Supplement: Shifting Search Scenarios, Extreme Bias Note: This is a supplement to the complaint: "The Case Against Google." It is intended to enhance understanding of the case. Shifting Search Scenarios There has been a significant shift in search scenarios that relates to the understanding of this case. Consider three search situations: 1. Google returns the URL of a Google AdWords customer or a Google property. 2. Google returns the URL of a publisher with an ad for an AdWords customer on the publisher's page. (The publisher is a Google AdSense "partner.") 3. Google returns the URL for a page with no AdSense ad and the owner of the page is not an AdWords customer or a Google property. Here are the economic implications of these three scenarios. Case #1 In Case #1, if Google places the URL on the first page of returned results and close to the top, it is very likely to get clicked on immediately. This means that Google will get paid 100 % of the price paid by the AdWords customer for the click, say 5 to 7 USD for a restaurant or a hotel ad. (Google will not get just 32% of the price if the click were on an ad placed on a bone fide publisher's page, as described in Case #2 below.) In short, Google will make a whole lot of money very fast. Case #2 Now consider Case #2, where an ad is placed on a publisher's page. First, the ad will be clicked on only about 2 to 3 times out of 1000 returns of the page as a search result. That is because users don't trust ads for information. No surprise there! So the page will have to be presented many times before a user clicks. But then when a user does click on the page, Google will get only 32% of the money paid by the AdWords customer for the click (68% goes to the publisher). So what one sees is very few clicks and far less money per click for Google. Case #1 is going to make lots of money fast for Google; Case #2, which honors the honest, old-fashioned advertising model, will make very little. Case #3 In Case #3, Google of course makes no money at all. Thus Google has no incentive to return such a URL, no matter how good it is. Also, returning such a URL will get in the way of money-making URLs, reducing Google's revenue. The higher up the URL in the search results, the worse the situtation is for Google's revenues. And note that in the honest, old-fashioned advertising model of Case #2, such a URL also gets in the way of high-revenue-generating URLs of Case #1. Thus it is much to the advantage of Google to push Case #2 and Case #3 URLs as far out of the way as possible. That is in fact what has been going on for some time now. Note that Case #1 URLs are rarely marked as ads anymore; hence they are biased (paid-for) search results that are also deceptive; Case #2 results may be unbiased (honest) search results if they are not artificially promoted over better results. Case #3 results are of course honest or unbiased results. But note that there is strong motivation to promote Case #1 URLs, as they are far more lucrative for Google. And that is the pattern that is seen these days: Top search results now consist almost exclusively of Case #1 URLs regardless of their merit. (Case #1 URLs for Google properties are a slightly different story. For a Google-owned property like Zagat, what you see is heavy promotion of page sponsorship on the returned page. In other words, if you want to get a high-page ranking on the Google property, you also have to pay. But keep in mind that until Google owns everything, the large majority of Case #1 URLs are AdWords customers.) The next section discusses just how biased Google's search results are. The relevance to this complaint is obvious: CoastNews represented a high-rated, unbiased Case #2 URL that got in the way of Google's high-revenue-generating scenario. When analyzed and understood in this way, it is clear that Google's URL placement scenario is really a deceptive scheme that harms both consumers and competitors. Extreme Bias As mentioned above, Google makes huge amounts of money by returning search results that are biased in favor of its AdWords customers or own properties. The question then arises: What percent of search results returned are biased (paid for) versus honest (unbiased) ones? Note that when Google began, most, if not all, returned results were honest; but clearly that has changed significantly over time. In a Business Insider article called "Here's The Evidence That Google's Search Results Are Horribly Biased," (http://www.businessinsider.com/evidence-that-google-search-results-are-biased-2014-10#ixzz3GSHJjHXc) Jim Edwards describes a piece of software (Chrome extension) developed by yelp and others that shows just how biased Google searches are. The title of the article says it all: "horribly." The software removes preference for Google+ results from Google searches, showing what the honest (unbiased) results would be. In one case with the extension enabled, a search on "hotels Balboa Spain" returned 2081 reviews. With it disabled (normal biased Google search), only 137 results are returned. I.e., Google returns only 6.58 percent of the results that it might in a fair search. (See Focus On The User at http://focusontheuser.eu/#introduction.) Let us now be specific to the complaint of CoastNews against Google (S. Louis Martin v Google Inc.). For a search on "San Francisco restaurant guide North Beach", the top 10 search results returned by Google are all AdWords customers or Google properties, other than one—yelp. In an FTC settlement with Google in 2013, Google agreed to be fairer to competitors such as yelp. It appears in the search described above, yelp is Google's token example of fairness. And among the returned results, one is not a guide at all but a restaurant. Moreover, that restaurant is rated as *** (mediocre) on the second listing, which is menupages.com. How much sense does that make? How much "editorial judgment" does that show? CoastNews's restaurant page is of course missing, though it is the only page that has photographs of all restaurants; lengthy descriptions, often including interviews with owners; and no reviews written on hand-held devices by semi-literate Millennials inebriated on a date, stating something like "Hey, dude, this place sucks!" or "Hey, dude, this place is f***ing awesome." On Media Post, see "Tool Claims To Show Google Search Bias": http://www.mediapost.com/publications/article/235369/yelp-demonstrates-googles-search-bias.html On the Wall Street Journal, see "Ads Tied to Web Searches Criticized as Deceptive": http://online.wsj.com/articles/ads-tied-to-web-searches-criticized-as-deceptive-1413226602 We rest our case that there is a big problem in the world of Google search caused by "bias," with the result that consumers are often badly misled and competition is either crippled or completely eliminated.
	By Dr. S. Louis Martin