What is A/B Testing

A/B testing is when you change one thing like a button shape, title text, or the color on a call to action area in a project, so you have an “A” version with an orange button for example and a “B” version with a blue button then share project A and B versions with two different groups. This is frequently used in email marketing by splitting an email list in half into two groups and sending a slightly different version of an email to each group to see how the emails test. If say more people click on the button when it is orange, then you have a data-driven reason to choose one color over another.

Why Should I A/B Test?

Many designers and marketers have a big challenge in proving not only the value of what they are doing but also the validity of design choices and the true impact a design decision can have on the success of a project. Moira Cullen used focus group responses, A/B testing, and other tools to gather real data to support the design decisions of her teams. This is highly effective when dealing with people who are more comfortable with numbers. If you have been struggling with persuading business people that your design decisions or other creative choices have merit use objective supporting materials to refocus the conversation away from subjective topics and back onto goals. While this can be taken to extremes (Google is known for valuing data above any other factor), many times it is a great way to show you aren’t just guessing or using some nebulous creative muse – you have hard data to back you up.

What Should I A/B Test

You can A/B test anything that you can change. Here are five items people often A/B test for infographics or call to action pages:

  • calls to action (statements or phrases)
  • infographic titles/names
  • key data presentation (a chart or visual vs. text or a number)
  • color of any interactive buttons
  • link or button activity

How to I Measure the Results?

Email marketers and web designers use goals and funnels. Click rates can also be measured. One of the best methods that can be used across print and digital media is creating a unique url or code. Ever heard or seen an add that mentions a promo code you can use or a special promo url? Media advertisers use them to test as well! They can measure the success of a specific campaign or regional version of a campaign to a type of audience or market this way. So here I have a list for you of common tools used to measure response and results:

  • Funnels and Goals in Analytic tools like Google Analytics
  • Smart lists in marketing automation tools
  • Unique urls
  • Promo codes (user entered or as part of a scannable QR code or coupon)
  • Unique contact emails or phone numbers
  • Heat maps for web pages to show where users focus most
  • Focus groups /User testing
  • Surveys

This type of testing does require a bit of extra effort, but when the stakes are high, having data to drive decisions or verify choices can be critical. If you haven’t used it before, start out small. Have two web pages and make the call to action button a different color or change the text. Send a different version of an email campaign to a section of your list, and swap out a chart for a lifestyle image. Start collecting results in lower impact areas can give you a lower risk way to experiment with you options.

Do you struggle with knowing how to consistently engage responses from visitors?