{ "dataset": "glossary", "source": "https://best-seo-software.org", "licence": "CC BY 4.0 — attribution required", "generated": "2026-08-09T14:47:56.209Z", "count": 331, "records": [ { "id": "redirect-301", "slug": "redirect-301", "term": "301 redirect", "aka": [ "Permanent redirect" ], "category": "technical", "shortDefinition": "A server-side permanent redirect that tells crawlers a URL has moved for good and that ranking signals should consolidate on the destination.", "definition": "Google treats 301 (and 308) as a strong canonicalisation signal, though the destination must be a genuine equivalent — mass-redirecting retired pages to the homepage is treated as a soft 404. Signal consolidation is not instant; it happens as Google recrawls the old URLs. Chains should be collapsed to a single hop, and internal links should be updated to point at final destinations rather than relying on redirects.", "example": "During an HTTPS migration every `http://` URL 301s to its exact `https://` equivalent, preserving path and query.", "related": [ "redirect-302", "redirect-308", "redirect-chain", "canonicalization", "site-migration" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/301-redirects" }, { "id": "redirect-302", "slug": "redirect-302", "term": "302 redirect", "aka": [ "Found", "Temporary redirect" ], "category": "technical", "shortDefinition": "A temporary redirect indicating the resource lives elsewhere for now; the original URL is expected to remain the canonical one.", "definition": "Because it is temporary, Google generally keeps the source URL indexed and does not transfer signals the way a 301 does — though if a 302 persists for a long time Google may treat it as permanent. 302 is correct for genuinely temporary situations such as maintenance pages, A/B tests and short-lived promotions. Using 302 for a permanent move slows consolidation and can leave both URLs competing.", "example": "A store 302s a sold-out product to its category page during a two-week restock, then removes the redirect.", "related": [ "redirect-301", "redirect-307", "canonicalization", "http-status-codes", "out-of-stock-handling" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/301-redirects" }, { "id": "redirect-307", "slug": "redirect-307", "term": "307 redirect", "aka": [ "Temporary Redirect" ], "category": "technical", "shortDefinition": "A temporary redirect that, unlike 302, guarantees the HTTP method and body are preserved; also the code browsers report for internal HSTS redirects.", "definition": "307 is the strict temporary counterpart to 308. In SEO work it is most often seen as '307 Internal Redirect' in browser devtools, which is not a server response at all but the browser upgrading http to https because of an HSTS policy or the preload list. Treat it as equivalent to 302 for indexing purposes.", "example": "DevTools shows `307 Internal Redirect` on an http request because the domain is on the HSTS preload list.", "related": [ "redirect-302", "redirect-308", "hsts", "https", "http-status-codes" ], "sourceUrl": "https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/307" }, { "id": "redirect-308", "slug": "redirect-308", "term": "308 redirect", "aka": [ "Permanent Redirect" ], "category": "technical", "shortDefinition": "A permanent redirect equivalent to 301 for indexing purposes, but which forbids changing the HTTP method when following it.", "definition": "308 was added to correct a historic ambiguity in 301, where clients often converted POST to GET. Google states it treats 308 the same as 301 for canonicalisation. It is the safer choice for API endpoints and form submission targets; for ordinary page moves 301 and 308 are interchangeable.", "example": "A framework's trailing-slash normalisation emits 308 rather than 301, with no difference in how Google consolidates signals.", "related": [ "redirect-301", "redirect-307", "canonicalization", "http-status-codes", "site-migration" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/301-redirects" }, { "id": "affiliate-seo", "slug": "affiliate-seo", "term": "Affiliate SEO", "aka": [ "Affiliate marketing SEO" ], "category": "content", "shortDefinition": "Earning search traffic for content monetised through affiliate commissions; permitted by Google when the content adds substantial original value.", "definition": "Google's spam policies target thin affiliate pages — product descriptions and reviews copied from the merchant with nothing added. Compliant affiliate content demonstrates first-hand testing, original photography or measurement, and honest comparison, and qualifies outbound monetised links with rel=sponsored. Successive core and reviews updates have compressed the market for undifferentiated affiliate roundups considerably.", "example": "A cycling site buys the products it reviews, publishes its measurement rig, and marks all merchant links rel=\"sponsored\".", "related": [ "sponsored-attribute", "product-reviews-system", "commercial-investigation", "thin-content", "spam-policies" ], "sourceUrl": "https://developers.google.com/search/docs/essentials/spam-policies" }, { "id": "agency-retainer", "slug": "agency-retainer", "term": "Agency retainer models", "aka": [ "Engagement models" ], "category": "business", "shortDefinition": "The commercial structures used for SEO services: monthly retainers, project fees, day rates, performance-based arrangements or hybrids.", "definition": "Retainers suit ongoing programmes and give continuity but can drift into activity reporting without outcomes; project fees suit defined deliverables such as an audit or migration; day rates suit advisory work where the client implements. Performance-based pricing is superficially attractive but hard to structure fairly given attribution difficulty, long lag times and factors outside either party's control. Scope, deliverables and decision rights matter more to outcomes than the pricing model.", "example": "A brand retains a consultant two days a month for advisory work and contracts a separate fixed-fee project for its migration.", "related": [ "in-house-vs-agency", "seo-roi", "seo-governance", "seo-strategy", "seo-audit" ], "sourceUrl": null }, { "id": "agentic-browsing", "slug": "agentic-browsing", "term": "Agentic browsing", "aka": [ "AI agents", "Autonomous agents" ], "category": "geo-ai", "shortDefinition": "AI systems that navigate websites on a user's behalf — reading pages, filling forms, comparing options and sometimes completing transactions.", "definition": "Agents interact with sites through rendered pages, structured data, feeds or purpose-built interfaces, and they are far less tolerant of ambiguity than humans: unlabelled buttons, JavaScript-only flows, aggressive bot challenges and inconsistent product data all cause failures. Google has begun documenting agents that act on user request, and several vendors ship agentic checkout and research features. Preparing means clean semantic markup, accurate structured data, stable URLs, and deciding which automated actors to permit.", "example": "An agent asked to buy a specific part abandons a retailer whose add-to-cart requires a hover-only menu, and completes at a competitor.", "related": [ "mcp", "chatgpt-user", "structured-data", "semantic-html", "synthetic-traffic" ], "sourceUrl": null }, { "id": "content-licensing-ai", "slug": "content-licensing-ai", "term": "AI content licensing", "aka": [ "Publisher AI deals" ], "category": "geo-ai", "shortDefinition": "Commercial agreements in which publishers license their content to AI companies for training, retrieval or attribution, as an alternative to blocking crawlers.", "definition": "Since 2023 a number of large publishers have signed licensing deals with model providers, and infrastructure vendors have introduced marketplace and pay-per-crawl mechanisms so smaller sites can charge for automated access. The strategic question for any publisher is whether its content is more valuable as licensed input or as freely retrievable material that earns citations and referral traffic. The answer differs sharply between reference publishers, news organisations and commercial sites.", "example": "A trade publisher licenses its archive for training while keeping current articles open to retrieval agents for citation value.", "related": [ "model-training-opt-out", "ai-crawler", "gptbot", "zero-click", "ai-referral-traffic" ], "sourceUrl": null }, { "id": "ai-crawler", "slug": "ai-crawler", "term": "AI crawler", "aka": [ "AI bot", "LLM crawler" ], "category": "geo-ai", "shortDefinition": "An automated agent operated by an AI company to fetch web content for model training, search indexing, or live retrieval when answering a user's question.", "definition": "These fall into three functional classes with different consequences: training crawlers, search-index crawlers that build a retrievable index, and user-triggered fetchers that retrieve a page in response to a specific request. Blocking them is a business decision, not a technical one — blocking a training crawler protects content but blocking a retrieval agent removes you from answers. Most declare a user-agent token and respect robots.txt, and several publish IP ranges for verification; bot-management rules at the CDN can block them regardless of robots.txt.", "example": "A publisher allows OAI-SearchBot and ClaudeBot's search fetches while disallowing training-oriented agents in robots.txt.", "related": [ "gptbot", "claudebot", "perplexitybot", "google-extended", "model-training-opt-out" ], "sourceUrl": null }, { "id": "ai-mode", "slug": "ai-mode", "term": "AI Mode", "aka": [], "category": "geo-ai", "shortDefinition": "Google's conversational search experience for complex, multi-part questions, returning a reasoned AI response with supporting links instead of a classic result list.", "definition": "AI Mode targets queries that would previously have required several searches — comparisons, planning, nuanced reasoning — and supports follow-up questions in context. Like AI Overviews it is grounded in Google's index and requires no special optimisation, and its impressions and clicks are reported in Search Console's generative AI performance views and within the Web search type. Because a single AI Mode session can decompose into many underlying queries, the relationship between a brand's classic rankings and its appearance there is indirect.", "example": "A user asks for a two-week itinerary balancing hiking and museums; AI Mode returns a structured plan citing a national park service page and two travel publishers.", "related": [ "ai-overviews", "query-fan-out", "grounding", "generative-ai-performance-report", "zero-click" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/ai-features" }, { "id": "ai-overviews", "slug": "ai-overviews", "term": "AI Overviews", "aka": [ "AIO", "SGE" ], "category": "geo-ai", "shortDefinition": "Google's AI-generated summaries shown at the top of many search results, synthesising information from multiple web pages with links to the sources.", "definition": "AI Overviews evolved from the Search Generative Experience labs product and are grounded in Google's core search index, so the pages that can be cited are those already crawled, indexed and considered relevant. Google states there is no separate optimisation and no markup requirement; snippet controls such as `nosnippet` and `max-snippet` do govern inclusion of a page's text. Overviews are counted in Search Console as impressions when links are scrolled or expanded into view, all occupying a single position.", "example": "A query about deductible limits returns a four-sentence overview citing an insurer, a regulator and a consumer charity, pushing organic results below the fold.", "related": [ "ai-mode", "zero-click", "grounding", "search-generative-ai-control", "generative-ai-performance-report" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/ai-features" }, { "id": "ai-referral-traffic", "slug": "ai-referral-traffic", "term": "AI referral traffic", "aka": [ "LLM referrals" ], "category": "geo-ai", "shortDefinition": "Visits arriving from AI assistants and answer engines, identifiable in analytics by referrer hosts such as chatgpt.com, perplexity.ai, claude.ai and copilot.microsoft.com.", "definition": "Volumes are typically small relative to search but often convert well, because a user arriving from a recommendation has already been pre-qualified. Attribution is imperfect: some assistants strip or vary referrers, in-app browsers can be misclassified as direct, and clicks from Google's AI features are reported inside ordinary Google Search data rather than separately. Building an explicit referrer-based channel group is the standard first step.", "example": "A GA4 channel group for AI assistants shows 1.2% of sessions but a conversion rate roughly double that of organic search.", "related": [ "ga4", "attribution", "answer-engine", "generative-ai-performance-report", "synthetic-traffic" ], "sourceUrl": null }, { "id": "ai-visibility", "slug": "ai-visibility", "term": "AI visibility", "aka": [ "AI search visibility" ], "category": "geo-ai", "shortDefinition": "The extent to which a brand appears, is cited, or is recommended across AI-generated answers, measured by sampling prompts rather than by rank positions.", "definition": "Because assistants are non-deterministic, personalised and unranked, visibility is measured statistically: run a defined prompt set repeatedly across models, record mentions, citations, sentiment and position within the answer, and track the distribution over time. There is no equivalent of a rank tracker's ground truth, so methodology — prompt set, region, model version, sampling frequency — must be documented or numbers are not comparable. Search Console's generative AI reports provide the only first-party view, and only for Google surfaces.", "example": "A team runs 120 category prompts weekly across four assistants and reports the share of responses in which its brand is named.", "related": [ "citation-share", "llm-share-of-voice", "prompt-volume", "brand-mention-monitoring", "generative-ai-performance-report" ], "sourceUrl": null }, { "id": "ai-generated-content", "slug": "ai-generated-content", "term": "AI-generated content policy", "aka": [ "Generative AI content" ], "category": "content", "shortDefinition": "Google's position that content is judged on quality and purpose, not method of production — but mass-produced AI content made to game rankings is spam.", "definition": "Google states that using automation, including AI, to generate content is not against its guidelines when the result is helpful and original; what violates policy is generating content primarily to manipulate rankings, which the scaled content abuse policy covers. Practically, this means AI-assisted drafting is fine when a human adds expertise, verification and originality, while publishing unreviewed model output at volume is a quality and accuracy liability. Google does not require AI disclosure but suggests it where readers would reasonably want to know.", "example": "A team uses a model to draft outlines and summarise sources, then has a subject expert rewrite, verify and add original testing data.", "related": [ "scaled-content-abuse", "helpful-content", "thin-content", "information-gain", "hallucination" ], "sourceUrl": "https://developers.google.com/search/docs/fundamentals/using-gen-ai-content" }, { "id": "algorithm-volatility", "slug": "algorithm-volatility", "term": "Algorithm volatility", "aka": [ "SERP volatility", "Ranking flux" ], "category": "algorithms", "shortDefinition": "Day-to-day fluctuation in rankings caused by continuous algorithm changes, index updates and testing, distinct from confirmed named updates.", "definition": "Google makes thousands of changes a year, most unannounced, so a degree of movement is constant background noise. Third-party volatility trackers measure aggregate position change across keyword panels and are useful for distinguishing 'the whole SERP moved' from 'we moved'. Reacting to daily noise is a common failure mode; positions should be assessed on rolling averages against a control set.", "example": "A site's rankings drop three positions overnight while volatility trackers show market-wide movement, so no change is made.", "related": [ "core-update", "position-tracking", "spam-updates", "average-position", "ranking-systems" ], "sourceUrl": null }, { "id": "alt-text", "slug": "alt-text", "term": "Alt text", "aka": [ "Alternative text" ], "category": "technical", "shortDefinition": "The alt attribute describing an image's content and function for screen readers, for cases where the image fails to load, and for image search.", "definition": "Google uses alt text alongside surrounding content and the filename to understand images for Google Images and image packs. Good alt text describes what matters about the image in context and is concise; decorative images should carry an empty `alt=\"\"` so assistive technology skips them. Keyword-stuffed alt attributes are treated as a spam signal and help nobody.", "example": "`alt=\"Cross-section of a sourdough loaf showing an open crumb\"` rather than `alt=\"bread bread recipe sourdough best bread\"`.", "related": [ "image-pack", "image-sitemap", "semantic-html", "accessibility", "structured-data" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/google-images" }, { "id": "anchor-text", "slug": "anchor-text", "term": "Anchor text", "aka": [ "Link text" ], "category": "links", "shortDefinition": "The visible, clickable text of a hyperlink, used by search engines as a description of the destination page.", "definition": "Anchor text is one of the oldest relevance signals and still informs what a linked page is about, both externally and internally. Over-optimised exact-match anchors at scale are a classic link spam signature, which is why natural profiles are dominated by brand, URL and descriptive phrases. Internally, anchor text is fully under your control and is one of the cheapest relevance levers available — 'click here' wastes it.", "example": "Changing internal anchors from 'learn more' to 'commercial lease break clauses' clarifies the target page's topic for crawlers.", "related": [ "internal-linking", "backlink", "link-spam", "link-equity", "accessibility" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/links-crawlable" }, { "id": "answer-engine", "slug": "answer-engine", "term": "Answer engine", "aka": [ "AI search engine" ], "category": "geo-ai", "shortDefinition": "A system that responds to a question with