Does Google Lie? 9 Examples: Antitrust and SEO Misinformation

Cara Bowles    By under SEO.

Updated August 25, 2020.

When this post was first published in August of 2014, I couldn't find any evidence of outright lies from Google, at least not within the context of search engine optimization (SEO), but I concluded that there were several cases where Google had still been deceitful. A lot has changed since then, and it is now clear Google has outright lied to this industry and the larger public.

This post has been dramatically expanded and rewritten to account for revelations found in documents from the antitrust-related House Committee on Online Platforms and Market Power.

Make no mistake, the way some SEOs operate creates conditions that lend themselves to conspiracy theories, which tend to emerge whenever people are both highly individualistic and lack a sense of control. An SEO strategy that depends too heavily on the specifics of how search engines work, combined with the lack of transparency about exactly how Google works, is a recipe for this kind of counterproductive thinking.

Even so, smart SEO does keep Google's motives and incentives in mind as they devise a strategy.

Let's take a look at examples from Google's history of deceit.

1. Google May Have Bought YouTube Explicitly To Limit Competition

youtube tablet

When Google bought YouTube for $1.65 billion in October of 2006, Sergey Brin said it was because:

Google's mission is to organize the the worlds information...and videois an important part of the worlds information.

But could Google have instead bought YouTube to limit competition? If so, it could violate the Clayton Antitrust Act, which says:

No person engaged in commerce...shall acquire...the assets of another person engaged also in commerce...where...the effect of such acquisition may be substantially to lessen competition, or to tend to create a monopoly.

In February of 2006, less than eight months before the acquisition, Eric Schmidt said in an email that they wouldn't be interested in buying YouTube for the much lower price of $500 million:

[Eric Schmidt:] Please proceed to meet with youtube to discuss a serious offer from Google to acquire them...

[Sean Dempsey:] I spoke with youtube...the net is that they want something in the $500M range...

[Eric Schmidt:] Thanks...We won't be pursuing them as an acquisition.

In another series of emails, Google executives discuss buying YouTube to defend themselves against Yahoo despite believing it was an inferior product to Google Videos. This discussion was apparently the motivation for the initial offer:

have we talked to the YouTube guys about coming here?...

we haven't...last we heard they...were thinking about a acq with Yahoo...

Is there any interest in acquiring them?...it would be nice for y! [(Yahoo!)] not to have them...

...their systems wouldn't be valuable to us. they aren't doing anything on their site where i say "wow they have some big video brains there"...their content quality is worse...they seem focused on the home video/community space while we want to be more like itunes/TV...if we pick them up it would be defensive vs yahoo but there are...more sites yahoo could go out and buy...

I think we should talk to them, if nothing else to make it more expensive for Yahoo.

A few months later, in May, after the initial offer was made, the executives discuss how YouTube was becoming a serious competitor:

We did look into them at the time and talk to them. They didn't want to be acquired and talked about a 500M valuation...So what did Eric say today at emg about video...

Just that youtube kicked our buts...

I was surprised he just noticed...

While all of this was going on, another thread reveals that Google was also considering a flip video camera partnership with Pure Digital as part of an effort to own as much of the online video market as possible in yet another defensive move against Yahoo, but considerations about buying YouTube apparently made this deal less important to them:

Peter Chane: ...Pure Digital...makes the disposable and one time use video cameras...I ran into their CEO today and he said they are close to signing a deal with Yahoo or YouTube to host their videos. They prefer to work with Google Video...their users will generate about 2M videos a year (our index today is about 350k)...a deal like this seems like a powerful way to kickstart a Google Video community...

Jennifer Feikin: ...i think we should acquire youtube, but pass on the pure dig deal...to date, our mission has been to host all video...the group's overwhelming concern when we first talked to pure digital...was that we'd be innundated [sic] with a lot of home video...youtube has some some great community features and they are ahead of us...then, if they do a deal with pure dig, we'll get it anyway...

Salar Kamangar: ...fyi -- if you have any thoughts on this let me know, but i want to be aggressive about deals that make google the default place to store photos and videos.

It's in this context, where YouTube was being discussed as a competitor, as a road to owning the online video market, and as a way of weakening Yahoo! as a competitor, that Google seems to have made the decision to buy it for what was seen as an "outlandish" $1.65 billion. The discussion seems to be almost exclusively about suppressing competition, not acquiring systems that supported their existing business efforts.

For context, while Google finally for the first time released YouTube's revenue numbers for the first quarter this year, they still haven't released information on profits. At least as late as 2015, sources were still saying YouTube wasn't actually running a profit for Google.

Given that YouTube seems to have been bought in the context of discussions about suppressing competition, and that they ran it for at least nine years with no apparent interest in using it to make a profit, the purpose of the acquisition and the business itself since seems to be entirely to stifle competition.

2. Google Likely Bought Invite Media To Stifle Advertising Competition

advertising

When Google announced that they bought Invite Media in June of 2010, they claimed it was about helping publishers and the whole industry space:

We're happy to announce that we've acquired Invite Media, an innovative start-up based in New York and Philadelphia...

We've developed, and are continuing to develop, tools that help both publishers and advertisers take advantage of the opportunities that display advertising offers. These investments are really bearing fruit: publishers are getting improved returns from our suite of publisher solutions; and advertisers and agencies are running effective and creative display campaigns with Google across the web...

Together with other participants in the industry, we're working to help them both grow significantly in the years ahead.

Invite Media was a real-time bidding platform that allowed advertisers and agencies to buy display ads across multiple advertising exchanges. In other words, it allowed other advertisers to compete with Google Ads (AdX at the time) all through one interface.

Just four months earlier, Neal Mohan, the Vice President of Product Management at Google who wrote the press release about the acquisition, was speaking with Henrique De Castro, the VP in charge of Google's media, mobile, and platforms organization:

Henrique: ...AdX is currently unable to track agency spend if the agency in question uses a 3rd party bidder. All networks are moving towards 3rd party bidders globally and we will lose all visibility of agency spend on AdX once this happens...What's needed: ...Google builds its own bidder - which is on our roadmap but is unlikely to deliver before end 2010...

Neal: ...Yes, we should build a bidder...but this reporting issue is not the primary benefit of owning our own bidder...The primary benefits...are eliminating the disintermediation risk and substantially increasing display spend with with Google from agencies...

What Henrique is saying is that, at the time, everyone was moving away from using Google's own bidding platform towards bidding platforms that pitted various advertising platforms against each other. His concern is that this would rob them of access to agency data. We'll see later in this post that Google explicitly considered their access to user data an unfair advantage over competitors.

But, more importantly, what Neal is saying here is that the primary reason Google wanted its own bidder was explicitly to stamp out competition and intermediaries in the advertising space and take a larger share of the pie.

At some point in the four months after this exchange, they decided it wasn't worth it to continue building their own bidding platform, and instead bought one of the most successful bidders they were talking about: Invite Media.

3. Yes, Google Does Use Click Data And User Behavior Metrics

click mouse

Google has told the SEO industry many times that they do not use UX and user behavior metrics for rankings. In a Reddit AMA in 2019, Gary Illyes said:

Dwell time, CTR, whatever Fishkin's new theory is, those are generally made up crap.

In 2017, Barry Schwartz asked John Mueller:

To be 100% clear, Google is not using click data, behavioral data, pogosticking, that type of stuff directly in their search ranking algorithm for ranking websites in their search results?

And Mueller responded by saying:

As far as I know, we don't use that.

The denials go back at least as far as 2008 when Matt Cutts said:

I'll just say that bounce rates would be not only spammable but noisy. A search industry person recently sent me some questions about how bounce rate is done at Google and I was like "Dude, I have no idea about any things like bounce rate. Why don't you talk to this nice Google Analytics evangelist who knows about things like bounce rate?" I just don't even run into people talking about this in my day-to-day life.

But a Google document from 2007 indicates that they were using click data and other user metrics for core rankings:

We have many promising ranking initiatives underway for the coming year. Recent gains indicate there is significantly more possible on core ranking. Initiatives include:

...

  • Continued investment in user signals, like clicks. Our search users create the first level of network effect for search quality and we are investing in this heavily.

Some of the most important phrases here are "core ranking" and "network effect." Google has admitted in the past that they use click data to test the quality of their algorithms, but denied that they used it as a core ranking signal. But this document suggests they were using clicks to measure "network effects," not just A/B testing different search result pages against each other, and it specifically mentions that this was being done in the context of core rankings.

The document goes on to make the data usage and the desire for more of it even more explicit:

Observation: Google has more users than any other search engine...If we could find a way to improve search that scaled...with the size of the user base, we could harness unparalleled advantage.

Remedy/Initiative to address: The best way to find such an advantage is to invest in and explore areas where this kind of finding seems possible.

  • Use more pervasively the data available to us, including toolbar and Orkut data...

...

  • Increase the number of signed in users...experiment with a tiered login system...invest in products that will incent people to...sign in...we recommend the acquisition of facebook.com, which we believe will double the number of signed in users...

Also, as we've mentioned here before, ex-Google employees have confirmed that Google uses Chrome user data:

Most of all though, and perhaps one of the biggest points of the session was that Google definitely uses Chrome user data and can track every click within it.

It is now perfectly clear that Google has not only been using this data but that they were hungry enough for user data that they were willing to try and buy Facebook just for the data.

4. Yes, Google's Algorithms Are Influenced By Human Quality Ratings

human ratings feedback

As recently as August 4th, Google still claims that

Once raters have done this research, they then provide a quality rating for each page. It's important to note that this rating does not directly impact how this page or site ranks in Search. Nobody is deciding that any given source is "authoritative" or "trustworthy." In particular, pages are not assigned ratings as a way to determine how well to rank them.

...

Instead, ratings are a data point that, when taken in aggregate, helps us measure how well our systems are working to deliver great content that's aligned with how people--across the country and around the world--evaluate information.

To be as fair as possible, the point most strictly being made here is that human quality ratings don't directly impact rankings, and as far as we know, that's technically true. But the claim that its purpose is to "measure how well our systems are working" at the very least implies that ratings don't influence rankings at all. Arguably, that implication is even stronger for "Nobody is deciding that any given source is 'authoritative' or 'trustworthy.'"

But the same document that makes it clear Google uses click data and user behavior metrics also makes it clear that human quality ratings were used to train Google's algorithms:

Rankboost: continue developing our learning system to take human rating data as input and predict new ranking signals

In other words, Google is trained to rank pages that humans have given high-quality scores well. That's how machine learning works. If a person rated a page well and the algorithm doesn't, Google takes that as a bad sign and adapts its algorithm. So, while it's apparently technically true that "nobody is deciding that any given source is 'authoritative' or 'trustworthy'" to directly change the rankings of those sites, Google is designed specifically to promote pages that people have marked as authoritative and trustworthy.

More brazenly, however, Google has previously denied that they even use human quality ratings to train their algorithms:

@jenstar: AFAIK, Google only uses the raters and data to evaluate possible algo changes based on whether the raters rate the results better or worse with each algo change they test. Maybe @johnmu @methode or @dannysullivan can weigh in if they are used for ML now too.

@dannysullivan: We don't use it that way.

Unless the project was phased out for some reason, this appears to be an outright lie, whether Danny Sullivan was aware it was or not.

5. Google Explicitly Said Their User Base Was An Unfair Advantage

mind the gap

Google's official position in antitrust situations is that, despite dominating the general search market share, they are vulnerable to competitors in vertical search. For example, in a submission to the Australian Competition & Consumer Commission, they said:

Google faces fierce competition from other providers, including vertical search sites like Amazon, Expedia, Domain and Carsales.com, many of which users access directly through mobile apps.

But again, the strategy document cited above stated, very clearly, that their goal was to:

Turn having the largest user base into an unfair advantage by building out technology that improves linearly with user base size -- initiatives that have this property include: using toolbar data for personalized search, integrating recommendations, question/answering, and creator/trust rank.

In other words, Google believed that they could maintain dominance over vertical search competitors due to their larger and more diverse data sets, a direct result of their search market share.

6. Google Does Track Search Clicks To Their Own Properties

analytics

In December of 2018, the House asked questions for the record from Google's CEO Sundar Pichai. Among them, he was asked:

What percentage of those search engine result pages resulted in the user:

a. Terminating their search because they were satisfied by the answer?

b. Clicking on a link which led to the user a webpage not hosted on Google.com?

c. Clicking on a link which led to the user being sent to a secondary page on Google.com, any of Google's top-level domains, or generating another search engine result on Google?

These were written follow up questions, so Sundar's response should be viewed in light of the fact that he wasn't being asked these questions on the spot. His answer didn't address the percentages at all:

Google doesn't necessarily know why users terminate a search...[Users] may be dissatisfied by the answer...found what they needed...realized that a different search query is more likely to help them...We will continue to work hard to fulfill our mission.

While this answer technically avoids claiming they don't have this data, the claim that they don't "know why users terminate a search" serves as a misdirection that sort of implies they don't.

But an old document about Google Shopping's predecessor and the plan to roll out product search suggests they definitely track this kind of information. In discussing how best to integrate their shopping results with Google search, they mentioned these data sources:

  • % English google.com queries that are commerce-related...
  • distribution of commerce-related English query types on google.com...
  • % google.com queries that are commerce related in [other languages]...
  • proportion of google.com users with a (US?) location in their cookie (so we can calculate shipping)...
  • proportion of google.com users with a Checkout account
  • histogram of # pages seen for users with merchant clickthroughs
  • histogram of # pages seen for users without merchant clickthroughs...

They name a specific percentage of Google search traffic to shopping results as a proposed milestone:

Proposed milestones...

  • 1% google.com traffic to Product Search index

Obviously, it's impossible to set a certain percentage of Google.com's total traffic as a milestone without tracking it.

They also discuss building a system to optimize the inclusion of their shopping results:

Universal Search Support (1) system that figures out how commercial a query is, whether we should put a link to product search, or even list a few products. Can e.g. do this manually for x queries that cover y% of search traffic temporarily.

While this document is old and is only about one of Google's properties, it seems to be a very clear indication that they care a lot about and track what proportion of their search traffic stays on Google's properties.

7. Links from Press Release Sites

press

On December 25, 2012, Matt Cutts wrote:

Note: I wouldn't expect links from press release web sites to benefit your rankings, however.

If we look at the phrasing closely, it's clear that Cutts isn't flat out saying "links from press release sites won't help your rankings." Instead, he's clearly saying you shouldn't count on them helping your rankings. That's an important distinction, so it's not really an outright lie.

However, it's important to realize that links from press releases can help your rankings (well, more accurately, we know they can pass anchor text). SEOConsult ran a test that proved this. The case study isn't on their site anymore, but here it is in the Wayback Machine.

8. Hummingbird

hummingbird

On August 21 and 22 of 2013, all of the major keyword tools spotted massive changes in the search results. When Search Engine Roundtable contacted Google about the update, they had "nothing to say about this topic." Many SEOs took this to mean that Google was denying an update.

About a month later, Google announced that they had completely revamped their algorithm with the Hummingbird update and that the changes had been made about a month before.

Again, this is not a case of outright lying, but it's clear that the update on August 21/22 was Hummingbird, and Google declined to say anything about it for a month.

9. "Not Provided" Keyword Data

closed

When Google started blocking keyword data in search results, Google claimed that it was to protect the privacy of searchers who were signed in to Google. They also specifically said that "the change will affect only a minority of your traffic." However, there is no reason why SSL search would prevent the reporting of keyword data (which is still found in AdWords), and today, essentially 100 percent of all keyword data is now "not provided."

The way the web is supposed to work, Google wouldn't pass the keyword data forward unless it were to a secure site. By default, this would have meant that webmasters could still access the keyword data if they just switched their sites to be encrypted. This would have encouraged webmasters to encrypt their sites, which in turn would have made the entire web a more secure place.

Instead of doing this, Google deliberately built a unique system that never refers to the keyword data, even to secured sites, except to Advertisers, even if they were unsecure. In short, they deliberately broke a web standard to avoid a backlash from unsecured advertisers.

Google began moving to encrypt all search, even for users that aren't signed in, in the wake of the PRISM scandal. Again, this data was not blocked from advertisers, even if they are not encrypted.

While the word "lie" might not quite fit, it's clear that the changes were not made in order to protect user privacy, or they would have blocked the data from advertisers as well. This move was a clear case of deception.

So, Does Google Lie?

pinocchio lie

When this post first went live in 2014, my conclusion was that I hadn't found any evidence that Google had outright lied about anything, although the spin they put on encrypted search came very close.

I have to say, in light of the documents now available to the public, it seems clear Google has lied. I don't know that they've done so in a way that's legally prosecutable, but on click data, on human quality ratings, to say that they lied seems like the only reasonable interpretation. Regarding the others, deception at minimum is certainly at play.

For us as SEOs, debating over the technicalities of what constitutes a lie isn't especially valuable, however. I think it's more important to realize that Google has very strong incentives:

  • Not to give away anything about the algorithm that people can exploit.
  • To discourage people from violating their guidelines.

Rather than wrap this up with my own thoughts, I will instead leave you with a quote from the definitive paper by Larry Page and Sergey Brin on PageRank, back when Google was a Stanford research project:

Furthermore, advertising income often provides an incentive to provide poor quality search results. For example, we noticed a major search engine would not return a large airline's homepage when the airline's name was given as a query. It so happened that the airline had placed an expensive ad, linked to the query that was its name. A better search engine would not have required this ad, and possibly resulted in the loss of the revenue from the airline to the search engine. In general, it could be argued from the consumer point of view that the better the search engine is, the fewer advertisements will be needed for the consumer to find what they want. This of course erodes the advertising supported business model of the existing search engines. However, there will always be money from advertisers who want a customer to switch products, or have something that is genuinely new. But we believe the issue of advertising causes enough mixed incentives that it is crucial to have a competitive search engine that is transparent and in the academic realm.

Want more? Check our Google Ranking Factors fact-check for dozens of other examples of statements by Google employees that are contradicted by other evidence when it comes to how Google ranks content.