Definition
Information gainInformation gain is how much new information a page adds beyond what a reader, or a search engine, has already seen on the same topic. The term comes from a Google patent that describes scoring documents this way. In practice, it means original data, firsthand experience, or specific details that other pages leave out.
How Searchable measures this
Searchable's Sources view shows the pages AI engines already cite for your tracked prompts, and Gap Analysis highlights sources that name competitors without citing you. Reading those pages first shows what's already covered, so a new page can add what they leave out.
Where information gain comes from
The term comes from a Google patent titled Contextual estimation of link information gain. It was filed in 2018 and granted in June 2022 as US 11,354,342. It describes scoring a document by how much additional information it offers beyond documents the user has already seen, so the next result shown can add something new.
The patent doesn't prove Google ranks pages this way. Patents describe ideas a company has explored, and Google hasn't said whether information gain scoring is part of Search.
Why information gain matters for AI search
AI answers are built from several sources at once. If ten pages say the same thing, an engine only needs one of them, and the other nine add nothing to the answer. A page that adds a fact the others don't have gives the engine a reason to use it, and often to cite it.
Research on generative engine optimization points the same way. The 2023 GEO paper tested content changes and found that adding statistics, quotations, and cited sources were among the methods that most improved a page's visibility in AI answers.
Examples of information gain
Information gain is anything a reader can't get from the pages already ranking or cited:
- Original data, such as a survey or an analysis of your own product data.
- Firsthand testing, such as what happened when you tried something, with screenshots.
- Specific facts that others round off or skip, like exact limits or dates.
- Expert input from someone who has done the work, quoted by name.
Here's an example from our own data. In September 2026, across 3,736 brands tracked in Searchable, the median brand was named in 93% of answers to branded prompts and only 7% of answers to unbranded ones. No other page had that figure, and it's the kind of detail a page on branded vs unbranded prompts can be cited for.
How to add information gain to a page
Start with what's already out there, then add to it:
- Read the current sources. Look at the top search results and the pages AI engines cite for the questions you want to win.
- List the consensus. Write down the points every source makes. You'll need to cover some of them, but they won't set you apart.
- Find the gap. Look for questions they skip, claims they don't support, and data they don't have.
- Put the new part near the top. Lead with what's new, since engines and readers both decide quickly whether a page is worth using.
Label new data clearly, with the date, the sample, and how it was collected. A specific, sourced figure is easier for an engine to quote than a vague claim.
Information gain vs E-E-A-T
The two ideas support each other. E-E-A-T describes the experience, expertise, authoritativeness, and trust behind content. Information gain describes what the content adds. Firsthand experience is often what produces new information, because it's the one thing other writers can't copy.
Frequently asked questions
Google hasn't confirmed it. A patent shows what Google has explored. It doesn't show what Google uses. Google's public guidance does ask whether content provides original information or analysis beyond the obvious, which is the same idea in plain terms.
See how AI engines talk about your brand
Get a free report on where ChatGPT, Perplexity, and Google AI name and cite you.