a synthesised answer and citations rather than a ranked list of documents — for example Perplexity, ChatGPT search or Google's AI Mode.", "definition": "Answer engines combine a retrieval layer over web or licensed content with a language model that composes the response. From a publisher's perspective they change the unit of visibility from position to citation, compress many results into a few sources, and remove most of the click-through that funded the underlying content. They also differ from each other in index, crawler and citation behaviour, so visibility in one says little about the others.", "example": "The same question yields three sources in one assistant and a largely non-overlapping set of five in another.", "related": [ "aeo", "geo", "citation-share", "zero-click", "perplexitybot" ], "sourceUrl": null }, { "id": "aeo", "slug": "aeo", "term": "Answer Engine Optimization", "aka": [ "AEO" ], "category": "geo-ai", "shortDefinition": "Optimising content to be the source of a direct answer in features and assistants that answer questions rather than list links.", "definition": "AEO predates generative AI — featured snippets, People Also Ask and voice assistants were answer engines already — and now covers AI Overviews, chat assistants and in-product copilots. The common requirement is extractability: a clear question-shaped heading, a direct answer stated compactly and unambiguously near it, and supporting detail below. The trade-off is that answering well can satisfy the user without a click, so AEO is worth most where a citation drives consideration rather than immediate traffic.", "example": "A support page restructures each article to open with a two-sentence answer under a question heading, and starts appearing as the cited source in assistants.", "related": [ "geo", "featured-snippet", "zero-click", "answer-engine", "semantic-chunking" ], "sourceUrl": null }, { "id": "assisted-conversions", "slug": "assisted-conversions", "term": "Assisted conversions", "aka": [ "Path contribution" ], "category": "analytics", "shortDefinition": "Conversions in which a channel appeared somewhere in the path but was not the final touchpoint before conversion.", "definition": "The metric exposes channels that initiate or support journeys without closing them, which is the usual shape of informational organic content. In GA4 the equivalent views live in the advertising and attribution reports rather than a dedicated multi-channel funnels section. Assisted figures should be read alongside path length and time lag, since a channel appearing in every long path may simply be branded search happening anyway.", "example": "Informational blog pages rarely convert directly but appear in 28% of converting paths as the first touch.", "related": [ "attribution", "conversion-rate", "ga4", "incrementality", "seo-roi" ], "sourceUrl": "https://support.google.com/analytics/answer/10596866" }, { "id": "attribution", "slug": "attribution", "term": "Attribution", "aka": [ "Attribution modelling" ], "category": "analytics", "shortDefinition": "Assigning credit for a conversion across the touchpoints that preceded it, using rules such as last click, first click, or data-driven models.", "definition": "Organic search is systematically undervalued by last-click attribution because it frequently appears early in research and again as a branded final step that other channels created demand for. GA4 defaults to a data-driven model within its own lookback window, but cross-device and cross-browser gaps, consent loss and dark-social referrals all limit any model's accuracy. Where budget decisions are at stake, incrementality testing and media mix modelling are more defensible than any attribution model.", "example": "A campaign appears unprofitable on last click but is shown to drive a third of assisted conversions in a path analysis.", "related": [ "assisted-conversions", "incrementality", "ga4", "conversion-rate", "seo-roi" ], "sourceUrl": "https://support.google.com/analytics/answer/10596866" }, { "id": "author-authority", "slug": "author-authority", "term": "Author authority", "aka": [ "Author E-E-A-T", "Bylines" ], "category": "content", "shortDefinition": "The credibility a named author brings to content, evidenced by verifiable expertise, publication history and presence across trusted sources.", "definition": "Google retired rel=author markup and authorship rich results long ago, so there is no direct mechanism to declare an author for ranking. What remains is substantive: real bylines linking to detailed bio pages, Person structured data with `sameAs` links, credentials stated plainly, and a body of work findable elsewhere. Quality raters explicitly research who created content and their reputation, and generative systems attribute more readily to named, corroborated experts.", "example": "Each article links its byline to a bio page listing the author's qualifications, professional registration number and other publications.", "related": [ "eeat", "trust", "entity", "editorial-guidelines", "ymyl" ], "sourceUrl": "https://developers.google.com/search/docs/fundamentals/creating-helpful-content" }, { "id": "locale-redirects", "slug": "locale-redirects", "term": "Automatic locale redirection", "aka": [ "IP-based redirects", "Language detection" ], "category": "international", "shortDefinition": "Redirecting users to a country or language version based on IP address or browser settings, a pattern Google warns can prevent crawling of other versions.", "definition": "Because Googlebot crawls predominantly from US IP addresses, aggressive IP redirection can leave every non-US version undiscovered. Google's guidance is to avoid automatic redirection or, if used, to make it non-blocking and to provide clearly linked alternatives such as a banner or country selector. Redirecting based on `Accept-Language` has the same risk and additionally frustrates users whose browser language differs from their preference.", "example": "A site replaces a hard IP redirect with a dismissible banner suggesting the local version, and its non-US pages begin indexing.", "related": [ "hreflang", "x-default", "geo-targeting", "international-seo", "redirect-302" ], "sourceUrl": "https://developers.google.com/search/docs/specialty/international/managing-multi-regional-sites" }, { "id": "average-position", "slug": "average-position", "term": "Average position", "aka": [], "category": "analytics", "shortDefinition": "Search Console's mean of the highest position a site achieved for each impression of a query, averaged across those impressions.", "definition": "It is an average of tops, not of every listing, and it aggregates across devices, countries and personalisations unless filtered — so a reported 8.4 may mean position 3 in one country and 20 in another. AI Overviews occupy a single position shared by all their links. Because averages hide distribution, position work should use filtered segments and, where precision matters, a dedicated rank tracker with fixed parameters.", "example": "Filtering a query by country reveals an average position of 2.1 in the UK and 27 in the US, behind a blended figure of 9.", "related": [ "impressions", "position-tracking", "search-console", "click-through-rate", "ai-overviews" ], "sourceUrl": "https://support.google.com/webmasters/answer/7042828" }, { "id": "avif", "slug": "avif", "term": "AVIF", "aka": [ "AV1 Image File Format" ], "category": "technical", "shortDefinition": "A modern image format based on the AV1 codec that generally compresses better than WebP, with wide but slightly less universal browser support.", "definition": "AVIF supports HDR, wide colour gamut and alpha transparency, and frequently halves WebP file sizes at equivalent perceived quality, at the cost of slower encoding. It is deployed with a `` element listing AVIF first, WebP second and JPEG last, or via CDN negotiation. For LCP-critical hero images the byte savings often translate directly into passing the 2.5 s threshold on mobile networks.", "example": "`\"…\"`", "related": [ "webp", "largest-contentful-paint", "srcset", "cdn", "core-web-vitals" ], "sourceUrl": "https://web.dev/articles/compress-images-avif" }, { "id": "backlink", "slug": "backlink", "term": "Backlink", "aka": [ "Inbound link", "External link" ], "category": "links", "shortDefinition": "A hyperlink from one website to another, historically the strongest external signal search engines use to judge a page's importance and relevance.", "definition": "Links are how PageRank flows and how crawlers discover content, and their anchor text supplies topical context. Not all links count: search engines discount those they judge to be manipulative, paid without disclosure, or from low-quality sources, and nofollow-class attributes signal that the publisher does not vouch for the target. Google representatives have said links matter less than they once did, but they remain a core signal and a primary driver of discovery.", "example": "A university department cites a company's research page, and the page's rankings for related terms improve over the following weeks.", "related": [ "pagerank", "anchor-text", "referring-domains", "link-equity", "nofollow" ], "sourceUrl": "https://developers.google.com/search/docs/essentials/spam-policies" }, { "id": "bert", "slug": "bert", "term": "BERT", "aka": [ "Bidirectional Encoder Representations from Transformers" ], "category": "algorithms", "shortDefinition": "A transformer language model Google applied to search from 2019, improving understanding of word order, prepositions and context within queries and passages.", "definition": "BERT reads a sequence bidirectionally, so it captures how 'to' and 'from' change a query's meaning — the classic example being a traveller query where 'to' determines the direction of travel. Google said it initially affected a substantial share of English queries and later expanded to all languages. It also improved featured snippet selection. As with other AI systems there is nothing to optimise for beyond writing naturally and precisely.", "example": "'Can you get medicine for someone pharmacy' began returning results about collecting a prescription for another person rather than general pharmacy pages.", "related": [ "neural-matching", "rankbrain", "mum", "semantic-search", "passage-ranking" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/ranking-systems-guide" }, { "id": "bing-webmaster-tools", "slug": "bing-webmaster-tools", "term": "Bing Webmaster Tools", "aka": [ "BWT" ], "category": "analytics", "shortDefinition": "Microsoft's free webmaster platform for Bing, offering search performance data, index explorer, site scan, IndexNow submission and URL inspection.", "definition": "Bing Webmaster Tools reports clicks, impressions, average position and crawl information for Bing, and includes tools Google does not offer such as Index Explorer, keyword research with actual query volumes, and direct IndexNow submission. Its relevance grew as Bing's index underpins Copilot and, at various times, other AI answer products, making Bing coverage a factor in AI visibility. Verification is via XML file, meta tag, DNS record, or import from Search Console.", "example": "A publisher notices its pages are indexed in Google but missing from Bing, and uses Index Explorer to find a blocked directory.", "related": [ "indexnow", "search-console", "crawling", "ai-visibility", "impressions" ], "sourceUrl": "https://www.bing.com/webmasters/help/webmaster-guidelines-30fba23a" }, { "id": "brand-mention-monitoring", "slug": "brand-mention-monitoring", "term": "Brand mention monitoring", "aka": [ "Mention tracking" ], "category": "geo-ai", "shortDefinition": "Tracking where a brand is named across the web and inside AI answers, including whether the description is accurate and the sentiment favourable.", "definition": "In the AI era monitoring extends beyond PR value: what third-party pages say about a brand becomes the raw material assistants use to describe it, so an outdated pricing article or a wrong founding date propagates into answers. The workflow is to find the influential sources for your category, check what they assert, and correct or update them. Unlinked mentions are also link reclamation candidates and contribute to entity corroboration.", "example": "An assistant repeatedly states a company still offers a discontinued free tier, traced to one widely-cited comparison article.", "related": [ "unlinked-mentions", "hallucination", "entity", "ai-visibility", "co-citation" ], "sourceUrl": null }, { "id": "brand-search", "slug": "brand-search", "term": "Brand search", "aka": [ "Branded queries" ], "category": "analytics", "shortDefinition": "Searches containing a company or product name, which behave differently from non-brand queries and should always be reported separately.", "definition": "Brand queries convert far better and are largely a function of marketing activity elsewhere, so including them in SEO reporting inflates apparent performance and hides non-brand trends. Segmentation is done with regex filters on query data. Brand search volume is also the most practical proxy for demand created by other channels, and is the basis of share-of-search measurement.", "example": "Excluding brand terms reveals that non-brand clicks fell 18% while total clicks appeared flat.", "related": [ "share-of-search", "impressions", "click-through-rate", "share-of-voice", "seo-roi" ], "sourceUrl": null }, { "id": "breadcrumb", "slug": "breadcrumb", "term": "Breadcrumb navigation", "aka": [ "BreadcrumbList" ], "category": "technical", "shortDefinition": "A secondary navigation trail showing a page's position in the site hierarchy; with BreadcrumbList markup it can replace the URL line in search results.", "definition": "Breadcrumbs distribute internal link equity upward to category pages, give crawlers a second path into the hierarchy independent of the main navigation, and clarify site structure for users arriving from search. Marking them up with BreadcrumbList JSON-LD makes Google eligible to show the trail instead of the raw URL. The markup must match the visible breadcrumb.", "example": "`Home > Running > Trail Shoes > Model X` appears in the result instead of `example.com/p/12345`.", "related": [ "internal-linking", "structured-data", "rich-results", "site-architecture", "json-ld" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/structured-data/breadcrumb" }, { "id": "broken-link-building", "slug": "broken-link-building", "term": "Broken link building", "aka": [], "category": "links", "shortDefinition": "Finding dead outbound links on third-party pages and suggesting your equivalent resource as a replacement.", "definition": "The tactic works because it offers the publisher something useful — a fix for a broken page — rather than just asking for a favour. Yield is low and falling, since many resource pages are unmaintained and the replacement must genuinely match what was lost. It remains a reasonable way to find opportunities within a niche, particularly where an entire well-linked resource has gone offline.", "example": "A defunct government guidance page has 400 inbound links; a firm publishes an accurate archive-based replacement and contacts the linking sites.", "related": [ "link-building", "link-reclamation", "backlink", "http-status-codes", "digital-pr" ], "sourceUrl": null }, { "id": "gbp-categories", "slug": "gbp-categories", "term": "Business categories", "aka": [ "GBP categories", "Primary category" ], "category": "local", "shortDefinition": "The classification assigned to a Google Business Profile, with one primary and up to nine secondary categories, strongly determining which searches it can appear for.", "definition": "The primary category carries the most weight and should describe what the business principally is, not what it aspires to rank for. Google maintains a fixed taxonomy that changes over time, so periodic review is worthwhile, and category availability varies by country. Choosing a category that does not match reality risks both poor relevance and suspension.", "example": "A practice changes its primary category from 'Dentist' to 'Cosmetic dentist' and its pack visibility shifts toward cosmetic queries and away from general ones.", "related": [ "google-business-profile", "local-pack", "local-seo", "proximity", "local-reviews" ], "sourceUrl": "https://support.google.com/business/answer/3038177" }, { "id": "gbp-suspension", "slug": "gbp-suspension", "term": "Business Profile suspension", "aka": [ "GBP suspension" ], "category": "local", "shortDefinition": "Google disabling a business listing for policy violations such as a fake address, keyword-stuffed name, ineligible business type or prohibited category.", "definition": "Suspensions are either soft, where the listing remains but is unverified and loses visibility, or hard, where it is removed entirely. Common triggers are virtual offices, keyword stuffing in the business name, multiple listings at one address, and edits that trip automated review. Reinstatement requires evidence the business is real and eligible, typically signage, utility bills or registration documents.", "example": "A firm adds its service keywords to its profile name, is suspended within days, and must submit registration documents to be reinstated.", "related": [ "google-business-profile", "service-area-business", "nap", "multi-location-seo", "local-seo" ], "sourceUrl": "https://support.google.com/business/answer/4569145" }, { "id": "cache-control", "slug": "cache-control", "term": "Cache-Control", "aka": [ "Caching headers" ], "category": "technical", "shortDefinition": "The HTTP header governing how browsers, CDNs and proxies store and revalidate a response, using directives such as max-age, s-maxage and stale-while-revalidate.", "definition": "Hashed static assets should be served `public, max-age=31536000, immutable`, while HTML usually needs a short or zero max-age with revalidation so content changes propagate. `s-maxage` targets shared caches specifically, and `stale-while-revalidate` lets an edge serve a slightly stale copy while refreshing in the background, which is a cheap TTFB win. Misconfigured caching is a classic cause of a site appearing not to update for search engines.", "example": "Setting `Cache-Control: public, s-maxage=600, stale-while-revalidate=86400` on article HTML gives an edge-cached response in single-digit milliseconds.", "related": [ "etag", "cdn", "ttfb", "http2", "edge-seo" ], "sourceUrl": "https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Cache-Control" }, { "id": "canonicalization", "slug": "canonicalization", "term": "Canonicalization", "aka": [ "Duplicate consolidation" ], "category": "crawling-indexing", "shortDefinition": "The process by which a search engine groups duplicate or near-duplicate URLs into a cluster and picks one representative URL to index and rank.", "definition": "Search engines cluster URLs that serve substantially the same content, then select a canonical using signals including the rel=canonical annotation, redirects, internal and external links, sitemap inclusion, HTTPS preference and URL cleanliness. The rel=canonical tag is a strong hint, not a directive: Google frequently chooses a different canonical, which Search Console reports as 'Duplicate, Google chose different canonical than user'. Non-canonical URLs in a cluster generally do not receive their own ranking.", "example": "`/shoes?color=red`, `/shoes?color=red&utm_source=news` and `/shoes/red/` all resolve to the same page; Google indexes `/shoes/red/` and folds the others into it.", "related": [ "rel-canonical", "duplicate-content", "url-parameters", "redirect-301", "url-inspection-tool" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/canonicalization" }, { "id": "checkout-noindex", "slug": "checkout-noindex", "term": "Cart and checkout exclusion", "aka": [ "Transactional page exclusion" ], "category": "ecommerce", "shortDefinition": "Keeping cart, checkout, account and order confirmation URLs out of the index, since they have no search value and can leak session data.", "definition": "These URLs are typically parameterised, personalised and infinite in variety, so they consume crawl budget and generate index bloat. The correct treatment is noindex on the pages plus, where the space is unbounded, a robots.txt disallow — remembering that a disallowed URL cannot be seen to be noindexed. Order confirmation pages occasionally get indexed with customer data, which is a privacy incident as well as an SEO problem.", "example": "A site finds 1,200 order confirmation URLs indexed with order references, and applies noindex plus authentication.", "related": [ "noindex", "index-bloat", "robots-txt", "url-parameters", "crawl-budget" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/block-indexing" }, { "id": "category-page-content", "slug": "category-page-content", "term": "Category page content", "aka": [ "Category copy" ], "category": "ecommerce", "shortDefinition": "Editorial content on a product listing page that helps buyers choose, as distinct from keyword-stuffed text appended below the product grid.", "definition": "The legacy pattern of a hidden block of keyword text under the fold adds little for users and is easily recognised as written for search engines. Content that works answers the decision the shopper is making — differences between types, sizing, compatibility, what to spend — and is placed where it helps, often as a short intro plus a deeper section below. Linking to relevant guides from the category is often more effective than expanding the copy.", "example": "A 'baby car seats' category opens with a two-paragraph explanation of the group standards and links to a fitting guide.", "related": [ "plp", "thin-content", "internal-linking", "search-intent", "ecommerce-seo" ], "sourceUrl": "https://developers.google.com/search/docs/specialty/ecommerce" }, { "id": "causal-impact", "slug": "causal-impact", "term": "Causal impact analysis", "aka": [ "Interrupted time series", "CausalImpact" ], "category": "analytics", "shortDefinition": "A statistical method that estimates the effect of an intervention by forecasting what would have happened without it, using control series as a counterfactual.", "definition": "Implemented in Google's open-source CausalImpact package using Bayesian structural time-series models, it takes a pre-period, a set of correlated but unaffected control series, and an intervention date, then reports the difference between observed and predicted values with credible intervals. It is the standard tool for evaluating migrations, template changes and algorithm updates when a true control group is impossible. Results are only as good as the control series; using a control that was also affected invalidates the estimate.", "example": "After a migration, analysis using unaffected site sections as controls estimates a 9% loss with a credible interval excluding zero.", "related": [ "incrementality", "seo-testing", "site-migration", "forecasting", "algorithm-volatility" ], "sourceUrl": "https://google.github.io/CausalImpact/CausalImpact.html" }, { "id": "cctld", "slug": "cctld", "term": "ccTLD", "aka": [ "Country code top-level domain" ], "category": "international", "shortDefinition": "A top-level domain assigned to a country or territory, such as .de or .fr, which carries an inherent and unchangeable geo-targeting signal.", "definition": "ccTLDs are the strongest country-targeting signal and often carry local user trust, but they fragment authority across domains, multiply technical and operational overhead, and cannot be geo-targeted to a different country in Search Console. Some ccTLDs are treated as generic by Google because of their widespread non-geographic use — .co, .io and .tv among them. The choice is usually driven by brand and legal considerations as much as SEO.", "example": "A retailer runs example.de and example.fr as separate ccTLDs, accepting duplicated technical work for stronger local trust.", "related": [ "subfolder-vs-subdomain", "geo-targeting", "international-seo", "hreflang", "internationalized-domain-name" ], "sourceUrl": "https://developers.google.com/search/docs/specialty/international/managing-multi-regional-sites" }, { "id": "chatgpt-user", "slug": "chatgpt-user", "term": "ChatGPT-User", "aka": [ "User-triggered fetcher" ], "category": "geo-ai", "shortDefinition": "OpenAI's agent that fetches a specific page in real time because a user or an agentic task requested it, rather than for indexing or training.", "definition": "User-triggered fetches are one-off requests on behalf of a person, analogous to Google's user-triggered fetchers. Blocking them means a user who explicitly asks the assistant to read your page gets nothing, which is usually undesirable even for publishers who block training. Traffic from these agents shows in logs but rarely in analytics, since they typically do not execute the JavaScript that analytics depends on.", "example": "A user pastes a URL and asks for a summary; the request appears in server logs as ChatGPT-User seconds later.", "related": [ "gptbot", "oai-searchbot", "agentic-browsing", "ai-referral-traffic", "log-file-analysis" ], "sourceUrl": "https://platform.openai.com/docs/bots" }, { "id": "crux", "slug": "crux", "term": "Chrome User Experience Report", "aka": [ "CrUX", "Field data" ], "category": "technical", "shortDefinition": "Google's public dataset of real-user performance measurements from opted-in Chrome users, and the official source of Core Web Vitals assessment.", "definition": "CrUX aggregates a rolling 28-day window of metrics per origin and, where traffic is sufficient, per URL, exposed through BigQuery, an API, the CrUX dashboard, PageSpeed Insights and Search Console. Because it requires a popularity threshold, low-traffic URLs fall back to origin-level data or have none at all. The 28-day window means fixes take weeks to show up fully, which is why lab tools are used for iteration.", "example": "After a performance release, PageSpeed Insights still shows red field data for three weeks while the CrUX window rolls forward.", "related": [ "core-web-vitals", "pagespeed-insights", "lighthouse", "inp", "largest-contentful-paint" ], "sourceUrl": "https://developer.chrome.com/docs/crux" }, { "id": "chunk-retrieval", "slug": "chunk-retrieval", "term": "Chunk retrieval", "aka": [ "Passage retrieval" ], "category": "geo-ai", "shortDefinition": "Retrieving individual segments of a document rather than whole pages, which is how most AI search and RAG systems select what to feed a model.", "definition": "Because context windows are finite and precision matters, retrieval systems index documents as chunks and return the best-matching ones. A chunk that depends on context stated elsewhere on the page — a pronoun, an implicit subject, a figure whose units appear in a distant heading — can be retrieved and then misinterpreted. Writing sections that stand alone, restating key entities and units, is the practical response, and it also improves featured snippet and passage ranking outcomes.", "example": "A pricing table without a currency symbol in the same block is retrieved and summarised with the wrong currency.", "related": [ "semantic-chunking", "rag", "grounding", "passage-ranking", "embedding" ], "sourceUrl": null }, { "id": "citation-share", "slug": "citation-share", "term": "Citation share", "aka": [ "Source share" ], "category": "geo-ai", "shortDefinition": "The proportion of AI answer citations across a tracked prompt set that point to a given domain, used as the AI-era analogue of ranking share.", "definition": "Citation share counts linked source attributions rather than mentions, which makes it more objective but narrower — a brand can be recommended by name without being cited. It is calculated per assistant, since citation behaviour and source pools differ sharply between them. Interpreting it requires care: a high share on a narrow prompt set means little, and citations to third-party review sites often matter more to a brand than citations to its own domain.", "example": "Across 200 prompts, a vendor's domain is cited in 6% of responses while a comparison site it appears on is cited in 34%.", "related": [ "ai-visibility", "llm-share-of-voice", "answer-engine", "brand-mention-monitoring", "link-bait" ], "sourceUrl": null }, { "id": "claudebot", "slug": "claudebot", "term": "ClaudeBot", "aka": [ "Claude-Web", "Claude-User", "Claude-SearchBot" ], "category": "geo-ai", "shortDefinition": "Anthropic's family of web crawlers and fetchers, including a general crawler and agents that retrieve pages for search results and for individual user requests.", "definition": "Anthropic documents its user agents and publishes IP ranges for verification, and states its crawlers respect robots.txt directives. As with OpenAI, the agents serve different purposes — general crawling, search indexing to support citations, and user-initiated retrieval — so a site owner should decide per agent rather than blocking the family. Blocking search-oriented agents removes the site from citation in Claude's answers.", "example": "A documentation site allows all Anthropic agents and sees ClaudeBot fetches spike after a release announcement.", "related": [ "ai-crawler", "gptbot", "perplexitybot", "model-training-opt-out", "robots-txt" ], "sourceUrl": "https://support.anthropic.com/en/articles/8896518-does-anthropic-crawl-data-from-the-web-and-how-can-site-owners-block-the-crawler" }, { "id": "click-through-rate", "slug": "click-through-rate", "term": "Click-through rate", "aka": [ "CTR" ], "category": "analytics", "shortDefinition": "Clicks divided by impressions, expressed as a percentage; in search it measures how compelling a result is at the position and in the SERP context it occupies.", "definition": "CTR varies enormously by position, query intent, brand recognition and SERP composition, so raw comparisons across queries are meaningless — it must be assessed against a position-adjusted expectation. Falling CTR at a stable position usually signals a SERP layout change, a snippet rewrite, or a competitor with a stronger title. Published CTR-by-position curves are averages across wildly different query types and should be treated as rough context only.", "example": "A page holding position three sees CTR drop from 12% to 5% after a shopping carousel is added above the organic results.", "related": [ "impressions", "average-position", "title-tag", "meta-description", "zero-click" ], "sourceUrl": "https://support.google.com/webmasters/answer/7042828" }, { "id": "cloaking", "slug": "cloaking", "term": "Cloaking", "aka": [], "category": "crawling-indexing", "shortDefinition": "Presenting different content or URLs to search engines than to users with the intent to manipulate rankings; a Google spam policy violation.", "definition": "Cloaking is defined by intent and by material difference: serving text to crawlers that users never see, redirecting bots and humans to different destinations, or detecting Googlebot's IP range to swap content. Legitimate variation — personalisation, A/B tests with proper canonicalisation, geolocation redirects applied equally to users and crawlers, paywalls implemented with the flexible-sampling and paywalled-content structured data — is not cloaking. Enforcement can result in a manual action or removal from the index.", "example": "A hacked site serves normal content to visitors but pharmacy spam to Googlebot, detectable only via the URL Inspection live test.", "related": [ "dynamic-rendering", "manual-action", "spam-policies", "doorway-pages", "url-inspection-tool" ], "sourceUrl": "https://developers.google.com/search/docs/essentials/spam-policies" }, { "id": "co-citation", "slug": "co-citation", "term": "Co-citation and co-occurrence", "aka": [], "category": "links", "shortDefinition": "The pattern of two entities being mentioned or linked together by third parties, which helps search engines infer that they are related.", "definition": "Co-citation is when two pages are cited by the same source; co-occurrence is when two terms or brands appear together in text without a link. Both contribute to entity association: repeatedly appearing in the same 'best tools for X' lists as established competitors helps a search engine and a language model place a brand in that category. It is one of the mechanisms behind why category-list coverage matters for AI visibility even without links.", "example": "A new tool appears in six roundups alongside three established competitors, and starts surfacing when users ask for alternatives to them.", "related": [ "entity", "unlinked-mentions", "brand-mention-monitoring", "citation-share", "backlink" ], "sourceUrl": null }, { "id": "commercial-investigation", "slug": "commercial-investigation", "term": "Commercial investigation intent", "aka": [ "Commercial intent" ], "category": "content", "shortDefinition": "Queries from users comparing options before buying — 'best', 'vs', 'alternatives', 'review' — sitting between informational and transactional intent.", "definition": "These queries carry high commercial value because the user has decided to buy but not what to buy, and the SERPs are dominated by comparison content, reviews and roundups. Google's product review guidance applies here: content should show evidence of actual use, quantitative measurement and comparison against alternatives. It is also the segment most disrupted by AI answers, which summarise comparisons directly.", "example": "'Best CRM for small business' returns nine roundups and one vendor comparison page, with an AI Overview summarising three of them.", "related": [ "search-intent", "product-reviews-system", "ai-overviews", "affiliate-seo", "serp-intent-match" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/ranking-systems-guide" }, { "id": "competitive-analysis", "slug": "competitive-analysis", "term": "Competitive analysis", "aka": [ "Competitor research" ], "category": "business", "shortDefinition": "Comparing a site's search position against competitors across visibility, content coverage, authority and technical capability to locate realistic opportunities.", "definition": "Search competitors are often not business competitors — publishers, marketplaces and comparison sites frequently occupy the SERPs a brand wants. Useful analysis identifies where competitors are winning, why structurally, and which of those advantages are replicable. Gap analysis at keyword level is the common starting point, but template and architecture comparison usually explains more of the difference.", "example": "Analysis shows a competitor's advantage rests on 3,000 indexable facet pages the brand's platform cannot currently produce.", "related": [ "share-of-voice", "organic-tam", "keyword-research", "seo-audit", "seo-strategy" ], "sourceUrl": null }, { "id": "compression", "slug": "compression", "term": "Compression (gzip, Brotli, Zstandard)", "aka": [ "Content-Encoding" ], "category": "technical", "shortDefinition": "Encoding text responses to reduce transfer size; Brotli typically beats gzip on HTML, CSS and JavaScript, with Zstandard now also supported by browsers.", "definition": "Compression is negotiated via `Accept-Encoding` and `Content-Encoding` headers and applies to text formats; already-compressed formats such as WebP, AVIF and WOFF2 should not be re-compressed. Static assets can use maximum Brotli levels at build time, while dynamic HTML uses a lower level to keep CPU cost down. Missing compression on HTML is one of the fastest wins in any technical audit.", "example": "Enabling Brotli reduces a 180 KB HTML document to 22 KB over the wire.", "related": [ "ttfb", "cache-control", "cdn", "largest-contentful-paint", "http2" ], "sourceUrl": "https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Content-Encoding" }, { "id": "consent-mode", "slug": "consent-mode", "term": "Consent management and Consent Mode", "aka": [ "CMP", "Consent Mode v2" ], "category": "analytics", "shortDefinition": "Systems for capturing and honouring users' tracking permissions, and Google's mechanism for adjusting tag behaviour and modelling gaps when consent is withheld.", "definition": "Under GDPR-style regimes a meaningful share of visitors decline analytics cookies, creating a structural undercount that varies by country. Consent Mode sends cookieless pings for non-consenting users so behavioural modelling can estimate the gap, and Consent Mode v2 added parameters required for advertising features in the EEA. The practical consequence for SEO reporting is that analytics traffic figures are always lower than server-side reality, and the ratio drifts as consent rates change.", "example": "A site's consent rate falls from 78% to 61% after a banner redesign, and organic sessions appear to drop with no ranking change.", "related": [ "ga4", "sessions", "data-sampling", "server-side-tracking", "attribution" ], "sourceUrl": "https://support.google.com/analytics/answer/9976101" }, { "id": "content-audit", "slug": "content-audit", "term": "Content audit", "aka": [ "Content inventory" ], "category": "content", "shortDefinition": "A systematic review of every page on a site against performance, quality and business value criteria, producing a keep/improve/merge/remove decision per URL.", "definition": "The audit joins a full crawl with Search Console, analytics, conversion and backlink data so decisions rest on evidence rather than opinion. Beyond pruning, it surfaces cannibalisation, orphans, outdated claims and gaps against the topic map. On sites over a few thousand URLs it is normally run at template level first, since most quality problems are template-wide rather than page-specific.", "example": "An audit of 4,200 URLs finds 61% have received no clicks in twelve months, of which half are near-duplicate tag archives.", "related": [ "content-pruning", "content-decay", "keyword-cannibalization", "index-bloat", "orphan-page" ], "sourceUrl": null }, { "id": "content-brief", "slug": "content-brief", "term": "Content brief", "aka": [ "SEO brief" ], "category": "content", "shortDefinition": "A specification given to a writer covering the target query set, intent, required subtopics, angle, sources, internal links and success criteria.", "definition": "A good brief encodes the SERP analysis so the writer does not have to redo it, and states what the piece must contribute that competing pages do not. Weak briefs are keyword lists with word counts, which produce interchangeable content. Briefs should also name the internal links in and out, so the page enters the site's link graph properly from day one.", "example": "A brief specifies three expert quotes, the two datasets to cite, five subtopics the top results all miss, and four internal links to add.", "related": [ "information-gain", "serp-intent-match", "internal-linking", "editorial-guidelines", "topic-cluster" ], "sourceUrl": null }, { "id": "content-decay", "slug": "content-decay", "term": "Content decay", "aka": [ "Traffic decay" ], "category": "content", "shortDefinition": "The gradual decline in a page's search traffic as it ages, competitors update, intent shifts or the underlying information goes out of date.", "definition": "Decay is normal and predictable, and is best managed by monitoring page-level clicks over rolling periods to catch declines before they become collapses. Causes differ: some pages lose relevance because facts changed, others because the SERP added features that absorb the clicks, others because a competitor published something better. The response is a substantive refresh, a merge, or retirement — not a date change.", "example": "A 2023 pricing comparison loses 70% of its clicks over eight months as every listed price becomes wrong.", "related": [ "content-refresh", "content-pruning", "freshness", "click-through-rate", "position-tracking" ], "sourceUrl": null }, { "id": "cdn", "slug": "cdn", "term": "Content Delivery Network", "aka": [ "CDN" ], "category": "technical", "shortDefinition": "A distributed network of edge servers that caches and serves content close to users, reducing latency and offloading the origin.", "definition": "For SEO the benefits are lower TTFB and higher crawl capacity, plus image optimisation and modern protocol support at the edge. The risks are cache misconfiguration serving stale or wrong-locale content, edge rules that block crawler user agents, and geo-routing that shows crawlers a different country's content. Bot management rules that challenge unknown agents can also silently block legitimate AI crawlers a publisher wants to allow.", "example": "A bot-protection rule starts returning 403 to ClaudeBot and GPTBot, cutting off the site's citations in AI assistants until it is allowlisted.", "related": [ "ttfb", "cache-control", "edge-seo", "claudebot", "gptbot" ], "sourceUrl": null }, { "id": "content-pruning", "slug": "content-pruning", "term": "Content pruning", "aka": [ "Content consolidation" ], "category": "content", "shortDefinition": "Systematically removing, merging or noindexing low-value pages so a site's indexed footprint reflects only content worth ranking.", "definition": "Pruning starts with an inventory joining crawl data, Search Console impressions and clicks, analytics and conversions, then classifies each URL as keep, improve, merge or remove. Merged pages should 301 to their successor; genuinely dead pages should return 410 or 404. The aim is not smaller for its own sake but concentrating link equity and quality signals, and reducing the share of a site that looks unhelpful.", "example": "An agency merges 34 overlapping posts about the same tax credit into three definitive guides and redirects the rest.", "related": [ "content-decay", "index-bloat", "thin-content", "keyword-cannibalization", "helpful-content" ], "sourceUrl": null }, { "id": "content-refresh", "slug": "content-refresh", "term": "Content refresh", "aka": [ "Content update" ], "category": "content", "shortDefinition": "Substantively updating an existing page — new data, corrected facts, restructured sections — to restore or improve its relevance and rankings.", "definition": "A real refresh changes the content, not just the visible date; Google has repeatedly warned against artificially updating dates to appear fresh. The most effective refreshes re-check the SERP for intent drift, add information the current top results lack, remove obsolete sections, and update internal links. Refreshing should be prioritised by opportunity: pages with impressions but falling CTR or position usually pay back faster than new content.", "example": "A 'best project management tools' post is rebuilt with 2026 pricing, three new entrants removed or added, and fresh screenshots.", "related": [ "content-decay", "freshness", "serp-intent-match", "content-pruning", "information-gain" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/publication-dates" }, { "id": "content-syndication", "slug": "content-syndication", "term": "Content syndication", "aka": [ "Republishing" ], "category": "content", "shortDefinition": "Republishing the same content on third-party sites for reach; without correct canonicalisation it creates duplicates that can outrank the original.", "definition": "Google's guidance is that syndication partners should block the syndicated copy from indexing or use rel=canonical pointing at the original, because Google will otherwise choose whichever version it judges most authoritative — often the larger partner. Syndication is not a spam policy violation in itself, but syndicated content that adds nothing can weigh on the receiving site's quality. Partial excerpts with a link back are the lowest-risk pattern.", "example": "A trade publication republishes a vendor's research with `rel=canonical` back to the vendor's original post.", "related": [ "duplicate-content", "rel-canonical", "noindex", "digital-pr", "site-reputation-abuse" ], "sourceUrl": "https://developers.google.com/search/docs/essentials/spam-policies" }, { "id": "context-window", "slug": "context-window", "term": "Context window", "aka": [ "Context length" ], "category": "geo-ai", "shortDefinition": "The maximum amount of text a language model can consider at once, which bounds how much retrieved content can be supplied when generating an answer.", "definition": "Even with large windows, retrieval systems are selective because relevance degrades and cost rises with volume, so only a handful of passages usually reach the model. This is why concise, self-contained, information-dense sections outperform sprawling pages in retrieval contexts: they fit, and they survive selection. Window size also constrains how much of a long document an agent will actually read before answering.", "example": "A 12,000-word guide is represented in an answer by two retrieved sections; the rest never reaches the model.", "related": [ "chunk-retrieval", "rag", "semantic-chunking", "knowledge-cutoff", "grounding" ], "sourceUrl": null }, { "id": "conversion-rate", "slug": "conversion-rate", "term": "Conversion rate", "aka": [ "CVR" ], "category": "analytics", "shortDefinition": "The proportion of sessions or users that complete a defined goal, such as a purchase, lead form submission or sign-up.", "definition": "The denominator matters: session-based and user-based rates differ substantially, as do rates calculated per landing page versus site-wide. In SEO analysis conversion rate should be segmented by query intent, because informational and transactional traffic convert at incomparable rates and blending them hides both. Micro-conversions are useful for pages whose role is assistance rather than closing.", "example": "A category page converts at 3.1% while a guide converts at 0.2%, but the guide contributes a quarter of assisted conversions.", "related": [ "assisted-conversions", "attribution", "ga4", "search-intent", "seo-roi" ], "sourceUrl": "https://support.google.com/analytics/answer/12844695" }, { "id": "core-update", "slug": "core-update", "term": "Core update", "aka": [ "Broad core algorithm update" ], "category": "algorithms", "shortDefinition": "A significant, broad change to Google's ranking systems, announced in advance and rolled out over one to several weeks, typically a few times a year.", "definition": "Core updates are not penalties and do not target specific sites; Google describes them as reassessments of how well pages serve their purpose, so a site can drop without doing anything wrong. Rollouts take days to weeks and rankings can fluctuate throughout, so analysis should wait until Google confirms completion. Google's stated recovery advice is to focus on the helpful content self-assessment rather than looking for a specific fix, and recoveries typically occur at a later update.", "example": "The March 2026 core update rolled out from 27 March to 8 April 2026, with a second core update following in May 2026.", "related": [ "helpful-content-update", "algorithm-volatility", "spam-updates", "helpful-content", "quality-rater-guidelines" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/core-updates" }, { "id": "core-web-vitals", "slug": "core-web-vitals", "term": "Core Web Vitals", "aka": [ "CWV" ], "category": "technical", "shortDefinition": "Google's set of three field-measured user experience metrics — LCP, INP and CLS — used as part of the page experience signals in ranking.", "definition": "Core Web Vitals measure loading (Largest Contentful Paint), interactivity (Interaction to Next Paint) and visual stability (Cumulative Layout Shift). A URL is judged 'good' when the 75th percentile of real-user samples is at or below 2.5 s LCP, 200 ms INP and 0.1 CLS, evaluated separately for mobile and desktop. Interaction to Next Paint replaced First Input Delay as the responsiveness metric in March 2024. Google updates the metric set on an announced, roughly annual cadence.", "example": "A retailer's mobile LCP sits at 3.8 s at p75; compressing the hero image to AVIF and preloading it brings it under 2.5 s.", "related": [ "largest-contentful-paint", "inp", "cumulative-layout-shift", "crux", "page-experience" ], "sourceUrl": "https://web.dev/articles/vitals" }, { "id": "crawl-budget", "slug": "crawl-budget", "term": "Crawl budget", "aka": [], "category": "crawling-indexing", "shortDefinition": "The number of URLs a search engine is willing and able to crawl on a site in a given period, set by crawl capacity limit and crawl demand.", "definition": "Crawl budget is the intersection of what a host can serve without degrading (crawl capacity limit) and what the search engine wants (crawl demand, driven by popularity, staleness and perceived quality). Google states that most sites under a few thousand URLs never need to think about it; it becomes a real constraint for large e-commerce, classifieds, and sites generating many low-value URLs. The practical levers are removing duplicate and faceted URL spaces, fixing slow responses, and returning accurate status codes.", "example": "A marketplace with 4 million filter-generated URLs sees Googlebot spend 70% of its fetches on parameter URLs, starving new product pages of crawls.", "related": [ "crawl-rate-limit", "crawl-demand", "faceted-navigation", "log-file-analysis", "index-bloat" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/large-site-managing-crawl-budget" }, { "id": "crawl-demand", "slug": "crawl-demand", "term": "Crawl demand", "aka": [], "category": "crawling-indexing", "shortDefinition": "How much a search engine wants to crawl a set of URLs, driven by their popularity, how often they change, and the site's perceived overall quality.", "definition": "Even when a server could serve far more requests, Google will not crawl URLs it does not value. Demand rises for URLs that are linked to, that appear in sitemaps with plausible lastmod dates, that historically change, and that sit on sites Google rates highly. Demand collapses for stale, duplicative or low-quality URL spaces — which is why crawl problems on large sites are usually quality problems, not server problems.", "example": "A blog that publishes daily sees its homepage crawled hourly while a five-year-old tag archive is crawled twice a year.", "related": [ "crawl-budget", "crawl-rate-limit", "freshness", "index-bloat", "xml-sitemap" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/large-site-managing-crawl-budget" }, { "id": "crawl-rate-limit", "slug": "crawl-rate-limit", "term": "Crawl rate limit", "aka": [ "Crawl capacity limit" ], "category": "crawling-indexing", "shortDefinition": "The maximum simultaneous connections and fetch frequency a crawler will use for a host, adjusted automatically based on server response times and errors.", "definition": "Google raises crawl capacity when a host responds quickly and lowers it when responses slow down or return 5xx and 429 errors. The old manual crawl-rate setting in Search Console was retired; capacity is now managed automatically, though site owners can signal overload with 503/429 responses. Because capacity is per host, sharing infrastructure across many hostnames can concentrate or dilute crawling.", "example": "After a database issue pushes median TTFB from 200 ms to 3 s, Googlebot's daily fetches drop by half until performance recovers.", "related": [ "crawl-budget", "crawl-demand", "ttfb", "http-status-codes", "log-file-analysis" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/large-site-managing-crawl-budget" }, { "id": "crawl-trap", "slug": "crawl-trap", "term": "Crawl trap", "aka": [ "Infinite space", "Spider trap" ], "category": "crawling-indexing", "shortDefinition": "A URL pattern that generates effectively unlimited unique URLs, causing crawlers to loop indefinitely and exhaust crawl budget.", "definition": "Classic examples are calendars with next-month links extending forever, relative-path bugs producing `/a/b/a/b/a/`, session ids in URLs, and unbounded filter combinations. Traps are visible in logs as crawl volume rising without indexation rising. Fixes include robots.txt patterns, nofollow on generative controls, bounding the generated range, and returning 404 outside sensible limits.", "example": "An events calendar lets Googlebot walk to the year 2147 one month at a time, generating hundreds of thousands of empty pages.", "related": [ "crawl-budget", "faceted-navigation", "index-bloat", "robots-txt", "log-file-analysis" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/large-site-managing-crawl-budget" }, { "id": "crawled-currently-not-indexed", "slug": "crawled-currently-not-indexed", "term": "Crawled — currently not indexed", "aka": [], "category": "crawling-indexing", "shortDefinition": "A Search Console status meaning Google fetched a URL but chose not to index it, usually a quality, duplication or value judgement.", "definition": "Unlike 'Discovered', the content has been seen and rejected or deferred. Frequent causes are near-duplicate templated content, thin pages, doorway-like location or variant pages, and pages that add nothing beyond what is already indexed elsewhere. Resubmitting via URL Inspection rarely helps; improving the page's distinctiveness and internal link support does.", "example": "Twelve near-identical 'plumber in [city]' pages are all crawled, but Google indexes only the two with genuinely local content.", "related": [ "discovered-currently-not-indexed", "thin-content", "doorway-pages", "duplicate-content", "helpful-content" ], "sourceUrl": "https://support.google.com/webmasters/answer/7440203" }, { "id": "crawling", "slug": "crawling", "term": "Crawling", "aka": [ "Spidering" ], "category": "crawling-indexing", "shortDefinition": "The process by which a search engine discovers URLs and downloads their content using automated software, before any indexing or ranking happens.", "definition": "A crawler starts from known URLs — sitemaps, previously crawled pages, links — and follows references to discover more. Each fetch is subject to robots.txt rules, host load limits and the crawler's own scheduling priorities. Crawling only obtains bytes; whether those bytes are stored and made retrievable is a separate decision made during indexing. A page that is never crawled can still be indexed URL-only if it is linked to, but Google will have no content to describe it with.", "example": "A news site publishes a new article, pings its sitemap, and Googlebot fetches the URL within minutes; the article is crawled but does not appear in search until it is also indexed.", "related": [ "indexing", "googlebot", "robots-txt", "crawl-budget", "xml-sitemap" ], "sourceUrl": "https://developers.google.com/search/docs/fundamentals/how-search-works" }, { "id": "cumulative-layout-shift", "slug": "cumulative-layout-shift", "term": "Cumulative Layout Shift", "aka": [ "CLS" ], "category": "technical", "shortDefinition": "A Core Web Vital scoring unexpected visual movement of page content, computed from the largest burst of layout shifts; good is 0.1 or less at p75.", "definition": "Each shift is scored as impact fraction times distance fraction, and CLS reports the worst session window of shifts within five seconds. Shifts caused within 500 ms of a user interaction are excluded. The usual causes are images and iframes without dimensions, injected banners and ads, late-loading web fonts causing reflow, and content inserted above existing content.", "example": "A cookie banner injected 1.2 s after load pushes the article down by 90 px, producing a 0.28 CLS until it is reserved space in the layout.", "related": [ "core-web-vitals", "font-loading", "third-party-scripts", "largest-contentful-paint", "crux" ], "sourceUrl": "https://web.dev/articles/cls" }, { "id": "data-sampling", "slug": "data-sampling", "term": "Data