In November of 2018, Google’s search team gave an internal presentation called “Ranking for Research.” Access was restricted to people actively working on search.
One slide tells them not to discuss how clicks are used in search, except on a need-to-know basis with people who understand not to discuss it outside the company.
Then one more line: “Google has a public position.”
The company you’re trying to rank on wrote a slide telling its own engineers what not to say about how you get ranked.
So today’s question is simple. How much of what you’ve been told about search is the practice, and how much is the position?
Hi, I’m Jeff Payne. You’re listening to The Jeff Payne Show, Episode #87: The Public Position.
That presentation became public because it was entered as evidence in the government’s antitrust case against Google. It’s a court exhibit now. Anyone can read it.
To be fair to Google, the reasons for the secrecy are right there on the slides. They’re sued constantly. Their emails get retained, subpoenaed, and leaked. And the slides name who would use any leak against them: people trying to manipulate their rankings.
That’s a reasonable business decision, but it has a consequence for you. Anything you hear that starts with “Google says,” whether it comes from Google, a vendor, or an agency pitch, is based on what Google chose to say. The part it calls its public position.
So if the rulebook isn’t available, what do you steer by?
A second exhibit from the same case answers that better than any public statement has.
It’s an internal Google presentation from 2016 about predicting clicks. One slide says Google’s reliance on user clicks in ranking had steadily increased over the previous decade. The same slide says that showing people results they want to click is not the goal, because chasing clicks alone would reward clickbait and pages that are appealing but irrelevant.
The next slide says showing results people want to click is close to the goal. It’s close enough that Google can do the almost-right thing extremely well, drawing on trillions of examples of real behavior. The strategy it lays out is to predict what people will choose, boost it, and then patch problems like quality and relevance on top.
Read that as a business owner. Google watches whether real people pick you. Then it corrects for whether you deserved to be picked.
That’s why “what are the ranking factors?” is the wrong question. Even inside Google, the factors are a patchwork around one central measurement: what people would choose, given a fair opportunity.
It’s also why gaming the system tends to backfire. Manufactured signals, like fake clicks or bought engagement, feed the exact part of the system Google already knows to distrust. The patches exist for that reason.
So here’s the line to remember: picked, not placed.
You don’t get placed at the top by knowing a secret. You get picked. Customers search for you by name, choose you when they see you, stay because you answered their question, and come back. That behavior is what the system was built to read.
This isn’t only true of Google. AI answer engines won’t explain how they choose sources either. What they share is a reliance on evidence the brand didn’t manufacture: what customers do, and what other people say about you.
So stop looking for the rulebook, and be skeptical of anyone who claims to have it. Ask a different question about your marketing: is this making us more likely to be chosen?
Not seen. Chosen.
Are more people searching for your company by name than a year ago? When you show up in results, do people pick you over the listing next to yours? When they land, do they get their answer, or do they bounce back to search?
Nobody can hand you those signals. You can’t fake them for long. You can only earn them, and they compound, because every person who chooses you makes the next choice easier.
And here’s the upside of Google’s secrecy: your competitors don’t have the rulebook either. The ones who keep hunting for it will keep losing ground to the ones who got busy being the obvious pick.
Three questions to take with you.
If Google published all its ranking factors tomorrow, how much would your marketing change? And if the answer is “a lot,” what have you been optimizing for?
When someone searches your category and sees you next to a competitor, what reason do they have to pick you?
And the last time someone told you “Google says,” did you ask whether that was the practice or the position?
I’m Jeff Payne. Thanks for listening. I will see you next time.
In November 2018, Google’s search team gave an internal presentation called “Ranking for Research.” Access was restricted to people actively working on search. One slide told them not to discuss how clicks are used in search, except on a need-to-know basis with people who understood not to talk about it outside the company. Then it added one more line: “Google has a public position.”
The company you’re trying to rank on wrote a slide telling its own engineers what not to say about how you get ranked. That raises a question every business owner should ask: how much of what you’ve been told about search is the practice, and how much is the position?
Why Google Keeps Search on a Need-to-Know Basis
That presentation became public because it was entered as evidence in the U.S. government’s antitrust case against Google. It’s a court exhibit now, and anyone can read it.
To be fair to Google, the reasons for the secrecy are on the slides themselves. The company is sued constantly. Its emails are retained, subpoenaed, and leaked. And the deck names who would use any leak against it: people trying to manipulate rankings.
That’s a reasonable business decision, but it carries a consequence for you. Anything you hear that begins with “Google says,” whether it comes from Google, a vendor, or an agency pitch, is built on the part Google chose to say.
Anything that starts with “Google says” is built on the part Google chose to say.
What Actually GETS MeasureD WHEN IT COMES TO GOOGLE RANKING FACTORS
A second exhibit from the same case says more than any public statement has. It’s an internal Google presentation from 2016 about predicting clicks.
One slide notes that Google’s reliance on user clicks in Google ranking had steadily increased over the previous decade. The same slide is clear that showing people results they want to click is not the goal, because chasing clicks alone would reward clickbait and pages that are appealing but irrelevant.
The next slide says showing results people want to click is close to the goal, close enough that Google can do the almost-right thing extremely well by drawing on trillions of examples of real behavior. The strategy it describes: predict what people will choose, boost it, then patch problems like quality and relevance on top.
Read that as a business owner. Even inside Google, the ranking factors are a patchwork built around one central measurement: what people would choose, given a fair opportunity.
WHY GAMING THE SIGNAL BACKFIRES
This is also why manufactured signals tend to fail. Fake clicks and bought engagement feed the exact part of the system Google already knows to distrust. The patches exist for that reason.
You don’t get placed at the top by knowing a secret. You get picked. Customers search for you by name, choose you when they see you, stay because you answered their question, and come back. That behavior is what the system was built to read.
The Same Rule Applies to AI Answers
This isn’t only true of Google. AI answer engines won’t explain how they choose sources either. What they share with search is a reliance on evidence the brand didn’t manufacture: what customers do, and what other people say about you.
That’s the practical upside of a black box. When no one can see the rulebook, the durable advantage goes to the business that doesn’t need one.
When no one can see the rulebook, the advantage goes to the business that doesn’t need one.
The Question to Ask Instead
Stop looking for the rulebook, and be skeptical of anyone who claims to have it. Ask a different question about your marketing: is this making us more likely to be chosen?
Are more people searching for your company by name than a year ago? When you appear in results, do people pick you over the listing next to yours? When they land, do they get their answer, or do they bounce back to search?
Nobody can hand you those signals, and you can’t fake them for long. You can only earn them, and they compound, because every person who chooses you makes the next choice easier. Your competitors don’t have the rulebook either. The ones who keep hunting for it will keep losing ground to the ones who got busy being the obvious pick.
A buyer will forgive being wrong about something. Nobody hires a company they have decided is unreliable.
Three Questions for Your Next Marketing Review
Sources: U.S. and Plaintiff States v. Google LLC (2020), trial exhibits UPX0204 (“Ranking for Research,” Nov. 16, 2018) and UPX0192 (“Unified Click Prediction,” 2016), published by the U.S. Department of Justice. This episode was prompted by Edward Sturm’s podcast episode “Google’s Leaked Internal Presentation: ‘Don’t Talk About Clicks.’”
DOJ trial exhibit list, U.S. v. Google (2020)
Exhibit UPX0192 (PDF)
COMPLETE THE FORM TO
BOOK A STRATEGY CALL
"*" indicates required fields
COMPLETE THE FORM TO
BOOK A STRATEGY CALL
"*" indicates required fields
Subscribe and Share – WE APPRECIATE YOUR SUPPORT