10 SEO Terms Explained: Query Fan-Out, GEO and More

Ten SEO terms now shape how pages get found and quoted: query fan-out (one question becomes many searches), information gain (what your page adds that others do not), keyword cannibalization (your own pages competing), crawl budget (how much Google will crawl), index bloat (too many low-value pages indexed), striking distance (rankings just off page one), passage ranking (Google ranking sections, not only pages), site reputation abuse (a spam policy), entity salience (how central a topic is in your text) and GEO (optimizing for AI answers). Each one maps to a practical task you can do this week.

This glossary defines every term in plain language, says whether it is an official Google concept or an industry term, shows how to check it and what to do about it. Claims are cited, and where Google has not confirmed something, we say so. Numbers in brackets point to the references at the end.

#Key takeaways

• Some terms are official, some are not. Query fan-out, crawl budget, passage ranking and site reputation abuse appear in Google's own communications. Cannibalization, index bloat and striking distance are industry labels. Information gain comes from a patent. GEO comes from academic research.
• Google says there is no special optimisation for AI features. Pages need to be indexed and snippet-eligible, and fundamentals still apply [1].
• Crawl budget matters mainly for very large sites, per Google's own guidance [2].
• Most of these come down to the same discipline: one clear page per intent, useful original content, clean indexing and a technically healthy site.

#The 10 terms at a glance

#1. Query fan-out

Definition: query fan-out is a technique where a system issues multiple related searches across subtopics and data sources to build one response. Google's documentation says both AI Overviews and AI Mode may use it [1].

Why it matters: a single question such as "best CRM for a small agency" can trigger separate searches for pricing, integrations, reviews and alternatives. Your page may be retrieved for a sub-question, not the original query.

How to check: list the sub-questions a searcher would have after reading your page. Look at "People also ask," related searches and your own Search Console queries for the long-tail variants.

What to do:

• Cover the obvious follow-up questions in clear sections with their own headings.
• Answer each in the first sentence or two, then expand.
• Link related pages together so the topic cluster is easy to follow.

Common mistake: chasing a magic fan-out format. Google states there are no additional requirements or special optimisations needed for AI Overviews or AI Mode [1].

#2. Information gain

Definition: information gain describes how much new information a document adds beyond what a reader has already seen. It comes from a Google patent titled "Contextual estimation of link information gain," which scores candidate documents by the new information they provide relative to what the user has already seen [6][7].

Caveat: a patent describes a capability, not a confirmed live ranking system. Sources disagree on its exact filing and grant dates, and Google has not said it is a ranking factor, so treat claims that it is "the number one ranking signal" with caution [6].

Why it matters anyway: the idea matches good editorial sense. If your page repeats what ten others say, there is little reason to show it.

What to do:

• Add something the top results lack: original data, a worked example, a screenshot, a process you actually ran, or a clear opinion with reasoning.
• Remove filler that restates common knowledge.
• Show first-hand experience and name your sources.

Common mistake: padding with extra words. More text is not more information.

#3. Keyword cannibalization

Definition: keyword cannibalization is when several pages on one site target the same search intent, so they compete with each other and neither ranks as well as one strong page would. It is an industry term, not an official Google penalty.

Why it matters: search engines have to guess which of your pages to show, rankings may swap between URLs, and links and signals are split.

How to check: in Google Search Console, open Performance, filter by a query, and view the Pages tab. If two or more of your URLs appear for the same important query, investigate whether they truly serve different intents.

What to do:

1. Merge near-duplicate pages into the strongest one and 301 redirect the others.
2. Differentiate pages that deserve to exist, so each targets a distinct intent (for example, "what is X" versus "X pricing").
3. Fix internal links and anchor text so the page you want to rank gets the main links.
4. Use canonical tags where pages must stay live but are near-identical.

Common mistake: deleting pages with traffic. Redirect them to the best match instead.

#4. Crawl budget

Definition: Google defines crawl budget as "the set of URLs that Google can and wants to crawl." It combines a crawl capacity limit (how much your server can handle, which falls with 5xx errors or HTTP 429 responses) and crawl demand (how much Google wants to crawl, based on size, update frequency, quality and relevance) [2].