sampling and thresholding", "aka": [ "Sampling" ], "category": "analytics", "shortDefinition": "Analytics platforms estimating results from a subset of data, or suppressing small values for privacy, both of which distort reporting if unnoticed.", "definition": "GA4 applies sampling in some exploration and ad-hoc queries above data thresholds, and applies thresholding that hides rows when they could identify individuals. Search Console applies its own filtering and anonymises rare queries, so query-level totals never sum to the reported overall figure. The remedies are BigQuery export for analytics, the Search Console API for search data, and always checking for the sampling indicator before quoting a number.", "example": "An exploration reports 62% sampling, and the same query run against BigQuery returns figures 9% different.", "related": [ "ga4", "search-console", "sessions", "impressions", "incrementality" ], "sourceUrl": "https://support.google.com/analytics/answer/13331684" }, { "id": "data-nosnippet", "slug": "data-nosnippet", "term": "data-nosnippet", "aka": [], "category": "crawling-indexing", "shortDefinition": "An HTML attribute that excludes a specific span, div or section of a page from being used in Google's snippets while leaving the rest eligible.", "definition": "Applied as a boolean attribute on span, div or section elements, data-nosnippet gives paragraph-level control instead of the page-level control of nosnippet. It is useful for hiding boilerplate, prices that change, or licensed text from snippets and generative summaries without suppressing the whole page. It affects presentation only — the content is still crawled and indexed for ranking.", "example": "`Members only: current price €39` keeps a member price out of the SERP snippet.", "related": [ "nosnippet", "robots-meta-tag", "featured-snippet", "ai-overviews", "structured-data" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/robots-meta-tag" }, { "id": "deduplication", "slug": "deduplication", "term": "Deduplication system", "aka": [ "Duplicate filtering" ], "category": "algorithms", "shortDefinition": "Google's system for ensuring results are not dominated by near-identical pages, filtering duplicates so users see a varied set of sources.", "definition": "Listed among Google's ranking systems, deduplication operates both at the clustering stage, where duplicate URLs are consolidated to a canonical, and at the result stage, where similar pages are filtered from the SERP. It is why syndicated copies of the same article rarely appear together, and why near-duplicate internal pages simply do not both surface. It also interacts with the site diversity system.", "example": "A press release published by twelve outlets shows only one or two versions in the results, with the rest filtered.", "related": [ "duplicate-content", "canonicalization", "site-diversity", "content-syndication", "ranking-systems" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/ranking-systems-guide" }, { "id": "digital-pr", "slug": "digital-pr", "term": "Digital PR", "aka": [ "SEO PR" ], "category": "links", "shortDefinition": "Earning editorial coverage and links from media outlets through newsworthy assets — original data, expert commentary, campaigns — rather than link placement.", "definition": "Digital PR sits at the intersection of press relations and SEO: the deliverable is coverage in publications journalists actually read, which produces links, brand mentions and, increasingly, the kind of third-party corroboration that AI systems cite. Effective programmes start from a genuine story angle and a dataset or expert nobody else has. Measurement should include unlinked mentions and share of voice, not only followed links.", "example": "An insurer analyses its own claims data by region and lands coverage across national and trade press, earning links from 40 domains.", "related": [ "link-building", "unlinked-mentions", "haro-sourcing", "brand-mention-monitoring", "link-bait" ], "sourceUrl": null }, { "id": "disavow-tool", "slug": "disavow-tool", "term": "Disavow links tool", "aka": [ "Disavow file" ], "category": "links", "shortDefinition": "A Search Console tool for asking Google to ignore specified inbound links or domains, intended for sites with a link-related manual action or known paid links.", "definition": "The tool accepts a plain-text file of URLs and `domain:` entries. Google's own guidance is that most sites never need it, because its systems already discount untrusted links, and that careless use can hurt by disavowing links that were passing value. It is not a general hygiene routine: the legitimate use cases are recovering from a manual action or pre-emptively disclosing links you or a previous agency paid for.", "example": "After buying links, a new owner uploads a disavow file listing the 60 purchased domains and files a reconsideration request.", "related": [ "toxic-links", "manual-action", "link-spam", "negative-seo", "search-console" ], "sourceUrl": "https://support.google.com/webmasters/answer/2648487" }, { "id": "discontinued-products", "slug": "discontinued-products", "term": "Discontinued product handling", "aka": [ "End-of-life products" ], "category": "ecommerce", "shortDefinition": "The policy for permanently retired products: keep as an archive page, redirect to a genuine successor, or return 404/410 — chosen by whether an equivalent exists.", "definition": "Redirecting to a successor product is right when one genuinely replaces it; redirecting everything to the category is a soft 404 pattern. Where a product has accumulated links and search demand — discontinued models people still research — keeping an informative archive page with specifications and successor links preserves the value. Where nothing equivalent exists and no demand remains, 410 is the cleanest signal.", "example": "A discontinued camera keeps its page with a 'replaced by' link and full specifications, retaining its review links and long-tail traffic.", "related": [ "out-of-stock-handling", "http-status-codes", "redirect-301", "soft-404", "content-pruning" ], "sourceUrl": "https://developers.google.com/search/docs/specialty/ecommerce" }, { "id": "discovered-currently-not-indexed", "slug": "discovered-currently-not-indexed", "term": "Discovered — currently not indexed", "aka": [], "category": "crawling-indexing", "shortDefinition": "A Search Console status meaning Google knows a URL exists but has not crawled it yet, typically because of crawl capacity or low crawl demand.", "definition": "Google reports this when a URL was found (via sitemap or link) and queued but deliberately deferred. It usually signals either that the site is slow enough that Google is throttling, or that Google does not consider the URL worth fetching. On large sites it is the clearest symptom of a crawl-demand problem: the remedy is fewer, better URLs, stronger internal links to the affected templates, and faster responses.", "example": "A 500,000-SKU catalogue shows 180,000 URLs in 'Discovered — currently not indexed' while its top-selling 20,000 products index fine.", "related": [ "crawled-currently-not-indexed", "crawl-demand", "crawl-budget", "url-inspection-tool", "internal-linking" ], "sourceUrl": "https://support.google.com/webmasters/answer/7440203" }, { "id": "discussions-and-forums", "slug": "discussions-and-forums", "term": "Discussions and forums", "aka": [ "Forum results" ], "category": "serp-features", "shortDefinition": "A SERP feature surfacing threads from forums, Q&A sites and social communities, supported by DiscussionForumPosting and QAPage structured data.", "definition": "Introduced as part of Google's push toward first-hand perspectives, the feature elevates community discussion for queries where personal experience is valuable. Sites hosting genuine discussion can mark threads up with `DiscussionForumPosting`, including author, date and interaction counts, or `QAPage` for question-and-answer pages. The feature contributed to the visibility of large communities such as Reddit in many result sets.", "example": "A cycling forum's thread about a specific frame creak appears in the discussions block above several manufacturer pages.", "related": [ "structured-data", "serp", "rich-results", "eeat", "ugc-attribute" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/structured-data/discussion-forum" }, { "id": "dom", "slug": "dom", "term": "DOM (Document Object Model)", "aka": [ "Rendered HTML" ], "category": "crawling-indexing", "shortDefinition": "The in-memory tree representation of a document that the browser builds from HTML and mutates with JavaScript; the rendered DOM is what search engines index.", "definition": "The source HTML delivered by the server and the DOM after scripts run can differ dramatically — tags injected, canonicals rewritten, content replaced. For SEO the practical rule is that the rendered DOM wins, but only for engines that render; anything relying on scripts is invisible to non-rendering agents. Comparing view-source HTML with the rendered DOM in URL Inspection is the standard diagnostic for JavaScript indexing problems.", "example": "View-source shows no meta description, but the rendered DOM contains one injected by a tag manager — Google may use it, while a simple HTML-only crawler will not.", "related": [ "javascript-rendering", "render-queue", "hydration", "prerendering", "url-inspection-tool" ], "sourceUrl": "https://developer.mozilla.org/en-US/docs/Web/API/Document_Object_Model/Introduction" }, { "id": "doorway-pages", "slug": "doorway-pages", "term": "Doorway pages", "aka": [ "Gateway pages", "Bridge pages" ], "category": "content", "shortDefinition": "Pages or sites created to rank for many similar queries and funnel users to a single destination, providing no unique value themselves.", "definition": "Google's spam policies name doorways explicitly: multiple domains or pages targeting slight query variations, near-duplicate city or region pages, and pages that funnel users to an intermediate page rather than the content they wanted. The test is whether each page is genuinely useful in its own right to the person who lands on it. Templated local landing pages cross into doorway territory when only the place name changes.", "example": "Four hundred pages reading 'emergency plumber in [town]' with identical copy and a shared phone number.", "related": [ "thin-content", "local-landing-pages", "scaled-content-abuse", "spam-policies", "programmatic-seo" ], "sourceUrl": "https://developers.google.com/search/docs/essentials/spam-policies" }, { "id": "duplicate-content", "slug": "duplicate-content", "term": "Duplicate content", "aka": [], "category": "crawling-indexing", "shortDefinition": "Substantively identical or near-identical content available at more than one URL, within a site or across sites.", "definition": "There is no general 'duplicate content penalty'; the normal outcome is consolidation, where one URL is chosen to represent the cluster and the others simply do not rank. Problems arise when the engine picks the wrong canonical, when link signals are split across variants, or when duplication is scaled deliberately, which can cross into scraped or scaled content abuse. Common internal causes are protocol and www variants, trailing slashes, parameters, print views and syndicated boilerplate.", "example": "Both `http://example.com/page` and `https://www.example.com/page` resolve with 200 status, so Google must guess which to index.", "related": [ "canonicalization", "rel-canonical", "url-parameters", "scaled-content-abuse", "content-syndication" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/consolidate-duplicate-urls" }, { "id": "dynamic-rendering", "slug": "dynamic-rendering", "term": "Dynamic rendering", "aka": [], "category": "crawling-indexing", "shortDefinition": "Serving a server-rendered HTML version to crawlers and a client-rendered version to users; Google now describes it as a deprecated workaround.", "definition": "Dynamic rendering detects crawler user agents and routes them to a prerendering service. Google explicitly labels it a workaround rather than a long-term solution because it is brittle, doubles the surfaces you must keep in sync, and sits uncomfortably close to cloaking. The recommended replacements are server-side rendering, static generation or hydration. It survives in practice mainly as a stopgap on legacy single-page applications.", "example": "A legacy Angular app routes requests whose user agent matches Googlebot to a Rendertron instance while humans get the SPA.", "related": [ "prerendering", "server-side-rendering", "cloaking", "javascript-rendering", "spa" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/javascript/dynamic-rendering" }, { "id": "eeat", "slug": "eeat", "term": "E-E-A-T", "aka": [ "Experience, Expertise, Authoritativeness, Trustworthiness", "EAT" ], "category": "content", "shortDefinition": "The framework in Google's Search Quality Rater Guidelines for assessing page quality: Experience, Expertise, Authoritativeness and Trustworthiness, with Trust at the centre.", "definition": "E-E-A-T is not a ranking factor or a score Google computes; it is the concept quality raters use to evaluate results, and Google says its automated systems are built to identify signals that align with it. The first E, Experience, was added in December 2022 to credit first-hand knowledge — having actually used the product, visited the place, lived the condition. Trust is described as the most important member of the family: the other three exist to support it. It matters most on YMYL topics, where the bar for demonstrated competence is highest.", "example": "A review page adds original photographs, a named author with a verifiable track record, testing methodology and disclosure of how units were obtained.", "related": [ "ymyl", "quality-rater-guidelines", "helpful-content", "author-authority", "trust" ], "sourceUrl": "https://developers.google.com/search/docs/fundamentals/creating-helpful-content" }, { "id": "ecommerce-seo", "slug": "ecommerce-seo", "term": "Ecommerce SEO", "aka": [ "Retail SEO" ], "category": "ecommerce", "shortDefinition": "Search optimisation for online retail, dominated by catalogue scale, faceted URL control, product data accuracy and the interaction between organic results and shopping surfaces.", "definition": "The distinguishing problems are structural rather than editorial: hundreds of thousands of URLs generated by facets and variants, products that appear and disappear, and the need for feed data to agree with on-page markup. Category pages usually carry more commercial value than individual products, and Google's own e-commerce documentation covers pagination, faceted navigation, product markup and out-of-stock handling. Merchant Center and free product listings sit alongside classic organic work.", "example": "A retailer decides that only 400 of 90,000 possible facet combinations are indexable, and enforces it in the URL layer.", "related": [ "faceted-navigation", "plp", "pdp", "product-feed", "merchant-center" ], "sourceUrl": "https://developers.google.com/search/docs/specialty/ecommerce" }, { "id": "edge-seo", "slug": "edge-seo", "term": "Edge SEO", "aka": [], "category": "technical", "shortDefinition": "Implementing SEO changes in CDN edge workers rather than the origin application, so fixes ship without touching the main codebase or release cycle.", "definition": "Edge workers can inject or rewrite title tags, canonicals, hreflang and structured data, serve redirect maps, modify robots.txt, add security headers, and route crawlers to prerendered HTML. The appeal is speed on platforms where engineering capacity is the bottleneck; the risk is a hidden layer of logic that drifts from the origin and confuses future debugging. Changes should be version-controlled and documented like any other deployment.", "example": "A team ships a 12,000-entry redirect map as a Cloudflare Worker two days after a migration, while the origin backlog is six weeks long.", "related": [ "cdn", "redirect-301", "prerendering", "site-migration", "cache-control" ], "sourceUrl": null }, { "id": "editorial-guidelines", "slug": "editorial-guidelines", "term": "Editorial guidelines", "aka": [ "Editorial policy" ], "category": "content", "shortDefinition": "A published statement of how a publisher sources, reviews, corrects and discloses its content, supporting trust for readers, raters and AI systems.", "definition": "Typical components are sourcing standards, fact-checking and expert review processes, correction and update policies, conflict-of-interest and affiliate disclosure, and rules for AI assistance. Publishing them is one of the few direct levers on the Trust component of E-E-A-T, and quality raters are instructed to look for exactly this kind of information about a site. They also give AI systems explicit, quotable evidence of process when assessing a source.", "example": "A finance site publishes a page describing its two-stage review by a chartered accountant and its policy on correcting published errors.", "related": [ "eeat", "trust", "author-authority", "ymyl", "quality-rater-guidelines" ], "sourceUrl": "https://developers.google.com/search/docs/fundamentals/creating-helpful-content" }, { "id": "embedding", "slug": "embedding", "term": "Embedding", "aka": [ "Vector embedding" ], "category": "geo-ai", "shortDefinition": "A numerical vector representing the meaning of text, an image or another item, positioned so that semantically similar items lie close together in the vector space.", "definition": "Embeddings let systems compare meaning rather than matching strings, which is what allows a query and a passage with no shared vocabulary to be recognised as related. They power vector search, clustering, deduplication, recommendation and the retrieval half of RAG. In SEO workflows, embeddings are used practically for detecting cannibalisation, clustering keywords by intent, mapping internal link opportunities, and measuring how close a page is to the topic it targets.", "example": "Embedding 4,000 URLs and clustering them reveals nine pairs of pages whose vectors are nearly identical — clear cannibalisation candidates.", "related": [ "vector-search", "rag", "semantic-search", "keyword-cannibalization", "chunk-retrieval" ], "sourceUrl": null }, { "id": "engagement-rate", "slug": "engagement-rate", "term": "Engagement rate", "aka": [ "Engaged sessions" ], "category": "analytics", "shortDefinition": "The share of sessions that lasted over ten seconds, had a conversion event, or included at least two pageviews — GA4's replacement for bounce rate.", "definition": "Bounce rate in GA4 is simply the inverse of engagement rate, not the Universal Analytics metric of the same name, so historical comparisons are invalid. Engagement thresholds are configurable. As an SEO metric it is a weak proxy at best: Google has consistently said it does not use analytics data in ranking, and engagement varies by content type and intent far more than by quality.", "example": "A reference page with a 20% engagement rate performs exactly as intended — users get an answer in five seconds and leave.", "related": [ "ga4", "sessions", "conversion-rate", "attribution", "search-intent" ], "sourceUrl": "https://support.google.com/analytics/answer/12195621" }, { "id": "entity", "slug": "entity", "term": "Entity", "aka": [ "Named entity" ], "category": "content", "shortDefinition": "A distinct, uniquely identifiable thing — a person, place, organisation, product or concept — that a search engine can reason about independently of the words used to name it.", "definition": "Search engines maintain entity graphs (Google's is the Knowledge Graph) in which entities have identifiers, attributes and relationships. Working with entities rather than strings is what lets an engine understand that 'Apple' the company differs from the fruit, and to answer questions by traversing relationships. For site owners the practical work is making your own entity unambiguous: consistent naming, Organization and Person markup, `sameAs` links to authoritative profiles, and coverage on third-party sources the graph already trusts.", "example": "A consultancy links its Organization markup via `sameAs` to its Wikidata item, LinkedIn page and Companies House record.", "related": [ "knowledge-graph", "semantic-search", "knowledge-panel", "schema-org", "brand-mention-monitoring" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/structured-data/organization" }, { "id": "etag", "slug": "etag", "term": "ETag and conditional requests", "aka": [ "If-None-Match", "If-Modified-Since" ], "category": "technical", "shortDefinition": "A validator token identifying a specific version of a resource, allowing clients and crawlers to revalidate with a cheap 304 Not Modified response.", "definition": "When a crawler sends `If-None-Match` with a stored ETag, or `If-Modified-Since` with a stored date, an unchanged resource can be answered with a 304 carrying no body. Google supports conditional requests, so correct ETag and Last-Modified handling saves bandwidth on both sides and effectively stretches crawl capacity on large sites. ETags must change when content changes, or crawlers will keep serving stale copies from cache.", "example": "A 500,000-URL catalogue implements ETags and sees 60% of Googlebot fetches answered with 304, freeing capacity for new products.", "related": [ "cache-control", "crawl-budget", "http-status-codes", "cdn", "log-file-analysis" ], "sourceUrl": "https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/ETag" }, { "id": "evergreen-content", "slug": "evergreen-content", "term": "Evergreen content", "aka": [], "category": "content", "shortDefinition": "Content whose relevance persists over time, as opposed to news or seasonal pieces, though it still requires periodic maintenance to stay accurate.", "definition": "Evergreen topics are those where the underlying question does not change even if details do — how something works, how to do something, what a term means. They compound in value because links and authority accumulate against a stable URL. The trap is treating evergreen as maintenance-free: prices, screenshots, regulations and best practice all drift, and decay follows.", "example": "A guide to reading an energy bill stays relevant for years but needs its tariff examples and screenshots updated annually.", "related": [ "content-decay", "content-refresh", "freshness", "topical-authority", "pillar-page" ], "sourceUrl": null }, { "id": "exact-match-domain", "slug": "exact-match-domain", "term": "Exact-match domain system", "aka": [ "EMD" ], "category": "algorithms", "shortDefinition": "Google's system preventing low-quality sites from ranking merely because their domain name matches a query, introduced as an update in 2012.", "definition": "Before 2012 a domain such as cheap-car-insurance-quotes.com could rank on its name alone. The system removed that advantage for weak sites while leaving genuine brands unaffected, and it remains listed among Google's ranking systems. Keyword domains still carry minor anchor-text and click-through effects but are generally a weaker branding choice than a distinctive name.", "example": "A thin site on a keyword-rich domain loses its rankings while an established brand on a made-up name retains them.", "related": [ "ranking-systems", "thin-content", "brand-search", "domain-rating", "url-structure" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/ranking-systems-guide" }, { "id": "expired-domain-abuse", "slug": "expired-domain-abuse", "term": "Expired domain abuse", "aka": [ "Domain resurrection" ], "category": "algorithms", "shortDefinition": "Buying an expired domain and repurposing it to host unrelated low-value content in order to inherit its accumulated ranking signals.", "definition": "Added to Google's spam policies in March 2024, the policy targets the practice of exploiting a domain's history rather than its current merit — for instance turning a defunct charity's domain into an affiliate site. Legitimate acquisitions with continuity of purpose are not the target. The policy also undercuts the economics of PBNs, which have relied on expired domain inventory.", "example": "A closed medical clinic's domain is bought and relaunched as a casino affiliate site, retaining nothing but the backlinks.", "related": [ "pbn", "spam-policies", "spam-updates", "site-reputation-abuse", "link-spam" ], "sourceUrl": "https://developers.google.com/search/docs/essentials/spam-policies" }, { "id": "faceted-navigation", "slug": "faceted-navigation", "term": "Faceted navigation", "aka": [ "Filtered navigation", "Facets" ], "category": "crawling-indexing", "shortDefinition": "Filter and sort controls on listing pages that can generate a combinatorial explosion of crawlable URLs from a small product set.", "definition": "Each filter combination typically produces a distinct URL, so a handful of facets can yield millions of near-duplicate pages. The standard approach is to decide which facet combinations have genuine search demand and make those indexable with unique content and internal links, while blocking or noindexing the rest and stripping ordering, view and session parameters. Google's e-commerce guidance recommends using URL patterns crawlers can recognise and, where a facet space is truly infinite, disallowing it in robots.txt.", "example": "`/running-shoes/mens/waterproof/` is indexable, while `/running-shoes?sort=price&view=grid&page=17` is disallowed.", "related": [ "index-bloat", "crawl-budget", "url-parameters", "pagination", "plp" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/crawling-managing-faceted-navigation" }, { "id": "faq-howto-deprecation", "slug": "faq-howto-deprecation", "term": "FAQ and HowTo rich result deprecation", "aka": [ "FAQ rich results removal" ], "category": "serp-features", "shortDefinition": "Google's 2023 change restricting FAQ rich results to well-known authoritative government and health sites and removing HowTo rich results entirely.", "definition": "Announced in August 2023 as part of simplifying results, FAQ rich results ceased to be shown regularly for most sites, and HowTo rich results were removed from desktop and then dropped from Search Console reporting and the Rich Results Test. Google stated existing markup does not need to be removed because unused structured data causes no harm, but it also produces no visible benefit. Teams should not invest in FAQPage or HowTo markup for SERP appearance, though FAQPage may still be useful as plain semantic description.", "example": "A retailer removes 6,000 auto-generated FAQ blocks written solely to trigger a rich result that no longer appears.", "related": [ "rich-results", "structured-data", "schema-org", "serp", "rich-results-test" ], "sourceUrl": "https://developers.google.com/search/blog/2023/08/howto-faq-changes" }, { "id": "featured-snippet", "slug": "featured-snippet", "term": "Featured snippet", "aka": [ "Position zero", "Answer box" ], "category": "serp-features", "shortDefinition": "A block at or near the top of Google's results that quotes an extracted answer from a page, with a link to the source.", "definition": "Featured snippets appear as paragraphs, lists or tables and are selected programmatically — there is no markup to request one. Google promotes the snippet's source from its normal organic position rather than duplicating it. Pages win them by answering the question directly and unambiguously near a matching heading, in the format the query implies. Sites can opt out with `nosnippet`, `max-snippet` or `data-nosnippet`, at the cost of ordinary snippets too.", "example": "A clearly headed 40-word definition paragraph is lifted as a snippet for 'what is a service area business'.", "related": [ "people-also-ask", "passage-ranking", "nosnippet", "zero-click", "ai-overviews" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/featured-snippets" }, { "id": "first-contentful-paint", "slug": "first-contentful-paint", "term": "First Contentful Paint", "aka": [ "FCP" ], "category": "technical", "shortDefinition": "The time from navigation start until the browser renders the first piece of DOM content; a diagnostic metric rather than a Core Web Vital.", "definition": "FCP marks the first visible feedback that something is loading — any text, image, canvas or SVG. web.dev treats 1.8 s or less as good. Because FCP is dominated by TTFB and render-blocking resources, it is chiefly useful as a way to attribute LCP problems: a slow FCP points at server or blocking-resource issues, while a fast FCP with a slow LCP points at the hero resource itself.", "example": "FCP of 0.9 s alongside LCP of 4.0 s indicates the page paints quickly but the main image loads far too late.", "related": [ "largest-contentful-paint", "ttfb", "render-blocking-resources", "lighthouse", "core-web-vitals" ], "sourceUrl": "https://web.dev/articles/fcp" }, { "id": "first-input-delay", "slug": "first-input-delay", "term": "First Input Delay", "aka": [ "FID" ], "category": "technical", "shortDefinition": "The retired responsiveness Core Web Vital measuring only the delay before the browser began processing a page's first interaction; replaced by INP in March 2024.", "definition": "FID captured input delay but not the time to run the handler or paint the result, so pages could score well while feeling unresponsive. Google announced its deprecation in 2023 and removed it from Core Web Vitals in March 2024, retaining it briefly in tooling before removal. It is included here because historical reports, dashboards and older articles still reference it; current work should target INP.", "example": "A dashboard built in 2022 still charts FID; the data stops in 2024 and must be rebuilt against INP.", "related": [ "inp", "core-web-vitals", "crux", "lighthouse", "page-experience" ], "sourceUrl": "https://web.dev/articles/fid" }, { "id": "freshness", "slug": "freshness", "term": "Freshness systems (QDF)", "aka": [ "Query Deserves Freshness" ], "category": "algorithms", "shortDefinition": "Google's systems that boost recent content for queries where recency matters, such as breaking news, sports scores or evolving topics.", "definition": "Freshness is query-dependent: for a breaking story recency dominates, for a definitional query it barely matters. Signals include publication and update dates, crawl frequency, link acquisition rate and the behaviour of the query itself. Because artificially changing dates is a known manipulation, Google's guidance on publication dates asks for accurate, consistent dates in visible text and structured data.", "example": "A tax-rate page updated the morning of a budget announcement overtakes older pages within hours.", "related": [ "content-refresh", "top-stories", "news-sitemap", "content-decay", "ranking-systems" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/publication-dates" }, { "id": "generative-ai-performance-report", "slug": "generative-ai-performance-report", "term": "Generative AI performance report", "aka": [ "Search Console AI report" ], "category": "analytics", "shortDefinition": "Search Console views introduced in 2026 showing impressions and clicks for a site's links inside Google's generative AI features on Search and Discover.", "definition": "An impression is counted when a link to the site is scrolled or expanded into view within a generative feature, aggregated per property so multiple links in one answer count once. Clicks on links to external pages in AI Overviews or AI Mode count as clicks, and an AI Overview occupies a single position shared by all its links. The data is also included in the overall Performance report under the Web search type, so it is additive detail rather than a separate channel.", "example": "A publisher finds a fifth of its impressions on a topic come from generative features, with a materially lower CTR than classic results.", "related": [ "search-console", "ai-overviews", "ai-mode", "impressions", "average-position" ], "sourceUrl": "https://support.google.com/webmasters/answer/16984139" }, { "id": "geo", "slug": "geo", "term": "Generative Engine Optimization", "aka": [ "GEO" ], "category": "geo-ai", "shortDefinition": "The practice of improving how often and how accurately a brand's content is retrieved, cited and represented in AI-generated answers.", "definition": "GEO extends SEO rather than replacing it: generative systems are grounded on retrieval, so being crawlable, indexed and authoritative remains the entry ticket. What differs is the unit of success — a citation or an accurate mention inside a synthesised answer rather than a blue link position — and the measurement problem, since there are no rank positions and few native analytics. Google's own guidance states that no special AI optimisations are needed beyond good SEO, while practitioners focus on clear structure, extractable facts, and third-party corroboration.", "example": "A B2B vendor tracks how often five assistants name it when asked for alternatives to a competitor, and works on the review sites those answers cite.", "related": [ "aeo", "llmo", "ai-visibility", "citation-share", "ai-overviews" ], "sourceUrl": "https://developers.google.com/search/docs/fundamentals/ai-optimization-guide" }, { "id": "geo-targeting", "slug": "geo-targeting", "term": "Geo-targeting", "aka": [ "International targeting" ], "category": "international", "shortDefinition": "Signalling which country a site or section is intended for, through ccTLD choice, hreflang region codes, server location, local links and language.", "definition": "Google retired the International Targeting report's country setting in Search Console, so explicit country targeting for generic domains now relies chiefly on hreflang, on-page localisation and other contextual signals. Geo-targeting is about audience, not language: a Spanish-language site can target Mexico specifically. Note that targeting one country makes ranking elsewhere harder, so it should only be applied where the content is genuinely country-specific.", "example": "A `/au/` section uses `en-AU` hreflang, Australian pricing, local phone numbers and links from Australian publications.", "related": [ "hreflang", "cctld", "subfolder-vs-subdomain", "international-seo", "locale-redirects" ], "sourceUrl": "https://developers.google.com/search/docs/specialty/international/managing-multi-regional-sites" }, { "id": "ga4", "slug": "ga4", "term": "Google Analytics 4", "aka": [ "GA4" ], "category": "analytics", "shortDefinition": "Google's event-based analytics platform, which replaced Universal Analytics in July 2023 and models sessions and users from event streams.", "definition": "GA4 records every interaction as an event with parameters rather than counting pageviews and sessions natively, uses a different session definition, and applies data thresholding and modelling that can make small segments unreliable. Key differences that trip up reporting are engaged sessions replacing bounce rate, a default 14-month retention window for exploration data, and cardinality limits producing 'other' rows. Its BigQuery export is the standard escape hatch for anything requiring unsampled, unthresholded data.", "example": "A team exports GA4 events to BigQuery because thresholding suppressed the demographic breakdown they needed.", "related": [ "sessions", "engagement-rate", "attribution", "data-sampling", "conversion-rate" ], "sourceUrl": "https://support.google.com/analytics/answer/10089681" }, { "id": "google-business-profile", "slug": "google-business-profile", "term": "Google Business Profile", "aka": [ "GBP", "Google My Business" ], "category": "local", "shortDefinition": "The free business listing that powers a company's presence in Google Maps, the local pack and business panels, renamed from Google My Business in 2021.", "definition": "The profile carries name, address, phone, categories, hours, attributes, services, products, photos, posts, Q&A and reviews, and is verified by video, postcard, phone or instant methods depending on the business type. Google retired the standalone GBP web dashboard in favour of managing profiles directly in Search and Maps, with the API remaining for multi-location operators. Accuracy and completeness are the primary controllable inputs to local prominence.", "example": "Adding twenty geotagged photos, complete service descriptions and correct holiday hours precedes a measurable rise in profile calls.", "related": [ "nap", "gbp-categories", "local-pack", "local-reviews", "service-area-business" ], "sourceUrl": "https://support.google.com/business/answer/3038177" }, { "id": "google-api-leak", "slug": "google-api-leak", "term": "Google Content Warehouse API leak", "aka": [ "Google API leak", "2024 documentation leak" ], "category": "algorithms", "shortDefinition": "The May 2024 publication of thousands of internal Google Search API documentation attributes, which named ranking-related features not previously