Optimising Google Sitelinks with Analytics

The Google Sitelinks were a welcome addition to the SERPs. They initially only appeared for brand searches and it gave webmasters a nice increase in their share of the space above the fold.

The Sitelinks are generated algorithmically and the results are usually a mixture of primary pages with a few bizarre additions.

Looking at the Sitelinks for Amazon.co.uk’s homepage as an example, Google has selected a majority of primary links but also one specific product, Skechers Womens Shape Ups Trainer.

amazon-sitelinks

In most cases, webmasters simply look at the selection and make decisions about which ones they think are irrelevant. In the Amazon example, the Skechers link stands out but the remaining selection of primary links may not be optimal.

There is a more systematic way to optimise your selection of  Sitelinks using data from your analytics package. By analysing the number of times each Sitelink is clicked and the resulting bounce rates and even conversions, you can experiment with alternatives to narrow down to the selection which provides the greatest overall value to your users.

All you need is to be able to view landing page data for specific keyword searches. This feature is available in Google Analytics which I will use for this example but this can replicated in your favourite analytics package.

  1. In Excel, put together a list of your Sitelinks URLs which you can get straight from Google’s SERPs or from within Webmaster Tools.
  2. In Google Asnalytics, navigate to the Keywords report under Traffic Sources.
  3. Find your brand search term, usually at the top of the list, and click on it to drill down to the specific data for this search term.
  4. Now select ‘Landing Pages’ fom the Dsimension drop down list.

    google-anaytics-landingpagedimension

  5. You will now be looking at the landing pages for your main brand search term.
  6. Copy the number of entry visits, bounce rate and conversion data for each landing page that is one of your Sitelinks into your Excel workbook. Here is an example of how the data might look for Amazon’s Sitelinks.

    amazon-sitelinks-table1

  7. You can now calculate the click-through-rate of each link by dividing the number of visits by the total number of visits.
  8. To understand the true value, you need to factor in the bounce rate or conversion rate as you might find that a link generates a lot of clicks but isn’t a good landing page. To do this, simply multiply the number of entry visits for each sitelink by the bounce rate or conversion rate. In the example below, I’ve used the coversion rate to establish the value.

    amazon-sitelinks-table2

  9. You can now easily see which Sitelinks work and which ones don’t. In this example, you would obviously remove the Skechers link but also the Music and PC & Video Games links. Without the analysis of the data, this would not have been obvious.
  10. If you discover that one of the links you previously blocked was a good performer then you can unblock it in webmaster tools and hopefully Google will reuse it.

A note of caution, if you have also have a double listing in the SERPs for your brand term, it’s likely that this second listing will generate entries which will corrupt your data. Usually the choice for your second listing is one of your main pages and probably won’t be one you want to remove from your Sitelinks so it isn’t necessarily a huge problem.

Google Analytics segmentation isn’t particularly accurate although it’s good enough for this kind of optimisation work providing you have a site with lots of data. If you want more accurate data then you can set up a custom profile to track keywords and landing pages. We’ll cover this in another post in the future.

Posted on August 9, 2009 at 8:55 pm by chris · Permalink
In: Analytics

Leave a Reply