Who needs to worry: Google's guide is aimed at sites with 1 million or more unique pages whose content changes about weekly, sites with 10,000 or more pages changing daily, and sites with many URLs reported as "Discovered - currently not indexed." Google adds that these are not exact thresholds [2]. Most small sites do not have a crawl budget problem.

What wastes crawl budget [2]:

• Duplicate content and URL variations that show the same page.
• Soft 404 pages.
• Long redirect chains.
• Unimportant URLs such as endless sort and filter combinations.

What to do [2]:

• Consolidate duplicates.
• Block unimportant URLs in robots.txt, and do not use noindex to save crawl time, since Google still has to fetch the page to see it.
• Return 404 or 410 for permanently removed pages.
• Keep sitemaps current and include <lastmod>.
• Speed up the site and return 304 Not Modified where appropriate.

Common mistake: assuming a crawl problem when the real issue is quality. A page that is crawled is not guaranteed to be indexed [2]. See also our guide to web performance and Core Web Vitals.

#5. Index bloat

Definition: index bloat is when a site has many low-value URLs in a search engine's index, such as parameter variations, thin tag pages, internal search results, old test pages and duplicates. It is an industry term.

Why it matters: it dilutes quality signals, wastes crawling and makes it harder for good pages to stand out.

How to check: use the Page indexing report in Search Console [11] to see which URLs are indexed and why others are not. Compare the number of indexed pages with the number of pages you actually want indexed. Look for patterns in the URLs.

What to do:

• Decide which URL types deserve to be indexed, and which do not.
• Remove or 404/410 pages that have no value, and redirect pages that have a good replacement [2].
• Use canonical tags for near-duplicates and noindex for pages that must exist but should not appear, remembering noindex does not save crawl budget [2].
• Block crawl traps such as infinite filter combinations with robots.txt [2].
• Keep your XML sitemap to indexable, canonical URLs only.

Common mistake: blocking a URL in robots.txt and expecting it to disappear from the index. Blocking stops crawling, so Google may never see a noindex.

#6. Striking distance keywords

Definition: striking distance keywords are queries where a page already ranks just below the first page of results, commonly somewhere around positions 8 to 20. The range is an industry convention, with no official definition.

Why it matters: these pages already have relevance, so modest improvements can move them onto page one faster than building new content.

How to find them: in Search Console, open Performance, then Queries. Add an average position filter for roughly 8 to 20 and sort by impressions. High impressions with a low click-through at those positions are the best opportunities.

What to do:

1. Check the search intent and compare your page with what ranks above it.
2. Improve the content: answer the question sooner, add missing sections, update outdated facts and add original information.
3. Strengthen the title, headings and meta description.
4. Add internal links from relevant, stronger pages.
5. Re-measure after Google has recrawled the page.

Common mistake: rewriting everything. Targeted improvements to near-page-one content are usually the efficient move.

#7. Passage ranking

Definition: passage ranking is a Google system that helps it understand and rank individual sections of a page, so a specific passage that answers a question can be surfaced even if it sits deep in a long page. Google announced it in October 2020 under the name "passage indexing," later renamed because it changes ranking, not indexing [5].

Facts: Google said it would affect about 7% of queries across all languages when fully rolled out, and it went live for US English queries on February 10, 2021 [5]. Google also said site owners did not need to make specific changes [5].

Why it matters now: the same idea powers how answer engines pick a section to quote.

What to do:

• Give each section a descriptive heading that matches a real question.
• Make sections self-contained: state the answer in the opening sentence, then explain.
• Keep paragraphs short and focused on one idea.
• Do not split one topic into many thin pages just to chase passages.

Common mistake: treating passage ranking as a trick. It rewards clear structure and quality writing, not special markup.

#8. Site reputation abuse

Definition: site reputation abuse is a Google spam policy aimed at third-party content published on an established site mainly to exploit that site's existing ranking signals. It is sometimes called "parasite SEO." Google announced the policy in March 2024 and enforcement followed in 2024 [8].