disclosed.", "definition": "The documents were briefly public in a code repository and described data structures and attribute names — including references to site authority, click-based signals and content classifiers — without describing weights or whether each was used in live ranking. Google confirmed the documents were genuine but cautioned against drawing conclusions about ranking from out-of-context, potentially outdated field names. The leak is best treated as vocabulary and hypothesis material, never as a specification.", "example": "The presence of a `siteAuthority` attribute prompted debate, though nothing in the documents established how or whether it is used in ranking.", "related": [ "navboost", "twiddlers", "ranking-systems", "domain-rating", "pagerank" ], "sourceUrl": null }, { "id": "merchant-center", "slug": "merchant-center", "term": "Google Merchant Center", "aka": [ "Merchant Center Next" ], "category": "ecommerce", "shortDefinition": "Google's platform for submitting product data, powering Shopping ads and free product listings across Search, Images and the Shopping tab.", "definition": "Merchant Center ingests feeds or crawls structured data directly from a site, validates against product data specifications and policies, and reports item-level disapprovals. It also surfaces competitive pricing and best-seller reports that are useful beyond advertising. Because free listings and organic results are separate systems with separate diagnostics, a product can rank organically while being disapproved in Merchant Center, or the reverse.", "example": "A retailer discovers 4,000 items disapproved for missing GTINs, none of which was visible in Search Console.", "related": [ "product-feed", "shopping-results", "merchant-listings", "price-schema", "ecommerce-seo" ], "sourceUrl": "https://support.google.com/merchants/answer/188924" }, { "id": "news-sitemap", "slug": "news-sitemap", "term": "Google News sitemap", "aka": [ "News sitemap" ], "category": "crawling-indexing", "shortDefinition": "A specialised sitemap containing only articles published in the last two days, using the news namespace to speed discovery for Google News and Top stories.", "definition": "News sitemaps use the `` extension with publication name, language and publication date. Google requires that URLs be removed once they are older than two days, though they remain indexed normally. The format is a discovery accelerator, not a ranking factor, and is only relevant to sites accepted into Google News surfaces.", "example": "A regional newspaper regenerates `/news-sitemap.xml` every five minutes with the 300 most recent stories.", "related": [ "xml-sitemap", "top-stories", "freshness", "sitemap-index", "indexnow" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/sitemaps/news-sitemap" }, { "id": "ranking-systems", "slug": "ranking-systems", "term": "Google ranking systems", "aka": [ "Ranking systems guide" ], "category": "algorithms", "shortDefinition": "Google's documented list of the systems that generate rankings, distinguished from one-off 'updates' to those systems, and including retired ones.", "definition": "The guide names current systems such as BERT, deduplication, exact-match domain handling, freshness, helpful content signals, link analysis, local news, MUM, neural matching, original content, page experience, passage ranking, product reviews, RankBrain, reliable information, reviews, site diversity and spam detection. It also lists retired systems — including Hummingbird, mobile-friendly ranking, page speed and Panda — whose functions were absorbed into core ranking. Reading it prevents chasing algorithms that no longer exist as separate entities.", "example": "A team checks the guide and finds the Helpful Content system is no longer listed separately because it was integrated into core ranking in 2024.", "related": [ "core-update", "bert", "rankbrain", "helpful-content-update", "deduplication" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/ranking-systems-guide" }, { "id": "search-console", "slug": "search-console", "term": "Google Search Console", "aka": [ "GSC", "Webmaster Tools" ], "category": "analytics", "shortDefinition": "Google's free property-level reporting and diagnostics service covering search performance, indexing status, enhancements, manual actions and security issues.", "definition": "Search Console provides query-level impressions, clicks, CTR and average position for 16 months, plus indexing reports, sitemap submission, URL Inspection, Core Web Vitals field data, structured data validation and manual action notices. Since 2026 it also exposes dedicated generative AI performance views for impressions and clicks from AI Overviews and AI Mode, which are additionally rolled into the Web search type totals. Data is sampled and filtered for privacy, so totals rarely reconcile exactly with analytics.", "example": "A team exports 16 months of query data via the Search Console API to build a share-of-search dashboard.", "related": [ "impressions", "average-position", "url-inspection-tool", "generative-ai-performance-report", "core-web-vitals" ], "sourceUrl": "https://support.google.com/webmasters/answer/9128668" }, { "id": "spam-policies", "slug": "spam-policies", "term": "Google spam policies", "aka": [ "Webmaster spam policies" ], "category": "algorithms", "shortDefinition": "Google's documented list of behaviours that can cause lower rankings or removal, covering cloaking, doorways, hacked content, keyword stuffing, link spam, scraped content and more.", "definition": "The policies define each violation and are the reference used both by automated systems and by the manual review team. They were expanded in March 2024 with scaled content abuse, expired domain abuse and site reputation abuse. Violations may be handled algorithmically, by manual action reported in Search Console, or both; reconsideration is only possible for manual actions.", "example": "A site issues its writers a checklist mapped directly to each policy section before publication.", "related": [ "manual-action", "spam-updates", "cloaking", "scaled-content-abuse", "site-reputation-abuse" ], "sourceUrl": "https://developers.google.com/search/docs/essentials/spam-policies" }, { "id": "google-extended", "slug": "google-extended", "term": "Google-Extended", "aka": [], "category": "geo-ai", "shortDefinition": "A robots.txt control token that governs whether a site's content can be used to improve Gemini models and grounded Gemini API responses, without affecting Google Search.", "definition": "Google-Extended is not a separate crawler — it is a token applied to content already fetched by Googlebot, letting publishers opt out of certain generative AI training and grounding uses while remaining fully in Search. Crucially, disallowing it does not remove a site from AI Overviews or AI Mode, which are Search features governed by ordinary snippet controls and the Search generative AI control. Confusing these two mechanisms is a widespread error.", "example": "`User-agent: Google-Extended` / `Disallow: /` opts a publisher out of Gemini training while its pages continue appearing in AI Overviews.", "related": [ "ai-overviews", "search-generative-ai-control", "nosnippet", "robots-txt", "model-training-opt-out" ], "sourceUrl": "https://developers.google.com/crawling/docs/crawlers-fetchers/google-common-crawlers" }, { "id": "googlebot", "slug": "googlebot", "term": "Googlebot", "aka": [ "Google crawler" ], "category": "crawling-indexing", "shortDefinition": "Google's main web crawler, which fetches pages for Google Search using two user agents: Googlebot Smartphone and Googlebot Desktop.", "definition": "Googlebot crawls from Google-owned IP ranges that are published as reverse-DNS-verifiable JSON files, so site owners can distinguish it from spoofers. Since mobile-first indexing, the primary crawler for nearly all sites is Googlebot Smartphone. Googlebot obeys robots.txt, supports HTTP/2 where the server offers it, and fetches at most a few megabytes of HTML per URL. It is distinct from Google's other agents such as GoogleOther, Google-Extended, AdsBot and Google-InspectionTool.", "example": "Server logs show `Mozilla/5.0 (Linux; Android 6.0.1; Nexus 5X Build/MMB29P) AppleWebKit/537.36 ... Googlebot/2.1` hitting a category page 40 times a day.", "related": [ "crawling", "mobile-first-indexing", "robots-txt", "google-extended", "log-file-analysis" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/googlebot" }, { "id": "gptbot", "slug": "gptbot", "term": "GPTBot", "aka": [], "category": "geo-ai", "shortDefinition": "OpenAI's crawler used to collect publicly available web content that may be used to improve future models; controllable via robots.txt.", "definition": "OpenAI documents GPTBot alongside OAI-SearchBot, which builds the search index used to surface and link sites, and ChatGPT-User, which fetches a page because a user or agent asked for it. Blocking GPTBot opts content out of training-oriented crawling without necessarily removing a site from ChatGPT's search results, since those use different agents. OpenAI publishes IP ranges so operators can verify requests rather than trusting the user-agent string.", "example": "`User-agent: GPTBot` / `Disallow: /` blocks training collection while `OAI-SearchBot` remains allowed for search visibility.", "related": [ "oai-searchbot", "chatgpt-user", "ai-crawler", "model-training-opt-out", "robots-txt" ], "sourceUrl": "https://platform.openai.com/docs/bots" }, { "id": "grounding", "slug": "grounding", "term": "Grounding", "aka": [ "Search grounding" ], "category": "geo-ai", "shortDefinition": "Constraining a language model's output to information retrieved from a specified source — a search index, a document set or a database — so answers can be attributed and verified.", "definition": "Grounding is what separates a search-connected assistant from a model answering purely from parametric memory: retrieved passages are placed in the model's context and the answer is generated over them, usually with citations back to the source documents. It substantially reduces but does not eliminate hallucination, since a model can still misread or over-generalise retrieved text. For site owners, being retrievable is the prerequisite for being grounded on, which is why crawlability and clear factual statements matter.", "example": "An assistant asked about a tax threshold retrieves the current guidance page and quotes the figure with a link rather than recalling last year's number.", "related": [ "rag", "hallucination", "chunk-retrieval", "ai-overviews", "citation-share" ], "sourceUrl": null }, { "id": "guest-posting", "slug": "guest-posting", "term": "Guest posting", "aka": [ "Guest blogging", "Contributed articles" ], "category": "links", "shortDefinition": "Writing articles for other publications; legitimate as an audience and authority tactic, but a link spam violation when done at scale primarily for keyword-rich links.", "definition": "Google's link spam policy explicitly names large-scale article campaigns and guest posts with keyword-rich anchor text in the content. The distinguishing factors are audience intent, editorial standards, and whether links are natural and disclosed. Sites that accept paid contributions at volume risk their own quality signals as well as the contributor's, and 'link inserts' bought into existing posts fall squarely inside the policy.", "example": "A specialist writes a genuinely original piece for an industry magazine with a brand-anchor bio link, rather than buying 50 templated placements.", "related": [ "link-spam", "digital-pr", "link-building", "site-reputation-abuse", "sponsored-attribute" ], "sourceUrl": "https://developers.google.com/search/docs/essentials/spam-policies" }, { "id": "hallucination", "slug": "hallucination", "term": "Hallucination", "aka": [ "Confabulation" ], "category": "geo-ai", "shortDefinition": "A confident, fluent but factually wrong or fabricated statement produced by a language model, including invented citations, prices, features or quotes.", "definition": "Hallucination arises because models generate plausible continuations rather than looking facts up; grounding reduces it by supplying retrieved evidence but does not remove it, since models can still misread or over-extrapolate from context. For brands the practical exposure is misinformation at scale: wrong prices, invented policies, or attributions to the wrong company. Mitigations are stating key facts explicitly and unambiguously on your own pages, keeping third-party sources current, and monitoring answers for recurring errors.", "example": "An assistant tells users a SaaS product has a 30-day free trial that was discontinued two years earlier.", "related": [ "grounding", "brand-mention-monitoring", "rag", "ai-visibility", "ai-generated-content" ], "sourceUrl": null }, { "id": "h1", "slug": "h1", "term": "Heading structure (H1-H6)", "aka": [ "H1", "Headings" ], "category": "technical", "shortDefinition": "The hierarchy of HTML heading elements that conveys a document's outline to users, assistive technology and parsers.", "definition": "Google has said multiple H1s are acceptable and that headings are used to understand page structure rather than as a strong ranking lever. Their real value is threefold: accessibility, snippet and passage extraction, and — increasingly — machine consumption, since headings help both search engines and language models locate the section that answers a question. Skipping levels or using headings purely for visual sizing undermines all three.", "example": "Converting a bolded paragraph acting as a section label into a real `

` makes the section eligible to be extracted as a featured snippet.", "related": [ "title-tag", "featured-snippet", "semantic-html", "meta-description", "passage-ranking" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/structured-data/article" }, { "id": "helpful-content-update", "slug": "helpful-content-update", "term": "Helpful content system and updates", "aka": [ "HCU" ], "category": "algorithms", "shortDefinition": "A Google system launched in August 2022 to demote sites with a high proportion of content made for search engines, integrated into core ranking with the March 2024 core update.", "definition": "The system originally applied a site-wide classifier that could weigh on an entire domain, and it ran as discrete updates with long recovery cycles — the September 2023 update was particularly severe for many publishers. With the March 2024 core update Google folded the signals into its core ranking systems, so 'helpfulness' is now assessed continuously and there are no separate HCU announcements. The self-assessment questions in Google's creating helpful content documentation remain the reference standard.", "example": "A recipe blog hit in 2023 rebuilt its content and saw partial recovery only at a subsequent core update, not in between.", "related": [ "helpful-content", "core-update", "thin-content", "content-pruning", "ranking-systems" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/core-updates" }, { "id": "helpful-content", "slug": "helpful-content", "term": "Helpful, reliable, people-first content", "aka": [ "Helpful content" ], "category": "content", "shortDefinition": "Google's stated content standard: content created primarily for people, that demonstrates first-hand expertise and leaves readers feeling their need was met.", "definition": "Google publishes a self-assessment list of questions covering originality, depth, expertise, accuracy and whether content exists mainly to attract search traffic. The separate Helpful Content system was folded into Google's core ranking systems with the March 2024 core update, so 'helpfulness' is now evaluated continuously rather than by a distinct periodic update. Google emphasises that the assessment is site-wide in character: large volumes of unhelpful content can weigh on a whole domain.", "example": "A publisher deletes 600 thin roundups written purely for keyword coverage and consolidates the useful parts into 40 substantial guides.", "related": [ "eeat", "helpful-content-update", "thin-content", "content-pruning", "quality-rater-guidelines" ], "sourceUrl": "https://developers.google.com/search/docs/fundamentals/creating-helpful-content" }, { "id": "hreflang", "slug": "hreflang", "term": "hreflang", "aka": [ "rel=alternate hreflang" ], "category": "international", "shortDefinition": "An annotation declaring alternate language or regional versions of a page, so search engines serve the appropriate one to each user.", "definition": "hreflang can be implemented in the HTML head, in HTTP headers, or in XML sitemaps, and its cardinal rule is reciprocity: every page in a set must reference every other, including itself. Values use ISO 639-1 language codes optionally combined with ISO 3166-1 Alpha-2 region codes — `en-GB`, `es-MX` — and using a country code alone is invalid. It is a targeting signal, not a ranking signal or a redirect: it swaps which version ranks, it does not create rankings.", "example": "A page set of `en-GB`, `en-AU`, `en-US` and `x-default` versions each list all four URLs, so Australian users receive the AU page.", "related": [ "x-default", "international-seo", "multilingual-sitemap", "geo-targeting", "canonicalization" ], "sourceUrl": "https://developers.google.com/search/docs/specialty/international/localized-versions" }, { "id": "x-default", "slug": "x-default", "term": "hreflang x-default", "aka": [ "x-default" ], "category": "international", "shortDefinition": "The hreflang value marking the fallback page for users whose language or region does not match any specified alternate.", "definition": "`x-default` typically points at a global page, a language selector, or the default-market version. It is not mandatory but is strongly recommended for large international sets, because without it users outside the declared markets are served whichever version the engine guesses. It must be included reciprocally in the set like any other annotation.", "example": "A brand's country selector page is declared as `x-default` so a visitor from an untargeted country lands there rather than on the US site.", "related": [ "hreflang", "international-seo", "geo-targeting", "multilingual-sitemap", "locale-redirects" ], "sourceUrl": "https://developers.google.com/search/docs/specialty/international/localized-versions" }, { "id": "hsts", "slug": "hsts", "term": "HSTS", "aka": [ "HTTP Strict Transport Security" ], "category": "technical", "shortDefinition": "A