Information Gain in SEO: What the Google Patent Actually Says (And What Most Guides Get Wrong)
Most guides on this topic repeat the same summary. Here's what the real Google patent says, why the common explanation misses the point, and a plain, jargon-free way to check your own content.
Type "information gain SEO" into Google right now, and you'll find a dozen articles saying more or less the same thing: add original data, share your experience, don't just rewrite the top 10. Good advice — but almost none of them explain what the actual patent behind this idea says, and that detail changes how you should think about it.
This guide covers both sides: the real story behind the patent, and a simple, no-jargon way to check whether your own content is adding anything new at all.
What Information Gain Really Means
Strip away the technical language, and information gain is just this: does your page teach the reader something they didn't already know from the other pages they read?
If ten articles on a topic all say the same five things, a reader who opens all ten only learns those five things once. The tenth article added nothing. Information gain describes that missing value — and, in theory, a way for search engines to notice it.
That part is well understood. Where most SEO guides go wrong is in explaining how Google actually measures this.
- 📄Low information gain repeats what's already sitting on hundreds of other pages.
- 💡High information gain adds a number, an example, or an experience the reader can't get from the other open tabs.
The Patent Everyone Quotes — But Few Have Actually Read
The patent behind this idea is called "Contextual Estimation of Link Information Gain." It was filed by two Google engineers, Victor Carbune and Pedro Gonnet, in October 2018, and granted in June 2022.
Here's the part most articles skip: the patent doesn't describe scoring a webpage against every other page on the internet for a topic. It describes something narrower — predicting how much new value a page would give to a specific user, based on the pages that same user already viewed earlier in their search session.
In plain terms: it's built to answer a question like, "This person just read three pages about home loan interest rates. If I show them a fourth page next, will it actually tell them anything the first three didn't?"
That's a personalization and ranking-diversity tool, not a blanket "originality score" applied to every page on the web. Most articles blur this distinction. There's no public confirmation that Google runs every URL through a single uniqueness formula — what the patent shows is Google's clear interest in avoiding repetitive results for a single searcher's journey, which is a related idea, but not the same claim.
"Google runs every page through a single uniqueness formula and scores it against the whole internet for that topic."
It predicts whether the next page would add value for one user, based on the specific pages that user already opened in their own session.
Why the Difference Still Matters to You
You might be thinking: does this technical detail change what I should actually do? Only partly — and here's why it's still worth knowing.
- 🎯It stops you chasing a mythical "information gain score" tool or formula that doesn't officially exist.
- 🔄It reframes the goal correctly: make sure someone who reads your page after three competitors still learns something.
- 🗣️It matches what Google has said openly for years — write for people, not to fill a template that already exists ten times over.
So the patent gives you the "why." The real work is figuring out the "what."
A Simple 3-Step Way to Check Your Content
Skip the frameworks and scoring formulas. Here's a plain method you can do with a notebook and twenty minutes.
📚 Build the "everyone already knows this" list
Open the top 5–6 ranking pages for your topic and write down every point that shows up in more than one. It's not useless — readers still need this baseline — but it isn't what will make your page stand out.
🧩 Ask what isn't on that list
Most writers get stuck here, hunting for new information instead of noticing what they already have: a repeated client question, a common mistake, a real number from your own work, or a time something didn't go as planned.
⬆️ Move it to the top half
A common problem: the one original insight is buried in paragraph fourteen. If a reader — or an AI system scanning your page — has to dig for it, it won't get credit, or read.
A Small Example
Picture two pages both explaining "how long SEO takes to show results" — a very common, very repeated topic.
Repeats the usual line: "SEO typically takes 3 to 6 months." True — but it's already on hundreds of pages.
Says the same thing, then adds: a low-competition local page showed first rankings in 5–6 weeks, while a page fighting ten-year-old established sites took closer to 8 months — and explains the one factor that made the difference.
Both pages answer the question. Only one gives the reader something the other nine open tabs didn't.
How Much New Information Is Enough?
There's a trap on the other side too: cramming a page with unrelated trivia to look "unique" usually backfires. If someone searches "how long does SEO take" and your page spends three paragraphs on an unrelated case study before answering the actual question, you've hurt the reader experience you were trying to improve.
A good rule of thumb: answer the core question the way any good page would, near the top. Then use the rest of the page to add the layer only you can add — your numbers, your mistakes, your specific experience. New information should support the reader's original question, not distract from it.
A Short Checklist Before You Publish
- ✓
Did I answer the exact question the reader searched for, clearly, near the top?
- ✓
Is there at least one thing on this page that isn't sitting on the other top-ranking pages too?
- ✓
Is that one thing a real number, a real example, or a real experience — not a rephrased opinion?
- ✓
Would someone who already read three other articles on this topic still learn something from mine?
- ✓
Have I said it plainly, in one or two sentences, rather than hiding it inside a long paragraph?
If you can't answer "yes" to most of these, the content isn't wrong — it's just not finished yet.
Frequently Asked Questions
Is information gain the same thing as originality?
Not exactly. Originality means the wording is your own. Information gain means the substance is something a reader couldn't already get elsewhere. You can write a fully original paragraph that still adds zero new information if it just restates common knowledge in your own words.
Does this apply to small or local businesses too?
Yes — arguably more so. A small business usually has direct, first-hand experience — actual customers, actual results, actual mistakes — that large sites covering a topic broadly don't have. That's a real advantage if you use it instead of writing generic advice.
Has Google confirmed it uses information gain as a ranking factor?
No. Google has never confirmed the patent is actively used in ranking, and the company doesn't comment on individual patents. What Google has said publicly, repeatedly, is that it wants to reward content offering real value beyond what's already available — which lines up with the spirit of the patent, even without a confirmed direct link.
Does this matter for AI Overviews and AI chat answers too?
Yes, arguably more. When an AI system pulls an answer together from several sources, a page that only repeats what others already said gives it nothing extra to pull from. A page with one clear, specific detail the others don't have is easier for that system to notice and use.
Add one thing the others didn't.
Not more words — one real number, one real example, one real experience. That's the entire idea, stripped of the jargon.