Use Google Analytics to Identify Ideal Pages for Google Website Optimiser Experiments

Thursday, August 13, 2009 | 5:56 PM

Labels: ,

We have seen how Google Website Optimiser is a capable testing tool for increasing landing page effectiveness. Using this tool you can test different page designs and content to find the best formula that works for your business. An effective landing page results in more conversions, which in turns boosts your return on investment (ROI). All this without needing to spend an extra Dollar (or Ringgit or Baht)!

A question I often get asked by online business owners is "which pages should we be testing?" This is not an easy question to answer, because the answer always varies depending on the nature of the business, the target audience, audience sophistication, market maturity, and so on.

I always encourage advertisers to consult their Google Analytics reports to help make this decision. As a business owner you may have an intuition about which pages need to be worked on, but it is often easy for your emotions to get in the way of identifying the right page. Data always beats opinion and that's why I always encourage using your web analytics reports to identify the right pages.

I rely on three Google Analytics reports to help me identify problem pages that need to be worked on: Top Landing Pages, Funnel Visualizations, and using the $ Index metric with Top Content.

Top Landing Pages
The Top Landing Pages report shows you which pages visitors are entering your site through. By switching to the comparison view and selecting "Bounce Rate" you can easily identify which pages have a higher bounce rate than others.

Bounces are single-page visits and means the visitor did not venture beyond the page that they landed on. This often signals that the user was dissatisfied with the content that they were exposed to. Perhaps the landing page did not match your ad's messaging or there is no clear call-to-action. These are pages that need to be addressed in order to encourage visitors to move beyond the landing page.

Pages with a red bar are pages that have a higher than average bounce rate. You should address these pages first.

Even though you should focus on the high bounce-rate pages, you shouldn't ignore the pages with lower than average bounce rates (signified by the green bars). You should inspect these pages and determine why they appeal to visitors. Does it have an eye-catching call-to-action? Does the page content align well with the ad's promise? Is the interface simple to navigate? Leverage these learnings and apply them to other pages in your site to reduce their bounce and exit rates.

Funnel Visualizations
Setting up goals and funnels in your Google Analytics account is vitally important for measuring how well your site is meeting its objectives and where in the conversion process your visitors are dropping out. Use the Funnel Visualization report to identify at which steps of the conversion process are most of your visitors dropping off.

These are the steps you should be addressing in your testing strategy to ensure that more visitors flow through the funnel and through to goal completion.

Ask yourself what can be done to encourage more visitors to move from the troublesome step to the next step. Does it involve including more helpful information to reduce confusion? Should you move your trust symbols (e.g. "Verified by Verisign") to a more prominent position? Or ask the extreme question: is this step even necessary?

Top Content report and the $Index metric
If you have set a value for your goals or are using e-commerce tracking, you should make use of the useful $ Index metric to determine which are your most "valuable" pages. In short, $ Index is a useful metric for determining the relative value of each of your pages. Pages with a higher $ Index value contribute more to your bottom-line than pages with low $ Index values. Visitors who visited these high-value pages went on to eventually convert and make high-value transactions.

Navigate to the Top Content report, and sort the report by the $ Index metric. You will note that there will be some pages that have an unusually high $ Index value. These are often pages that are in the conversion funnel (e.g. credit card confirmation page) and should be filtered out using the filter box at the bottom.

Next, switch to comparision view, and compare exit rates.

We have now identified pages that are of high value to you yet have higher-than-average exit rates. These are pages that you should also be focusing on. If you can reduce the number of visitors leaving the site from these pages it should help drive incremental revenue for your online business.

These are three key reports that I use to identify problematic pages that would be ideal Google Website Optimiser experiment candidates. These reports don't simply identify problematic pages, but rather, they highlight high-potential pages that, if optimised, should lead to large gains in ROI. The next step is to start testing these pages with Google Website Optimiser.

So how do you use Google Analytics data to identify pages for testing? We would love to hear from you in the comments below.