What is typically not abuse: according to secondary coverage, examples include wire services, syndication between news publications, forums, opinion columns and affiliate links with proper link treatment [9]. Verify against Google's spam policies page, since details have been updated over time.

2026 update: Google updated the policy in August 2026. Secondary reports say the change affects how manual actions apply for users in the European Economic Area, not the policy's core definition [9]. Check Google's documentation for the current position in your region.

What to do if you run a site with guest or partner content:

• Audit sections that host content from third parties, especially coupon, review and "advertorial" style pages.
• Make sure that content is genuinely relevant to your site and has real editorial involvement.
• Consider removing, noindex-ing or properly separating content that exists mainly to borrow your domain's authority.
• Mark paid and affiliate links appropriately.

Common mistake: assuming the policy only hits large publishers. Any site can have a section that looks like a rented ranking.

#9. Entity salience

Definition: an entity is a distinct thing, such as a person, company, place or product. Salience is a score for how important or central an entity is to the whole text. Google's Natural Language API reports salience as a number from 0 to 1.0, where "scores closer to 0 are less salient, while scores closer to 1.0 are highly salient" [3][10].

Caveat: Google's API documentation describes the score but does not say Google Search uses it as a ranking factor. Treat it as a diagnostic for whether your page is clearly about what you intend, not as a lever Google has promised to read.

Why it matters: if your page is meant to be about "passkeys" but the text spends most of its words on other topics, a machine reading it sees a muddled subject. A clear main entity helps both search engines and AI summaries understand the page.

How to check: paste your text into an entity analysis tool, such as Google's Natural Language demo, and see which entities rank highest.

What to do:

• Name your primary topic early, in the title, headings and opening sentences.
• Use consistent names for the same thing, and spell out acronyms once.
• Keep each page focused on one main subject.
• Make structured data match the visible text, which Google specifically recommends [1].

Common mistake: keyword stuffing. Repetition is not salience; clarity and focus are.

#10. GEO (Generative Engine Optimization)

Definition: GEO is the practice of optimising content so it is more likely to be used and cited in answers from generative AI search systems. The term comes from a 2023 research paper by Aggarwal and colleagues, accepted to KDD 2024, which proposed GEO as a framework and a benchmark called GEO-bench. The authors report that GEO methods "can boost visibility by up to 40%" in generative engine responses, and that effectiveness varies across domains [4].

Caveats:

• The 40% figure is a maximum from the authors' own benchmark, not an average or a guarantee [4].
• Google says AI Overviews and AI Mode have no extra requirements and no special optimisations needed beyond SEO fundamentals, and that no special schema, AI text files or new markup is required [1].
• "GEO" is also used as a marketing label, so be wary of anyone selling secret tactics.

Practical advice (our guidance, consistent with the fundamentals above):

• Answer the main question in the first two or three sentences.
• Make claims verifiable with named, linked sources.
• Include specific facts, definitions, steps and comparison tables that are easy to quote.
• Structure content with clear headings and FAQ sections.
• Keep the page indexable and fast.
• Earn mentions and links from reputable places, since systems favour sources they can trust.

For how language models read and use content, see how large language models work and embeddings explained.

#How the 10 terms fit together

Think of them as a pipeline:

1. Crawl: Google must reach your pages (crawl budget, index bloat).
2. Index: the right pages must be stored, without duplicates (index bloat, cannibalization).
3. Understand: the page must be clearly about a topic (entity salience, passage ranking).
4. Rank and answer: useful, original pages win positions and citations (information gain, striking distance, query fan-out, GEO).
5. Stay clean: do not host content that breaks spam policies (site reputation abuse).

#A 7-step checklist to apply this week

1. Open Search Console's Page indexing report and note why URLs are not indexed.
2. Find queries where two of your pages both appear, and fix cannibalization.
3. List your striking distance queries (positions 8 to 20) with the most impressions.
4. Improve the top five of those pages with original information and clearer answers.
5. Make sure every key page opens with a direct answer and has descriptive headings.
6. Remove or consolidate thin, duplicate or outdated URLs.
7. Audit any third-party content hosted on your domain.

