You know those moments when you Google something simple—like What temperature should chicken be cooked to?—and instead of getting a direct answer, you’re dropped into a 1,500-word article that includes a history of poultry, the risks of salmonella, and a giant pop-up newsletter signup box? Yeah, we’ve all been there. And it’s a perfect example of the tension at the heart of the U.S. government’s antitrust case against Google.
The Department of Justice (DOJ) , backed by nearly every state and territory, sued Google for abusing its monopoly in the search and advertising markets. While much of the case focused on deals with phone manufacturers and browsers, the most fascinating aspect for those of us in SEO and digital marketing came from the internal documents and testimony about how Google’s search engine works. For the first time, we got a behind-the-scenes look at how content gets ranked, what signals Google trusts most, and what determines whether your page shows up—or disappears—on page one.
As someone who’s spent decades watching Google evolve, I found these revelations both validating and frustrating—validating because Google’s engineers confirmed many best practices in SEO. It’s frustrating because it’s now clearer than ever how the system is fine-tuned in ways that sometimes feel like they’re meant to serve Google more than the user.
This article breaks down the key takeaways from the DOJ case—not in encyclopedic terms, but in plain language for marketers, business owners, and content creators trying to understand what’s really going on when someone hits search.
Table of ContentsThe Core Ranking Signals: Anchors, Body, ClicksA Linear Formula, Not a Black BoxHand-Tuning Over Deep LearningUser Behavior Matters, But Is NormalizedManual Interventions for Public InterestData from Human Raters Is Used to Train Ranking ModelsWhole-Page Ranking and the Role of TangramNo Explicit Favoritism for Google-Owned PropertiesMy Advice for Businesses
The Core Ranking Signals: Anchors, Body, Clicks
We still rely on ABC: Anchors, Body, and Clicks. Anchors are what others say about a page, Body is the content itself, and Clicks measure how users respond to it.
Hyung-Jin Kim, Google Search VP (DOJ Exhibit, Testimony)
At the heart of Google’s ranking system is something it calls the ABC framework. Think of it like the core of a restaurant review:
Anchors: These are the links pointing to your page from other sites. If trustworthy websites are referencing your content, Google assumes you’re credible. It’s like getting recommended by a top food critic.
Body: This is your actual content. Is it clear? Is it focused? Is it relevant to what the person is searching for? It’s the menu and experience—what you’re serving your readers.
Clicks: This is how people respond. Do they click and stay? Do they bounce right back to try something else? If users are leaving right away, Google assumes your content didn’t satisfy the query.
Now, here’s the nuance: Google doesn’t treat all three signals equally for every search. If someone is researching a health condition, they may give authoritative links more weight. If it’s a local restaurant review, user engagement might tip the scale. Although the model appears simple, it’s applied flexibly.
From a marketer’s perspective, this framework affirms the basics: be credible, be useful, and be worth the click. But here’s where things start to get more complex—and honestly, more human.
A Linear Formula, Not a Black Box
Most of our signals are combined in a linear way. They use log-scaling, but it’s still a weighted sum… and we aim for monotonicity.
Google Search Engineer (DOJ Evidence, Ranking Design Document)
Contrary to what many believe, Google’s algorithm isn’t an unknowable AI overlord throwing dice. It’s more like an elaborate spreadsheet where different factors get scored, scaled, and added up. So if your content is strong and your site is well-structured, that’s going to help you—predictably and measurably.
This is encouraging. It means SEO isn’t about gaming the system. It’s about building assets that move the right dials. And it reinforces something I’ve told clients for years: Google doesn’t reward tricks; it rewards quality. There’s still room to compete if you focus on fundamentals.
Hand-Tuning Over Deep Learning
We hand-craft most ranking signals. We do it so we can fix things when they go wrong. Deep learning has made debugging harder.
Hyung-Jin Kim, Google Search VP (DOJ Testimony)
One of the more surprising admissions was just how much Google resists letting AI fully drive search rankings. Instead, engineers hand-tune most of the signals. Why? Because if something breaks, they want to know how to fix it.
That’s both reassuring and eye-opening. It’s reassuring because it means there are real humans behind the algorithm trying to maintain quality. But it’s also frustrating—because when Google decides a signal isn’t working right, they can tweak it overnight, and your traffic could crater without warning.
In my work, I’ve seen clients hit by updates not because they did something wrong, but because Google decided a new threshold was more accurate. This is why SEO is often more of an art than a science. There’s always a bit of chasing the algorithm’s latest mood.
User Behavior Matters, But Is Normalized
Click-based signals are adjusted for position bias. A link getting clicked just because it’s #1 doesn’t mean it’s better.
Internal Google Ranking Notes (DOJ Exhibit)
Click data does matter—but not in the way you might think. If you’re in the top spot, of course, you’re going to get more clicks. Google is aware of this, and they’ve built systems to normalize the data. So, just having a high CTR doesn’t automatically help you rank higher.
