do you A/B test your messages?
Three A/B test scenarios
If you have ample time and web-development resources, you can A/B test landing-page options and measure outcomes such as clicks, conversions and the bounce rate of site visitors. In this scenario, you can set up two somewhat similar landing pages to compare the new idea against the status quo. Half of the visitors to your site can be directed to your test page, and half to your control page. Even subtle differences in your page metrics can tell you a lot about whether the new variation is worth pursuing.
If you have some resources but little time, you might want to consider A/B testing with digital advertising. Most self-service advertising platforms, such as Google AdWords and Facebook’s Ads Manager and Power Editor, let you test two variations of the same ad to similar audiences. Clear results often can be seen within hours—not days or weeks—and you can further refine your test messages on the fly.
Finally, for the shoestring warriors who need answers now but don’t have the budget for advertising or additional web development, there is email marketing. Simply take two equal-sized randomized samples of email addresses from your database to create your A and B audiences, then send each group different message variations that you want to test. Click-through data will reveal if one option yields a better response. You can also use open-rate data to compare the effectiveness of two different message subject headers.
Test and test again
While it can provide definitive, black-and-white answers for two alternatives, A/B testing is most effective when used in an iterative process. Breakthroughs come from testing your existing message against new alternatives, measuring results, adapting and then testing again.
That’s the way to bring science to the art of communicating your brand.