Performing AB testing on the landing page

Contributors

@paul-boag


Business Benefits

Optimize your landing pages for conversion.


Use a tool like Clickfunnels or Kartra with built-in A/B split test features to make a new variation of your landing page.

Duplicate your original page and use that as a starting point. If the page builder you are using does not have this feature, search for software that does, like Zoho PageSense, Leadformly or Convertize. Alternatively, perform A/B split testing with Google Analytics for free.

Decide what you want to test on your landing page.

Only make one change at a time so you can easily determine what changes are making improvements. Testing every possible aspect on your landing page will take a lot of resources and time, so instead pick those that have the highest conversion impact. For example, page layout, call to action, button colors, button position.

Implement the first change on your page variation.

For example, if you want to test a one-step checkout page vs a two-step checkout page, implement one-step checkout on variation A and a two-step checkout on variation B.

Set the distribution of landing page traffic to be even between variations - 50% and 50% - and turn the A/B testing feature on.

A/B testing software lets you specify the percentage of traffic going to variation A and B.

Run traffic to your landing page URL. The split testing software will automatically distribute traffic between landing page variations.

The page variations will probably not get the same number of impressions at the start, but these numbers should level out with more traffic to your website.

Check your software to see which landing page variation gets the higher conversions. The one with a higher conversion rate is a winner.

Your software should tell you when you have enough results to declare a statistically accurate winner. However, you can make the decision yourself if you are seeing a distinct trend emerging earlier.

Declare the winning landing page variation in the split test software.

The software will send 100% of the traffic to the winning landing page variation

Continue split testing different variations of your landing pages.

Using your previous winner as a baseline, start exploring other changes. You can either test completely different page elements or explore other variations of the previous element you changed.

Also, consider running A/B testing on other pages in the same funnel as the landing page. For example, test the page following a user completing a call to action on your landing page.

Working on small improvements in conversion has a multiplicative effect on your overall conversion. Therefore you should always be running split testing.