sales planning

How to A/B Test Your Hypothesis

This is the last post in our series of five steps to landing page success. Let’s do a quick recap of what we’ve covered so far:

  1. Starting by defining your landing page goals and metrics
  2. Then creating a successful landing page (and what that looks like)
  3. Measuring the performance of your landing page
  4. Then creating a test hypothesis for landing page optimization.

Today, we’re finishing up the series with our final step –  how to A/B test your hypothesis.

A/B testing is another term for split testing. It allows you to have a control landing page (the one without any changes) and a testing landing page (the one set up to test your hypothesis). Having these two pages gives you a more accurate idea of what works and what doesn’t.

Many businesses use the same page to make changes. This can cause quite a bit of confusion in your results and create more work for you in the long run. With an A page and a B page, you can continue to make small tweaks to test the landing page’s success without tainting your starting landing page.

How to Test Your Landing Page Success

With the criteria outlined in your hypothesis from step four, you’re ready to make a change.

To test, create a second web page with exactly the same elements. Then, only make the change you defined in your hypothesis statement. Doing any other changes at this point will taint your results. You won’t know what changes worked and what changes failed.

Testing Criteria

Once you have both pages set up, it’s time to start driving traffic to each landing page. To have clean test results, you must have the same amount of traffic going to page A as you do going to page B.

Not only do you want to have the same amount of traffic going to each page, you also want to have a high enough number of conversions before trying to test other elements. Don’t try to make any other changes to your test landing page until you have at least 100 conversions on both pages.

It’s equally important that you let your test run for long enough. We recommend testing your landing page for a minimum of one week for each change you make.

Ready to Get Started?

Testing your landing page is a fun way to achieve more success in your campaigns. You can bring your entire team into the mix to gather ideas for changes to make, and then hold an office contest to see who was right about which changes work the best. Keep it playful and exciting while also brainstorming for how you can make your landing page perform at its maximum potential.

That’s it! Bookmark this series so you always have the step-by-step guide handy for your campaigns. Here’s to your success!

Image: Photospin