Europe last week officially filed formal charges against Google for violating European antitrust laws. Specifically, regulators say that when European users search for products, Google favors its own Google Shopping service over competing sites and services.
America’s own Federal Trade Commission (FTC) made similar accusations before concluding nebulously that Google did favor its own services, but consumer harm could not be proved.
I’ve been interviewing search experts and journalists on the question as well as talking to Internet users. Responses to these charges fall into three general groupings:
1. The charges are false. Google results objectively rank shopping services according to their relevance.
2. The charges are true. Google favors its own shopping services.
3. It doesn’t matter if the charges are true or false. Google should be able to promote its own services higher than competitors because it’s Google’s search engine.
The most interesting version of response 2 came from my recent interview with Amir Efrati, senior reporter with The Information online news site. His opinion is that search engine rankings are all about relevance. Relevance is inherently subjective—an opinion. And that it’s Google’s opinion about relevance that is Google’s product—it’s the benefit you get from using Google Search.
It’s an interesting point, especially in light of the fact that (as Efrati wrote in a recent column) the European case against Google is about antitrust, but unlike most antitrust cases it’s not about harm to the consumers. As the FTC concluded, no evidence can be found that users are being harmed by Google’s rankings.
Europe’s five-year investigation into Google’s “fairness” in placing search results in the order that it does concluded with Europe officially disagreeing with Google’s opinion about which links are most relevant given a specific set of searches.
Once you deeply explore Europe’s case against Google—or the FTC’s case, for that matter—an elephant always shows up in the living room, which few acknowledge: The issue isn’t about results that are fair or unfair, relevant or irrelevant, biased or unbiased.
The issue is: Who should decide what the results are when somebody searches Google?
Who decides now?
The average person on the street might believe that Google currently determines the rankings that its search engine produces. But Google actually is the third most influential actor here.
Most results are determined by the actions of Website owners and by Google Search users.
For example, when I enter the term “cupcake” into Google Search, the second result is the Wikipedia entry for “cupcake.” On Bing, it’s the first result.
The Wikipedia entry ranks high because any good search engine would develop an algorithm for a highly relevant, universally applicable resource like Wikipedia that is also visited by a huge number of users worldwide.
The high ranking of Wikipedia is determined almost entirely by Wikipedia and by users. The search algorithms simply facilitate the influence of Wikipedia and users on rankings.
Personalization is another way for users to influence results. The No. 1 Google Search result for the word “cupcake” for me is a local chain called Kara’s Cupcakes. Because I live in Petaluma, Calif., and have demonstrated through my activity on Google Search and other sites my interest in organic foods, Google is showing me this particular result as the top one.
The person most responsible for that result is me.
Most of Google’s algorithm kung fu exists to remove Google from direct involvement in determining what the search results are—most, but not all.
Successful sites that rank or filter things algorithmically from Google to Facebook to dating sites must and do tests obsessively using real users. Google, of course, does such testing and tweaks its algorithms, as the company claims, to provide the best user experience.
Perceived relevance is one aspect of user experience. But there are others.
Shopping results may go up or down the rankings depending on factors other than content or even popularity. The range of information provided or even the formatting of that information may result in what Google deems a bad user experience, so they might be “penalized” with lower ranking.
I’ll pause here to remind you of what you might already know: Google’s relentless improvement of the user experience is the only reason why Europeans voluntarily choose Google for more than 90 percent of their search queries.
Who Should Decide the Relevance of Search Engine Results?
In a leaked report from the FTC investigation (similar to the European Union one), it emerged that when Google favored itself in tests over competitors, users didn’t find the results as relevant. So through trial and error, Google found a way to favor its own sites without users being too unhappy about it.
Google is likely to approve of its own sites, its own content, its own formatting and the accessibility of its own data. So Google is unlikely to get penalized for these usability problems. Plus, Google may also favor its own sites because it wants to boost its own assets and revenue.
It’s clear to me that even if Google were choosing to favor its own sites, the answer to the question “who decides search results?” is mostly users and secondarily the sites. Google’s discretionary meddling in search engine results is severely constrained by the need to succeed and barely registers as a factor compared with the overwhelming influence of users and Website creators.
Yes, results on different search engines vary. And that variation can only be attributed to what you could call “values,” which is defined as “the regard that something is held to deserve; the importance, worth, or usefulness of something.”
It’s all about values. Websites are built based on the values of the companies, their products or people who create them. User activity, which impacts both search engine rankings and personalized results, is based entirely upon how users value them.
It’s not possible to rank search engine results—or anything, for that matter—without those rankings being based entirely on values. Objectivity in search engine results is impossible.
So what Europe is actually proposing if and when it proposes material changes in Google’s search results is the introduction of new sets of values—the values of new groups—being factored into the mix.
They want the values of Google’s competitors to override to some degree Google’s own values in a product where Google’s collective set of value judgments actually is the product.
They want the values of bureaucrats, regulators, lobbyists, lawyers and others to override to some degree the values of users who determine search results in the course of their search activity and choices.
French politicians want to inject other values.
The upper house of the French parliament last week passed a bill that would force Google to provide competing search engines on the Google.com home page, and at least one must be a French company, which would probably be a search engine called Voila—a lousy search engine that French people mostly ignore. French lawmakers will vote on this bill in the weeks ahead.
This outcome would force the values of protectionist French nationalists as well as the values of some Google competitors onto Google Search.
If this sounds like crazy talk, remember that Europe has already enacted the right-to-be-forgotten rules, which enables a European citizen to “scrub” Google search results that link to Websites containing information that stigmatize the person.
Placing the right for the individual to be forgotten over the right of the public to remember and over the purpose of search engines to accurately reflect what’s on the Internet is a value that has been imposed on Google and other search engines and a European value that the EU wants Google to impose on the world.
Who should decide?
Regulators say they want search engine results to be “fair” and “objective.” But this is an impossible delusion.
If France succeeds in forcing Google to offer the Voila search engine on its home page, is that fair to France’s second and third biggest search engines?
On what criteria is one site objectively deemed “better” than another?
At the risk of overstating it, regulators are either lying or delusional when they call for “fair” and “objective” search engine results.
The truth is that they are calling for the coercive introduction of their own values—and the values of Google’s competitors—to take precedence over the values of users, of every Website and of Google itself.
How is this better?