The Do's and Don'ts of A/B Testing

A/B testing is a kind of website optimization test and is a simple concept. You have two designs of the website: A and B. Usually A is the existing site and B is the new design with a changed variable. Between these sites you'll split your traffic between them and measure their performance using certain metrics (bounce rate, sales, conversion rate etc.). Ultimately you'll select the version that performs best.

So why should you conduct A/B testing? A/B testing is important because it allows you to learn how to drive more traffic to your website and generate more leads from the visits you're getting. A simple change to your website can drive big results.

This may sound easy enough but there are some do's and don'ts of A/B testing. If you keep these simple and practical things in mind, you'll never make a mistake or fall victim to an A/B testing pitfall!


Only change one thing at a time

In order to determine how effective an element on your page will be, you have to isolate that variable in your test. This means you should only test one thing at a time. That means testing each element of the website - navigation, header, color scheme, etc. - separately.

Show repeat visitors the same variation

The tool you use to conduct your a/b test should have a way of remembering which variation a visitor has seen. This prevents any potential confusions such as different promotional offers or product prices.

Decide what is worth testing

Don't limit yourself to only testing the background color or text size. Take a deeper look at all the elements on your site and their possibilities for design, wording and layout.

Conduct many A/B tests

Your first A/B test may not yield any significant results, so try, try, try again! You may consider testing the same variable again but with different variations. An A/B test can have three outcomes: no result, a negative result or a positive result. Just remember that this testing process focuses on continuous improvement.

Elements worth testing:

  1. Any Call-to-Action button (wording, size, color)
  2. Headline or product description
  3. Form lengths and fields
  4. Layout and style of website/landing page
  5. Product pricing and promotional offers
  6. Images


Don't surprise regular visitors

Only include new visits in your test if you are testing a core part of your website. You don't want to shock your regular visitors.

Don't wait to test the variation until after you've tested the control

Always test both versions at the same time and always split traffic between the two versions.

Don't be biased towards one variation

Everyone has their preferences, and you'll most likely have a variation in your test that you'd like to win. Just ask yourself if you'll be able to accept the results of the test, even if your preference doesn't win. If not, don't bother testing.

Don't give up or draw conclusions too early

Don't read too much into the results if you only have a few conversions for each variation. Also, don't forget that this process takes time.

Elements NOT worth testing:

  • Minor text changes - changing the word "a" to "the" isn't going to make a difference
  • General improvements

Hopefully with these do's and don'ts in mind, you'll have everything you need for a successful A/B testing strategy. Just keep in mind not to draw conclusions too early and that this form of testing will take a bit of time!