Measure changes over weeks, not days, and test one thing at a time. Our guides to A/B testing and product metrics can help you set that up.

#Want a technical SEO and site performance review?

We build fast, crawlable, prerendered sites and fix the indexing issues that hold rankings back. Get a scope estimate for a technical SEO and site performance review, or see our work on AI agents and automation.

#FAQ

#What is query fan-out in SEO?

Query fan-out is when a search system breaks one question into several related searches across subtopics and data sources to build an answer. Google's documentation says AI Overviews and AI Mode may use it.

#What is information gain?

Information gain is the new information a page adds beyond what a reader has already seen. It comes from a Google patent, but Google has not confirmed it as a live ranking factor, so use it as an editorial principle.

#What is keyword cannibalization and is it bad?

It is when several pages on your site compete for the same search intent. It can dilute rankings, so merge or differentiate the pages. It is an industry term, not an official Google penalty.

#Do I need to worry about crawl budget?

Usually not. Google's guidance is aimed at very large sites (about 1 million or more pages with weekly changes) or sites with 10,000 or more pages that change daily, and says the thresholds are not exact.

#What is index bloat?

It is having many low-value URLs in the search index, such as parameter variations, thin tag pages and duplicates. Fix it by removing, redirecting, canonicalising or noindexing pages that should not rank.

#What are striking distance keywords?

Queries where you rank just below page one, often around positions 8 to 20. Improving those pages is usually faster than creating new ones.

#What is the difference between passage ranking and featured snippets?

Passage ranking helps Google understand and rank specific sections within a page. Featured snippets are a result format that displays an extracted answer. They can overlap, but they are different things.

#What is GEO and is it different from SEO?

GEO is optimising for visibility in generative AI answers. Google says no special optimisation is required for its AI features beyond SEO fundamentals, so GEO and SEO largely overlap in practice.

#What is entity salience?

A score from 0 to 1 for how central an entity is to a text, reported by Google's Natural Language API. It is a useful diagnostic for topic clarity, but Google has not said it uses it as a ranking factor.

#References

1. Google Search Central. "AI features and your website." developers.google.com/search/docs/appearance/ai-features
2. Google Search Central. "Managing crawl budget for large sites." developers.google.com/search/docs/crawling-indexing/large-site-managing-crawl-budget
3. Google Cloud. "Natural Language API: Entity (salience field)." docs.cloud.google.com/natural-language/docs/reference/rest/v1/Entity
4. Aggarwal, P., et al. "GEO: Generative Engine Optimization." arXiv:2311.09735, accepted to KDD 2024. arxiv.org/abs/2311.09735
5. Search Engine Land. "Google passage ranking now live in US English search results." 2021. searchengineland.com/google-passage-ranking-now-live-in-us-english-search-results-346034
6. Search Engine Journal. "Google's Information Gain Patent for Ranking Web Pages." searchenginejournal.com/googles-information-gain-patent-for-ranking-web-pages/524464
7. Google LLC. "Contextual estimation of link information gain," patent application US 2020/0349181. freepatentsonline.com/y2020/0349181.html
8. Google Search Central. "Spam policies for Google web search" (site reputation abuse). developers.google.com/search/docs/essentials/spam-policies
9. Search Engine Roundtable. "Google Won't Enforce Its Site Reputation Policy In The European Economic Area." seroundtable.com/google-site-reputation-policy-eea-41968.html
10. Google Cloud. "Natural Language API basics." docs.cloud.google.com/natural-language/docs/basics
11. Google Search Console Help. "Page indexing report." support.google.com/webmasters/answer/7440203

Method and limits: definitions and figures come from the primary sources above, read in October 2026. Cannibalization, index bloat and striking distance are industry terms without official Google definitions, and the thresholds given are conventions. Search policies and features change often, so verify details against Google's current documentation.