Find and Banish Bad Landing Pages with Google Analytics

Laptop screen with two red circles

We’ve gone on and on about the benefits of Google Analytics in this chapter, but it is in this section of the chapter that you can really see its practical application.

That’s because Analytics cannot just be used for tracking and analysing conversion data, but it can also be utilised to test the effectiveness of your landing pages. Why is it important? Well, after you’ve spent all that time, money and effort on paying for your ads, you’ll want to make sure your site is working as you’ve intended it to and that your visitors are converting just fine.

In other words, once people get to your site – all thanks to your super-awesome ad writing skills – you want to ensure that they find the information and/or the products they’re looking for.

No matter what your goals for conversions are – be it purchasing a pair of running shoes or signing up for a monthly newsletter – Google Analytics testing can help you optimise your landing pages to the max and help your site’s performance. And it’s got a fancy name too – Google Content Experiments. Yet, despite the elaborate title, it’s much like a middle school science experiment testing a simple hypothesis – which of the landing pages works better and why?

Testing Your Landing Pages

Without getting too deep into statistical analysis and technical jargon, Google Content Experiments work on the principle of seeing how different versions of a landing page change the actions that your visitors make. So, if one landing page gets people to convert and the other one makes them look around, but decide to ‘bounce’ to another page, then you’ll want to optimise the bad landing page and delete the elements that might not be working.

But don’t worry – not all of your traffic will be made into testing bunnies. You can control pretty much everything about the Google Content Experiments, making it not only a highly effective, but a flexible analytics tool as well. Namely, you can:

  • Use a random sample of the total number of your visitors, giving your statistically reliable data;
  • Select which goal or goals you want to test;
  • Indicate how much of your traffic should participate in the experiment.

Start Experimenting

Before you can start, you must make sure that your AdWords and your Analytics accounts are linked and that you have goals set up and being tracked on AdWords. Then decide which landing page and which aspects of it you’d like to test.

But how? Well, you can simply use the Google Analytics data that you’ve now integrated into your AdWords to make that decision. Check out the bounce rate and the average time on site, plus the conversion rate to pin-point the landing pages that aren’t engaging the users, converting them or both.

For example, you could have a call-to-action button that doesn’t stand out so visitors can’t figure out how to convert or imagery that just isn’t trustworthy enough, making visitors bounce. Other things you can test include, but aren’t limited to, copy, its placement, length, the presence (or absence) of images and graphics, length of fill-in forms, et cetera.

Once you’ve decided which aspects you would like to test, set up an experiment like so:

  1. Select Campaigns on AdWords.
  2. Choose the campaign that you wish to experiment with.
  3. Click on the Settings tab.
  4. Under Advanced Settings near the bottom of the page, click Experiment.
  5. Click Specify Experiment Settings
  6. Choose a name, which proportion of your traffic will be tested and the time-frame of the experiment.
  7. Click Next Step and enter the URL of the variation page(s) that you’ve set up.
  8. Manually or automatically paste the Google Experiments code into the original landing page of the test.

A Short Example of Using Content Experiments To Increase Conversions

Don’t worry – we won’t be talking about antique fishing lures anymore. Let’s say you’ve got a business fixing computers, offering basic maintenance, software installation and updates and solving complex hardware issues. If you make most of your money from signing people up for basic computer maintenance to tend to their computers twice a year, then getting more of such computer care contracts signed would boost your business.

People land on your site via an AdWords ad for ‘computer maintenance plan’, but which takes them to your basic home page outlining all of the services you offer. Since you’re not exactly targeting the biggest chunk of your potential business – the people interested in an on-going computer care plan – you should set up different landing pages that’d outline the basic maintenance services that you offer in slightly different ways.

Experiment with font sizes, colours, length of text, different imagery and anything else that would make sense to you.

Then, you should run a Google Content Experiment with your newly set up landing pages to see which one of them works the best, i.e. which makes people buy most of your basic care computer plans. When your experiment is running, a selected fraction of visitors will randomly be directed to differing landing pages. After you figure out which one of those landing pages works the best for you, you can go ahead and make that page live so that all of your site’s visitors see it and have the biggest chance of converting to boost your business.