What matters more is something called dwell time—how long people stay on your site after clicking. If they stick around, that’s a good sign. If they bounce immediately, that’s a red flag. This is why satisfying the user’s intent is more important than tricking them into clicking.
This is one of those things that frustrates me. A perfect, concise article that doesn’t take long to digest can be viewed as the opposite of engagement. IMO, this is why AI is starting to eat away at Google’s market share. If answering the user wasn’t important… why does Google now put AIO front and center on every SERP?
Manual Interventions for Public Interest
We can adjust signals for specific edge cases or public/media attention issues.
Google Internal Communication (DOJ Evidence)
Google says it doesn’t manually promote or demote specific sites—but they do make interventions when things go off the rails. If a spammy site starts ranking for a sensitive topic, or if there’s a media outcry about misinformation, Google engineers can step in and tweak the signal behavior to correct the problem.
This isn’t necessarily nefarious—it’s just how large systems are managed. However, for website owners, it means you may be affected by adjustments that weren’t intended for you. I’ve seen sites get caught in these crossfires, particularly during major core updates. It’s part of why SEO requires resilience.
Data from Human Raters Is Used to Train Ranking Models
We use labeled data from human quality raters to guide what we want the algorithm to learn.
Google Ranking Engineer (DOJ Testimony)
You may have heard of Google’s Search Quality Rater Guidelines. The DOJ case confirmed what we suspected: Google uses real human evaluations to teach the algorithm what good content looks like.
These raters don’t directly change rankings, but their assessments help train the models that do. So if your content aligns with their guidelines—demonstrating expertise, trustworthiness, and usefulness—you’re building long-term ranking potential.
Whole-Page Ranking and the Role of Tangram
Tangram optimizes the layout of the search page based on engagement. It uses data from systems like Navboost and Glue.
Google Search Team Member (DOJ Exhibit)
One of the more subtle shifts in search is that it’s no longer just about ranking ten blue links. Google uses a system called Tangram to determine how the entire page is arranged—what gets featured, what gets boxed, and what shows up as a map or video snippet.
This is why even if your content ranks well, it might be buried under a local pack, an answer box, or a flight widget. It’s not always about quality. Sometimes, it’s about what Google believes will get the most engagement.
That brings us back to the chicken example. Google may decide to answer your query in a summary box—and never send you to a website at all. That’s great for the user. However, it’s brutal for publishers who invest in content and have explicitly written it according to Google’s quality content guidelines.
No Explicit Favoritism for Google-Owned Properties
The same signals apply to all websites. We do not have algorithmic labels for competitor or owned property.
Google Search Executive (DOJ Testimony)
Google insists it doesn’t give itself special treatment in rankings. And technically, that’s true—the same signals judge YouTube and Google Flights. But here’s the rub: those properties often have more user data, better-structured content, and higher engagement. So they rise naturally in a system tuned to reward those things.
It’s like playing poker with someone who’s not cheating—but who also happens to own the casino. The cards may be fair, but the advantage is built into the environment. Conversely, I think it’s annoying that Google focuses so much on our site’s experience while integrating their analytics and ad platforms may have a derogatory impact on CWV and user experience.
My Advice for Businesses
Relevance and Authority Still Reign: High-quality content and strong backlinks continue to be the foundation of good rankings.
Optimize for Engagement: Google watches what users do—so build pages that satisfy the query and hold attention.
No More Shortcuts: Tricks like CTR manipulation or keyword stuffing don’t work—and might backfire.
Structure Content Smartly: Use headers, natural language, and schema to help Google understand your topic.
SERPs Are Competitive Real Estate: Even top content can get crowded out by features like maps, videos, or AI summaries.
Expect Constant Change: Google’s hand-tuned updates require you to stay adaptable and closely monitor your performance.
For years, SEO experts have speculated about what truly drives Google’s decisions. Thanks to the DOJ’s antitrust case, we now have confirmation straight from the source. And the truth is… the basics still matter. Good content, trusted backlinks, and a satisfying user experience are still the most clear paths to ranking well.
But there’s also a growing tension. Google is trying to serve users faster, cleaner answers. That’s good for searchers—but it means fewer clicks for publishers. AI-driven summaries, answer boxes, and SERP features are crowding out traditional content. And that raises a bigger question: is Google still a gateway to the web, or is it becoming the destination itself?
If you’re a business owner or marketer, the answer is that you have to play both sides. Create content that earns trust and engagement. But also anticipate how Google might repurpose that content—because increasingly, your best shot at visibility might be getting cited in a summary, not just ranking #1.
SEO is evolving. And if we want to stay visible in a world where AI rewrites search, and Google owns the page, we need to evolve with it.
©2025 DK New Media, LLC, All rights reserved | DisclosureOriginally Published on Martech Zone: Inside Google’s Ranking Algorithms: Insights from the DOJ Antitrust Case