Skip to Content

Site Extensions. The Good, The Bad and The Ugly.

The author

James Holding

Head of Analytics

The Adwords “Ad sitelinks” feature, allowing advertisers the possibility of additional links to specified site areas, is a feature we have used on several accounts to varying effect. Initial results suggest that adverts displayed with site links have benefitted from improved click through rates. In some select cases, conversion rates have changed for both the good and the bad.

(an example of sitelinks on a British Airways advert)

The Good In many areas, the conversion rates have remained at similar levels, however click through rate has improved, thus benefitting the overall campaign as a whole. Also logically if a user sees a link more related than their initial choice of search term (by this I mean something they were looking not quite the same as the search term), then the site links are likely to have a positive effect on conversion rates. The Bad Sitelinks are set at a campaign level, therefore are not always as ideally suited to the search term, as the link provided by the original advert (assuming you’re doing a good job with your adverts). This has the possibility to alter conversion rates for worse, which is a rather unwanted side effect. The Ugly The only reasonable way to assess the performance of sitelinks is, of course, to monitor the effect that they have. The standard sitelinks report in adwords does not offer conversion information; however the URL’s can be specified, allowing for the addition of variables for analytics packages to pick up. Tracking the site links effectiveness will be vital in selecting the destination links, wording of the links and especially the decision to keep sitelinks and improve them for the benefit of the overall campaign. Whilst the ideal method would be to have the information within adwords (hopefully one for the future), the information can still be recorded. Ultimately site links provide a useful tool; however as with much of online advertising, performance should be based on results obtained and not just blind faith.