Google’s Search Rankings Exposed as Subjective

February 26, 2011

Media and antitrust scrutiny of Google’s public representations — that its search rankings are unbiased and objective — will increase greatly for two reasons:

  • First, Google’s search dominance does make Google the Internet’s content “kingmaker” and it does empower Google to de facto pick winners and losers on the Internet just by subjectively adjusting its ranking algorithm; and
  • Second, the more people look into Google’s search ranking algorithm, the more obvious it becomes that it is highly subjective, and not objective and unbiased as Google represents.

As the glare of public and investigative scrutiny focuses on Google’s “black box” search ranking algorithm, its mystery and mystique fade away, because people come to understand that a search algorithm is just mass automation of the application of subjective variables/biases, subjective judgements of “quality” content and “quality links,” subjective judgements of intent, and subjective human ratings of websites and content.

  • Simply, automation of subjectivity does not produce objectivity.
  • When a monopoly search engine automates its subjective bias, the system reflects the subjective discriminatory bias of its programmers.
  • Google’s self-assertion and self-characterization that its search processes are completely objective, only affect the perception of their objectivity not the reality of the inherent bias in their search algorithm.

Google is unabashed that it was subjectively demoting “low quality” content in its latest changes to its search algorithm that affected 12% of all searches.

  • As the eBay phenomenon conclusively proves, one person’s low quality junk is another person’s high quality treasure.
  • Quality is inherently subjective because users value things differently.
    • A big problem with Google’s subjective algorithmic changes here is that they uniformly demote the rankings of low quality content regardless of whether a specific searcher values that type of content or not.

Pejorative Subjectivity: Google pejoratively characterizes content it wants to demote in rankings as “spam,” “low quality,” or “cheaters,” and that assessment depends largely on Google’s subjective judgement. It also has the the effect of discouraging an objective assessment of the content in question.

  • In a competitive market concerns of subjectivity would not be a problem because content providers could just take their content elsewhere, but that self-healing capacity of competition and choice does not work in search because of Googleopoly.

Why questions of Google’s objectivity will continue is because Google as so much motive and opportunity to subjectively rank information to benefit Google or Google’s world view.

We’ve learned from the Google-Overstock episode that .edu links are given super rank status over commercial links. Agree with it or not, Google has subjectively decided that not-for-profit links to information should always have more “authority” and weight than for-profit brand links. (This bias for free content over paid content also massively favors Google’s free content advertising model and disfavors Google competitors’ paid subscription business models.)

  • The largest subjective bias in Google’s search engine is that it is based primarily on the prevalence of web-links, not on objective external assessments of authority like audited subscriber audience size numbers or objective surveys of brand recognition.
    • A search engine based on links has a deep embedded subjective bias for academic content, and for tech-sector/blog generated content because that community naturally links relatively much more than other communities on the web, especially more than paid content communities who don’t depend heavily on links to increase ad revenue.

Subjective Self-Dealing: We’ve learned from the various antitrust suits and outstanding detailed quantitative research by Ben Edelman, Gary Reback, and Foundem that Google:

  • Subjectively ranks themselves above everyone else with a manual hardcoding to ensure selected Google products and services always rank first on Google’s search; and
  • Subjectively blacklists potential competitors to Google search.

Subjectivity of Human Raters: What we don’t know is what the subjective rating guidance is that Google’s army of “human raters are given by Google’s leadership to subjectively determine what “quality” is.

  • Without any transparent objective framework or standard that we know of, quality is in the eye of the beholder — the Google human rater could have their own biases and subjective views about what content people should see first and most, and what content Google wants sent to the “back of the arena.”

Adding to Google’s problem of impairing or degrading certain content’s access to Google’s monopoly world-wide Internet audience is that Google is well-known as the corporate champion of net neutrality — i.e. that broadband providers would economically-discriminate against particular content unless preemptively regulated by the FCC to prevent it.

The Googleopoly itself argued that competitive broadband providers should not be able to discriminate against content, or prioritize content based on their own ownership or financial interest in the content. Google argued that only users should be able to prioritize content.

  • At a minimum, shouldn’t Google have to openly disclose that they have a financial interest in Google-affiliated content that ranks first or high in Google’s rankings?
  • Without disclosure, oversight or third-party accountability, how is the user to know whether or not Google is ranking certain information based on how much money Google makes from the information or not?
  • If this neutrality was such a serious principled concern for Google in the competitive broadband provider context, why can’t Google see the analogous problem with its own monopoly search engine?

Let me be clear, as I have long said, my problem with Google is not that in discriminates against content in its ranking process, my problem is that it has consistently and blatantly misrepresented to the public that their search results were objective and unbiased when they knew they were not objective or unbiased — all to build up trust of an unsuspecting public.

In sum, we are witnessing the Google analog of the famous scene in the Wizard of Oz when Dorothy pulled back the big green curtain and exposed that the all-powerful Wizard of Oz was just using sound and visual PR effects to appear to be an all-powerful wizard — when he really was just like everyone else.

  • The point here is Google is not this perfectly moral, no-self-interest, super-human innovative search wizard, Google has all the normal biases, self-interests, and flaws that every other person or company can have.

Once again, Google’s real serious problem is that they have publicly represented themselves for years as completely free of bias or self interest in order to get everyone to trust them, when the overwhelming evidence is piling up that Google indeed has powerful biases and self-interest to favor Google-owned/favored content, and to disfavor competitive content or content Google does not agree with.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: