Pandu Nayak: Google’s Head of Search Quality and Dry Snitching

Pandurang P. Nayak is a vice president at Google in charge of search quality. In this article we will talk about his personal life, professional career, and how recent statement he made incriminating his boss led to the creation of this website.

Personal Life

Nayak is a 57 years old resident of California that grew up in India. According to his website, he received a bachelor or technology in computer science and engineering from IIT Bombay. He moved to the United States sometime in the 1980s before attending Stanford University and receiving a Ph.D. in computer science. He appears to have been married at some point and might still be.

Where He Lives

Public records list several residences in recent years. The most recent are in Palo Alto and San Francisco, California. Since we are not sure which of these residences he spends most of his time we are including both of them.

Palo Alto

Nayak purchased property at 3485 Middlefield Road, Palo Alto, California 94306 with a woman in 2016 for $4,600,000. We know her name, but are not posting it because our beef is with Pandu and not her. The property is believed to be worth over $4,800,000 today. It features about 3,613 square feet of living space on a 13,891 square foot lot.

San Francisco

Public records link Nayak to the property at 1472 Dolores Street, San Francisco, California 94110 as far back as 1999 and as recently as this year. The property owner is listed as Nigam Nayak 2003 Trust. It is common for high value targets to purchase property through trusts in an effort to conceal their location by keeping it from being linked to their names in property records. Nayak began working for Stratify in 1999. Stratify is located in the San Francisco Bay Area. The property was purchased for just over $300,000 and is believed to be worth just over $606,000 today, but some estimates online are as high as $2 million. Those seem like low figures for someone of Nayak’s stature, so we suspect him of spending most of his time at his multimillion dollar property in Palo Alto these days. He likely keeps his old house to use whenever he needs to stay in San Francisco, or he might rent it out.

Contact Information

Our background check service lists [email protected] as a possible email address. The land line for the San Francisco property appears to be (415) 826-2642 and the land line for the Palo Alto property appears to be 650-813-0127. Our background check service links the mobile numbers 650-804-1286 and 650-804-1286 to him with a confidence rating of 4/5.

Professional Career

Nayak began his professional career working for NASA before moving into the private sector where he’s worked for Stratify and Google.

NASA

Nayak began working as a research scientist for NASA in 1992. He worked at the Ames Research Center in Mountain View, California. A lot of people think that Google created Mountain View, but it turns out that it existed before Google.

Stratify

Nayak left NASA to become Chief Architect at Stratify when the company was founded in 1999. He eventually became Stratify’s Chief Technology Officer.

Google

Nayak started working at Google around 2004 or 2005. He is now in charge of search quality. We are not sure exactly when Nayak started working in search or when he took over search quality, but online research indicates that there were not many articles written about him until 2019.

As head of search his job is to develop algorithmic solutions for determining content quality. Usually that involves developing new machine learning solutions to help Google better understand language. Nayak says the goal is to truly understand the meaning of what content authors are saying beyond simply analyzing keywords or phrases and looking for indicators of low quality content like bad grammar. His more recent work involves synonymous search which often results in users being served articles of similar meaning even if there are plenty of exact keyword matches to their query. Unfortunately, synonymous search has been known to make it harder for users to find what they are looking for because it often results in irrelevant articles ranking higher because they are believed to be of the same meaning and of higher quality. We highlighted one such example in our last piece.

You can learn more about Nayak’s work in the video below:

More recently, Nayak has admitted that employees in his department have been manually sabotaging search results for the purpose of preventing specific sites from ranking well for specific searches. We call it sabotage because Google’s “mission is to organize the world’s information and make it universally accessible and useful.” Suppressing search results to the point that they rank below large quantities of irrelevant content contradicts that mission. That is especially true when the information is suppressed for being too useful (ex: mugshots).

Dry Snitching

A dry snitch is someone who tells on someone indirectly and often unintentionally without actually informing on them to an authority. Nayak dry snitched on Google CEO Sundar Pichai in a blog post last month by saying that employees in his department were manually manipulating search results after Pichai testified under oath that “it is not possible for an individual employee or groups of employees to manipulate our search results.” The relevant portions of Nayak’s post with the most damning parts in bold are as follows:

We design our ranking systems to surface high quality results for as many queries as possible, but some types of queries are more susceptible to bad actors and require specialized solutions. One such example is websites that employ exploitative removals practices. These are sites that require payment to remove content, and since 2018 we’ve had a policy that enables people to request removal of pages with information about them from our results.

Beyond removing these pages from appearing in Google Search, we also used these removals as a demotion signal in Search, so that sites that have these exploitative practices rank lower in results. This solution leads the industry, and is effective in helping people who are victims of harassment from these sites…

To help people who are dealing with extraordinary cases of repeated harassment, we’re implementing an improvement to our approach to further protect known victims. Now, once someone has requested a removal from one site with predatory practices, we will automatically apply ranking protections to help prevent content from other similar low quality sites appearing in search results for people’s names. We’re also looking to expand these protections further, as part of our ongoing work in this space.

This change was inspired by a similar approach we’ve taken with victims of non-consensual explicit content, commonly known as revenge porn. While no solution is perfect, our evaluations show that these changes meaningfully improve the quality of our results.

– Pandu Nayak

The “specialized solutions” Nayak is referring to are obviously a series of actions targeting some types of sites for some types of queries. The first action occurs when a Google employee reviews a request to remove search results from sites with allegedly “exploitative removal practices.” The second action involves registering the name of the requester as a “known victim” if the employee agrees that the site in question has an “exploitative removal practice.” The third action involves blacklisting the site itself for having an “exploitative removal practice.” We will review each action in detail.

Action 1: Reviewing the Request

We evaluated this step by submitting a pair of complaints to Google regarding information about our people. The first involved a site that does not charge removal fees. In that case we received an email from an anonymous Google employee saying that they found no evidence of an “exploitative removal practice.” The second request involved a site that allowed users to seek removals through an arbitration service. In that case we received an email from Google saying that they would remove the result citing their policy on sites with “exploitative removal practices.” This was somewhat surprising because the site didn’t specifically say it would remove something for a fee. Just that you could purchase arbitration services. We say somewhat surprising because some of our sites that were hit never removed content for a fee, but did allow users to purchase noindex meta tags which is technically adding content and not removing it even though those tags direct search engines not to include specific pages in their results. This part of the process is clearly a manual action as it requires a human employee to personally review the request.

Action 2: Registering the “Victim”

The name of the person who filed the successful removal request is added to a list which includes the names of every other person whose filed a successful request. This could be done manually by the employee who approves the request or it might be done automatically by using a script to insert the name of the person that filed the request into a database after the request is approved. We suspect this of being done manually due to all the input fields on the Google form. They ask for several things like your name, if you are submitting the request on behalf or someone else and if so their name, the URLs involved, and the queries involved. Users are notorious for screwing up on forms like that, so we think Google employees probably add the names manually to make sure they have the right ones. Doing otherwise could result in the wrong names being added or being added with errors.

Action 3: Blacklisting the Site

The website labeled by a Google employee as exploitative is blacklisted so that it does not rank highly in Google results for names of individuals registered as “known victims” and it also suffers a demotion that causes it to rank lower in names searches generally. This action could be manual or automatic. We suspect that it is a manual action in which a Google employee personally adds the site to a list, but we also have data indicating that this could be an automatic step based on the number of successful removals. A New York Times story discussing the impact of the change noted that a new site which doesn’t overtly advertise paid removal options was not impacted by the change. Their theory was that the site may not have received enough removal requests to qualify for the blacklist. That was over a month ago and because the site was mentioned by name we suspect that plenty of removal requests followed. Today we looked at Google searches for a handful of people whose names are listed on that site and most of them have first page results from that site. Obviously Google employees did not take the manual action of blacklisting it, but that doesn’t rule out the minimal removal threshold theory because Google employees probably have not been approving removal requests for that one.

The third action proves that this has more to do with public relations than improving the quality of search results. When the only differentiating factor between multiple sites with substantially similar content are their removal policies then clearly quality is not the issue.

Why This is Illegal

This is illegal because Pichai’s testimony made it impossible for Google employees to manually manipulate search results without engaging in a deceptive trade practice. We discussed this in depth recently. In his effort to get some good publicity for the company, Nayak admitted not only that it was possible for Google employees to manipulate search results, but also that they had been secretly doing so for quite some time. On top of that the three actions needed for this mechanism to work are the same three actions that would be needed to suppress any other category of site for any other list of search terms (ex: right wing media and known conservatives).

It does not matter that Nayak only recently admitted to the change because Pichai did more than simply deny that Google employees were manipulating results while under oath in 2018. He said it was not possible for them to manipulate search rankings. Even if Google employees were not manipulating search ranking back then it was certainly possible for them to start. If it were not possible for them to start then they would not have been able to make the changes they’ve admitted to making.

Imagine if you had a tall cherry tree in your yard and I asked, “is it possible for you to pick cherries from the top of the tree?” To which you answer, “it is not possible.” Then a few days later I see you picking cherries from the top of the tree and call you liar. You respond, “it was not possible at the time because I left my ladder at work.” Not being in direct possession of your ladder at the time would not make your original answer any less false. The possibility of you being able to pick the cherries always existed, you just needed to go get your ladder. What if your work were closed for the weekend, you didn’t have a key, and your boss wouldn’t let you in? Then the correct answer would be, “yes, its possible, but I need to pickup my ladder from work and I won’t be able to do that for a few days.” Even if you did not own a ladder you still could not correctly deny the possibility of being able to pick those cherries because you know that all you need is a ladder.

The above analogy might not be exactly the same, but it is close enough. Google employees have always had the ability to manually manipulate search rankings even if the mechanism Nayak described was not available exactly as is back in 2018. Surely the CEO of Google would know that people working in the search department were capable of manually manipulating search in this way. They already had manual mechanisms in place to demote sites caught manipulating search ranking by engaging in link schemes and other activities in violation of their quality guidelines. According to Google’s manual actions policy of September 4, 2018, “Most issues reported here will result in pages or sites being ranked lower or omitted from search results without any visual indication to the user.” That alone proves that it was both possible and common practice for employees to manipulate search rankings before December of that year.

Where the manual manipulations become deceptive is when they are done for reasons not included in their webmaster guidelines and do not result in a manual action report being sent to the webmaster. Google’s policy states, “If your site is affected by a manual action, we will notify you in the Manual Actions report and in the Search Console message center.” This left us scratching our heads wondering what was wrong because surely if Google took manual action against us they would notify us per their policy. Yet here we are today with Nayak admitting to manual actions that do not include notifications. Is this because the manual actions were not in response to violations of the webmaster guidelines? As you can see there is no mention of “exploitative removal practices” on the manual actions list, so Google might try to rebut this claim simply by saying that the manual action taken against us was not the type of manual action on the list, so no notification was sent. That does not make the list any less deceptive since anything that might result in a manual action should be on that list.

The FTC Statement on Deception says, “the Commission will find deception if there is a representation, omission or practice that is likely to mislead the consumer acting reasonably in the circumstances, to the consumer’s detriment.” When Google’s CEO represents that it is not possible for his employees to manually manipulate search rankings then consumers like us will reasonably conclude that we will never have to worry about our sites being demoted manually for political reasons. Had Pichai simply testified honestly our business plan would have been drastically different. That conclusion leads people to reasonable believe that they are free to offend as many people as they wish without their Google rankings suffering as a result. In fact, being offensive often improves Google rankings by increasing backlinks, social media buzz, and media coverage. Google’s algorithm would have likely boosted our rankings just for being mentioned by The New York Times. It took a human to rob us of that.

What is Legal

It is perfectly legal for Google to write programs that might cause sites like ours not to rank highly, but they would still likely rank highly for niche terms with little competition. That is why Google illegally manipulated the results because they realized no matter how poorly the quality of the content were categorized by the algorithm it would still rank in name searches for people with little web presence. Still, they could use a lot of stuff that typically shows up in that type of content as a ranking indicator without deceptively engaging in manual manipulation.

Keywords and phrases such as cheater, homewrecker, skank, whore, slore, sloot, and cheater report to name a few could easily be flagged as indicators of low quality content just like names of medications or other spammy things. Programming the algorithm to rank pages with those types of words lower in general would not require manual actions on a per site or per term basis.

The best rules of thumb for them to make sure they are not engaging in perjury or unlawfully deceptive trade practices are simple: First, don’t like under oath; Second, don’t make false statements in pubic; Third, don’t do anything that contradicts your sworn or pubic statements.

Conclusion

Pandu Nayak incriminated his boss by proving that he committed perjury and admitting to deceptive trade practices while sabotaging his company’s core service. Search quality would be better if Google adopted the warning label proposal from our manifesto.