Microsoft touts eye-tracking study as proof of Google’s EU antitrust naughtiness

Google(S goog)’s EU antitrust adventure is getting a bit repetitive – the company proposed concessions and was told to try harder, then Google tried again and again. And now the companies whose complaints kicked this whole thing off, Microsoft(s msft) being the ringleader, have said they’re still not happy with what Google is proposing.

This time round, the Initiative for a Competitive Online Marketplace (ICOMP) – one of Microsoft’s astroturfing operations – has commissioned an eye-tracking study to demonstrate how Google’s latest suite of settlement proposals “actually makes the abuse worse.”

Watching where users look

The study was conducted by the media research department at the German Sports University in Cologne. According to ICOMP, it shows that, quote:

  • Google’s ‘Sponsored’ results consistently attract the largest amount of the users’ total visual attention
  • ‘Alternative search sites’ do not draw enough visual attention to prompt the users to click on them
  • Visual attention for organic links is negligible compared to the ‘image enhanced’ Google elements placed above them

All this relates to one of the four accusations against Google that the European Commission is investigating: that Google promotes its own services through search results in a way that’s not transparent, leaving users to believe that Google’s search results are neutral. The Commission was particularly concerned that alternative, industry-specific search services were seeing their own results downgraded in Google’s results, sometimes being pushed off the first page.

Now, this is a serious matter – Google has a much greater share of the European search market than it does of the U.S. market, so in my opinion the antitrust authorities are right to be pushing for change. However, at some point a compromise must be made.


Here are two examples of what the German sports university found when showing test subjects the type of search results that Google is proposing:

“Results for the search term ‘iPod’ reveal that thumbnail product pictures guide the visual attention of users to ‘Google Shopping Results’ with 56% of participants clicking into this area. While the ‘alternative search sites’ caught less visual attention and only clicked once, indicating little interest from users.

“During a search for ‘Map London’, the Google Maps area and Google Images thumbnails receive more, earlier, and longer visual attention than all other page elements, including competing mapping providers, with 46% of participants clicking on the Google map and a further 36% clicking ‘Google Images’, which is itself another Google service. Comparatively, the 1st organic link on the results page, ‘’ (a Google Maps clone), received only two clicks and the official ‘Transport for London’ link positioned below, received no clicks.”

Google iPod heatmap
Two points need to be made here. Firstly, never trust a study commissioned by an interested party. While I’m sure the German Sports University maintains the highest standards, you would hardly expect such a study to come out in Google’s favor.

Secondly, in a region where Google has 90 percent market share, most people have been conditioned to use the company’s services in a certain way. Google’s search results pages have always promoted Google services in the past, so users will automatically gravitate towards those results. The only way to avoid that would be to completely redesign Google’s results pages to the point where they are unrecognizable. It’s as though Microsoft and friends want Google to pretend it doesn’t have a connected suite of services.

In the iPod(s aapl) example pictured above, vertical search engines’ results are way up there. Yes, they’re below Google’s own results, but those are clearly marked, both with the words “Google Shopping” and with a “sponsored” tag.

This has dragged on long enough. I’m not saying Google’s latest concessions are perfect in all respects, but they’re certainly a big improvement in terms of visual placement, and commissioned eye-tracking studies such as this one just come across as petulant. Let’s all move on soon, shall we?