The Idea

What is a split test

A split test is when you get data from doing things one way (A) versus doing it another way (B). That’s why it’s commonly called A/B testing. There are many tools that allow you to split test your social media ads, website copy, email campaigns and most other forms of digital marketing. But split testing can and should be done with physical forms of advertising as well, including proposals, direct mail etc.

Image Credit: Optimizely.com

How to do an A/B test

  1. Create Goals: In order to do this right you’ll have to have a goal like sign-ups, conversions, phone calls etc. You’ll also need a way to track the results like a special discount code, software to track it or just asking verbally.
  2. List things that may affect conversion: Most parts of your ad or collateral can affect the effectiveness of the campaign. The common things to test are headlines, images and placement but keep an open mind.
  3. Change one major thing at a time: The idea here is to change something in a controlled way and gather enough data to know which one did better. For example you could create two webpages that are the exact same, except
  4. Measure results: There are many ways to measure which version performed better. It doesn’t matter which way you choose, as long as it’s scientific. Using a specialized tool that does this for you is the easiest way to save time and money in the process.

Thing to Test

  • Website images and copy
  • Button color and placement
  • Tone/style
  • Headings
  • Call-to-actions
  • Email follow-ups
  • Requests for referrals
  • Speed to respond

Metrics

  • Ease: Tough
  • Cost: $10+
  • Scope: High
  • Speed: Normal

  • Reach: Global