This is a quick guide to help you test your Pay-per-click landing pages. Google has made this easier than ever by allowing you to run experiments through Google analytics. Of course the real task is to make sure you are testing the right elements!

Step 1: Identify landing pages to test

Most campaigns should have multiple landing pages that correspond to your various Ad groups. So the first step is to choose which landing pages should be tested to optimize your conversions. Here is how I did it:

  • First export your destination page report for analysis:
    • In your campaign go to your Dimensions tab and change the “view” tab to: “destination URL”
    • Make sure your column include: Clicks, Impressions, Cost, Conversions
    • Export the report into excel
  • In excel separate sort by landing page URL
  • For each grouping of landing pages calculate the total clicks, impressions and conversions
  • Now for each landing page group divide conversions by clicks to get your conversion rate
  • Now you should be able to identify what landing pages have the highest clicks, impressions and conversion rate
  • Choose a landing page based on your requirements (choosing a landing page with the high clicks and low conversion rate is typically the best way to go).

Step 2: Identify landing page elements to test

Once you have identified what landing page to test you can make a change to the landing page to test. Some examples include:

  • Main landing page image (product/testimonial/picture of a person)
  • Call to action (button size, button color, button position, button phrasing)
  • Copywriting (features vs. benefits)
  • Trust symbols (including client logo’s, accreditation or secure ssl certificates)

Step 3: Test your landing page in Google Analytics

Go into your Google analytics account and select the content tab. Within the content tab you will see a sub-category labelled experiments. Select this tab and create an experiment:

  • Enter in the URL of the page you want to test
  • Make sure you use a descriptive experiment name (help to differentiate if you have many experiments running)
  • Now add your test landing page URLs. You can add multiple variations but keep in mind that the more variation the longer the experiment needs to run before its statistical significance.
  • For the next step select your experiment goal (this is commonly your “conversion” goal)
  • Choose the appropriate % of traffic you would like diverted to your test pages. If you are risk averse then consider sending less traffic to your new landing pages (example: 10%). Keep in mind that the less traffic you send to your experimental landing page will extend the length of your study because it will take longer to get statistically significant results.
  • The last step is to add the experiment code to your original landing page (the code should be placed right after the opening head tag).

Now sit back and enjoy the fruits of your labour: higher converting landing pages!

Have a faster or more efficient way to test landing pages? Would love to get your input in the comments.

Nick Johansson is an SEO & Internet Marketing expert at Pixelmade, a Vancouver web design + internet marketing firm.You can find him on Twitter and Google+.