response header instructing browsers to only contact a host over HTTPS for a set period, eliminating the initial insecure request.", "definition": "`Strict-Transport-Security: max-age=31536000; includeSubDomains; preload` tells browsers to upgrade http URLs internally, which shows as a 307 internal redirect rather than a network round trip. Domains can be submitted to a browser preload list so the policy applies even on first visit. HSTS applies to browsers, not crawlers, so server-side 301s from http to https are still required for search engines.", "example": "After enabling HSTS with preload, http requests never leave the browser, cutting one redirect hop from TTFB for returning users.", "related": [ "https", "redirect-307", "ttfb", "mixed-content", "redirect-chain" ], "sourceUrl": "https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Strict-Transport-Security" }, { "id": "http-status-codes", "slug": "http-status-codes", "term": "HTTP status codes", "aka": [ "Response codes" ], "category": "technical", "shortDefinition": "Numeric codes in an HTTP response that tell crawlers how to treat a URL: 200 index it, 3xx follow the redirect, 404/410 drop it, 5xx come back later.", "definition": "Google treats 200 as indexable, 301/308 as permanent moves that consolidate signals, 302/307 as temporary, 404 and 410 as gone (410 slightly faster to drop), 429 and 503 as back-off signals, and other 5xx as errors that reduce crawl rate. Serving the wrong code is a frequent bug: 200 on missing pages creates soft 404s, and 503 left in place for weeks eventually causes deindexing. Codes are also what redirect chains and loops are diagnosed with.", "example": "During a two-hour deploy, a site returns 503 with a `Retry-After` header so Googlebot pauses rather than recording thousands of errors.", "related": [ "soft-404", "redirect-301", "redirect-302", "crawl-rate-limit", "site-migration" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/http-network-errors" }, { "id": "http2", "slug": "http2", "term": "HTTP/2", "aka": [ "h2" ], "category": "technical", "shortDefinition": "A binary, multiplexed version of HTTP that carries many concurrent requests over one TCP connection with header compression.", "definition": "Multiplexing removes the head-of-line blocking that made domain sharding and sprite sheets worthwhile under HTTP/1.1, so those old optimisations are now counterproductive. Googlebot crawls over HTTP/2 where the server supports it, which reduces connection overhead for both sides. HTTP/2 still suffers TCP-level head-of-line blocking on lossy networks, which is what HTTP/3 addresses.", "example": "Consolidating assets that were split across four sharded subdomains onto one HTTP/2 origin reduces handshakes and improves TTFB.", "related": [ "http3", "ttfb", "https", "cdn", "googlebot" ], "sourceUrl": "https://developers.google.com/search/blog/2020/09/googlebot-will-soon-speak-http2" }, { "id": "http3", "slug": "http3", "term": "HTTP/3", "aka": [ "QUIC" ], "category": "technical", "shortDefinition": "The version of HTTP that runs over QUIC on UDP, removing transport-level head-of-line blocking and speeding up connection setup.", "definition": "QUIC combines transport and TLS handshakes, supports 0-RTT resumption, and keeps streams independent so a lost packet does not stall unrelated responses. The gains are largest on mobile and lossy networks, which is where Core Web Vitals are usually hardest to pass. It is typically enabled at the CDN with an `Alt-Svc` header advertising availability; origins can remain on HTTP/2.", "example": "Enabling HTTP/3 at the edge reduces p75 mobile TTFB by roughly a fifth on a site with heavy traffic from congested networks.", "related": [ "http2", "ttfb", "cdn", "https", "core-web-vitals" ], "sourceUrl": "https://developer.mozilla.org/en-US/docs/Glossary/HTTP_3" }, { "id": "https", "slug": "https", "term": "HTTPS", "aka": [ "TLS", "SSL" ], "category": "technical", "shortDefinition": "HTTP over TLS, encrypting traffic between browser and server; a lightweight Google ranking signal since 2014 and a baseline expectation today.", "definition": "Beyond the ranking signal, HTTPS is required for many browser APIs, HTTP/2 and HTTP/3 in practice, and referrer data preservation. SEO issues arise from certificate errors, mixed content, incomplete redirects leaving http URLs live, and canonical tags still pointing at http. Free automated certificates have made HTTPS universal, so its practical value now lies in avoiding the penalty of getting it wrong.", "example": "A certificate expires on a Sunday; Googlebot records TLS errors and crawling of the host drops sharply until it is renewed.", "related": [ "hsts", "mixed-content", "redirect-301", "page-experience", "http2" ], "sourceUrl": "https://developers.google.com/search/blog/2014/08/https-as-ranking-signal" }, { "id": "hummingbird", "slug": "hummingbird", "term": "Hummingbird", "aka": [], "category": "algorithms", "shortDefinition": "Google's 2013 rewrite of its core ranking infrastructure, shifting emphasis from individual keywords toward the meaning of whole queries.", "definition": "Hummingbird was an architectural replacement rather than a filter, enabling query rewriting, synonym handling and conversational query understanding, and paving the way for entity-centric search alongside the Knowledge Graph. Google now lists it among retired systems because its capabilities were absorbed into successor systems. Its conceptual legacy is that pages answer questions, not keyword strings.", "example": "'Where can I buy an iPhone charger near me' began returning local retailers rather than pages containing that exact phrase.", "related": [ "semantic-search", "rankbrain", "bert", "knowledge-graph", "ranking-systems" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/ranking-systems-guide" }, { "id": "hydration", "slug": "hydration", "term": "Hydration", "aka": [ "Rehydration" ], "category": "technical", "shortDefinition": "Attaching client-side JavaScript behaviour to server-rendered HTML so a static page becomes interactive.", "definition": "Hydration is the point where SSR's fast paint meets the cost of shipping and executing the framework, and it is a common cause of poor INP: the page looks ready but does not respond until hydration finishes. Techniques such as partial, progressive, islands and selective hydration reduce this by hydrating only interactive regions. A hydration mismatch, where client output differs from server HTML, can also cause content to be replaced after render.", "example": "A 400 KB bundle blocks the main thread for 900 ms after paint, so early taps on the menu do nothing and INP measures over 500 ms.", "related": [ "server-side-rendering", "inp", "spa", "javascript-rendering", "progressive-enhancement" ], "sourceUrl": "https://web.dev/articles/rendering-on-the-web" }, { "id": "image-sitemap", "slug": "image-sitemap", "term": "Image and video sitemap extensions", "aka": [ "Image sitemap", "Video sitemap" ], "category": "crawling-indexing", "shortDefinition": "Namespace extensions to XML sitemaps that declare images and videos associated with a page so they can be discovered for Images and Video search.", "definition": "The image extension lets you list image URLs per page; the video extension carries thumbnail, title, description, content or player URL and duration. They matter most when media is loaded by JavaScript, served from a separate CDN host, or lazy-loaded in ways a crawler may miss. For video, the sitemap and VideoObject structured data can both be supplied and should agree.", "example": "A recipe site adds `` entries with thumbnail and duration so its how-to clips are eligible for video results.", "related": [ "xml-sitemap", "image-pack", "video-carousel", "structured-data", "lazy-loading" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/sitemaps/video-sitemaps" }, { "id": "image-pack", "slug": "image-pack", "term": "Image pack", "aka": [ "Image results block" ], "category": "serp-features", "shortDefinition": "A horizontal or grid block of image results embedded within the web SERP, linking through to Google Images.", "definition": "Images qualify through the usual image indexing signals: crawlable and indexable files, descriptive filenames and alt text, relevant surrounding content, and structured data where applicable. Google Images results show the source page, so the traffic value depends on the click continuing to the site. Lazy loading implemented in JavaScript that crawlers never trigger is a frequent reason images are missing entirely.", "example": "A product photography set ranks in the image pack for a model number because filenames, alt text and captions all name it.", "related": [ "alt-text", "image-sitemap", "lazy-loading", "serp", "webp" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/google-images" }, { "id": "impressions", "slug": "impressions", "term": "Impressions", "aka": [], "category": "analytics", "shortDefinition": "In Search Console, the number of times a site's links appeared in results for a query, counted when the link is scrolled or expanded into view.", "definition": "Impressions are counted per property and per query, and links that require scrolling or expansion count only once they enter view. Impressions from generative AI features are included on the same basis. Because impressions capture visibility independent of clicks, they are the more robust top-line metric in a zero-click environment, and impression trends often reveal SERP layout changes that clicks alone would misattribute to ranking loss.", "example": "A page's impressions rise 40% while clicks fall, revealing that an AI Overview now sits above it.", "related": [ "click-through-rate", "average-position", "search-console", "zero-click", "generative-ai-performance-report" ], "sourceUrl": "https://support.google.com/webmasters/answer/7042828" }, { "id": "in-house-vs-agency", "slug": "in-house-vs-agency", "term": "In-house vs agency", "aka": [ "Resourcing model" ], "category": "business", "shortDefinition": "The choice between building an internal SEO team, retaining an external agency or consultant, or operating a hybrid of both.", "definition": "In-house teams carry context, relationships and the ability to influence roadmaps, which is where most large gains sit; agencies bring cross-client pattern recognition, specialist skills and elastic capacity. The hybrid model — internal ownership with external specialists for migrations, digital PR or technical depth — is the most common at scale. The decisive factor is usually whether the organisation can act on recommendations, since advice without implementation capacity delivers nothing.", "example": "A retailer hires one internal lead to own the roadmap and retains a technical specialist for its replatform.", "related": [ "agency-retainer", "seo-maturity-model", "seo-governance", "seo-strategy", "seo-roi" ], "sourceUrl": null }, { "id": "incrementality", "slug": "incrementality", "term": "Incrementality testing", "aka": [ "Lift testing" ], "category": "analytics", "shortDefinition": "Measuring the additional outcomes caused by an activity, by comparing a treated group against an untreated control, rather than inferring causation from attribution.", "definition": "In SEO, true randomised tests are hard because you cannot randomise search results, so practitioners use geo holdouts, template-level split tests across matched URL groups, and switchback designs. The value is in answering whether a change caused an effect, as opposed to whether it coincided with one. Designs must control for seasonality, algorithm updates and cross-contamination between groups.", "example": "A retailer applies new title templates to a random half of 6,000 product pages and compares clicks against the untouched half.", "related": [ "causal-impact", "attribution", "seo-testing", "forecasting", "seo-roi" ], "sourceUrl": null }, { "id": "indented-results", "slug": "indented-results", "term": "Indented results", "aka": [ "Nested results", "Multiple results from one site" ], "category": "serp-features", "shortDefinition": "A layout where a second result from the same site appears indented under the first, indicating Google judged both relevant to the query.", "definition": "Indented results are a display treatment, not a separate ranking system; they typically arise when two closely related pages from one domain both qualify. Google generally limits how many results a single site can occupy for a query, with more allowed when it judges diversity would not help. In practice indentation is a signal that two pages overlap in topic, which sometimes indicates cannibalisation worth resolving.", "example": "A guide ranks first with its own chapter page indented directly beneath it for a long-tail query.", "related": [ "serp", "keyword-cannibalization", "sitelinks", "site-diversity", "internal-linking" ], "sourceUrl": null }, { "id": "index-bloat", "slug": "index-bloat", "term": "Index bloat", "aka": [], "category": "crawling-indexing", "shortDefinition": "A condition where a site has far more URLs indexed or crawlable than it has genuinely useful pages, diluting crawl attention and site-level quality signals.", "definition": "Common sources are faceted navigation, session and tracking parameters, paginated archives, internal search results, tag pages and auto-generated location or variant pages. The cost is twofold: crawlers spend their budget on low-value URLs, and quality evaluation systems see a site dominated by thin pages. Remediation combines noindex, canonicalisation, robots.txt for genuinely infinite spaces, and outright deletion.", "example": "A WordPress site has 60 real posts but 9,000 indexed URLs because every tag, author and date archive is crawlable and paginated.", "related": [ "faceted-navigation", "thin-content", "crawl-budget", "noindex", "content-pruning" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/large-site-managing-crawl-budget" }, { "id": "indexing", "slug": "indexing", "term": "Indexing", "aka": [ "Index selection" ], "category": "crawling-indexing", "shortDefinition": "Storing and organising a crawled page in a search engine's index so it becomes eligible to be retrieved and ranked for queries.", "definition": "After crawling and rendering, a search engine parses the page, extracts text, links, structured data and media, resolves the canonical among duplicates, and decides whether the page is worth storing. Indexing is selective: search engines routinely decline to index pages they judge low value, duplicative or thin. Being indexed is a prerequisite for ranking but guarantees nothing about position.", "example": "Search Console's Page indexing report shows 12,000 pages 'Crawled - currently not indexed', meaning Google fetched them but chose not to store them.", "related": [ "crawling", "canonicalization", "index-bloat", "crawled-currently-not-indexed", "url-inspection-tool" ], "sourceUrl": "https://developers.google.com/search/docs/fundamentals/how-search-works" }, { "id": "indexnow", "slug": "indexnow", "term": "IndexNow", "aka": [], "category": "crawling-indexing", "shortDefinition": "An open protocol for pushing URL change notifications to participating search engines, adopted by Microsoft Bing, Yandex, Seznam and Naver.", "definition": "A site hosts a key file at its root, then submits added, updated or deleted URLs by HTTP GET for a single URL or POST with a JSON array of up to 10,000 URLs. Participating engines share submissions with each other, so one call reaches all of them. Google has run an evaluation but does not use IndexNow for Google Search. The protocol is a discovery accelerator only — it does not guarantee crawling or indexing.", "example": "A CMS fires an IndexNow POST containing the 12 URLs touched by a publish action immediately after deployment.", "related": [ "xml-sitemap", "crawling", "bing-webmaster-tools", "url-inspection-tool", "freshness" ], "sourceUrl": "https://www.indexnow.org/documentation" }, { "id": "infinite-scroll", "slug": "infinite-scroll", "term": "Infinite scroll", "aka": [ "Endless scrolling" ], "category": "crawling-indexing", "shortDefinition": "A UI pattern that appends more items as the user scrolls; without backing paginated URLs the additional items are usually invisible to crawlers.", "definition": "Crawlers do not scroll, so anything loaded by a scroll event is not fetched. The supported pattern is to back the experience with real, linkable paginated URLs updated via the History API, so each chunk of items exists at a crawlable address. Failing that, items beyond the first viewport batch are typically discovered only through sitemaps, which delays or prevents indexing of deep inventory.", "example": "A jobs board loads 20 listings per scroll; adding `/jobs?page=2` URLs with real anchors makes listings 21-40 crawlable.", "related": [ "pagination", "javascript-rendering", "crawl-budget", "internal-linking", "orphan-page" ], "sourceUrl": "https://developers.google.com/search/docs/specialty/ecommerce/pagination-and-incremental-page-loading" }, { "id": "information-gain", "slug": "information-gain", "term": "Information gain", "aka": [ "Novel information" ], "category": "content", "shortDefinition": "The amount of genuinely new information a document adds beyond what the user has already seen in other results for the same query.", "definition": "The term entered SEO vocabulary via a Google patent describing scoring documents for additional information relative to those already presented. Whether or not that specific mechanism is live, the principle is directly useful: a page that only restates the consensus of the top ten results gives a search engine no reason to add it, and gives a generative system nothing distinct to cite. Original data, first-hand testing, expert interviews and unique synthesis are the practical sources of gain.", "example": "A review adds bench measurements no other outlet publishes, which competitors then cite — a citable, non-commodity contribution.", "related": [ "eeat", "helpful-content", "citation-share", "topical-authority", "content-refresh" ], "sourceUrl": null }, { "id": "inp", "slug": "inp", "term": "Interaction to Next Paint", "aka": [ "INP" ], "category": "technical", "shortDefinition": "A Core Web Vital measuring overall responsiveness as the latency of the worst (approximately) interaction on a page; good is 200 ms or less at p75.", "definition": "INP observes clicks, taps and key presses through the page's whole lifetime and reports a high-percentile latency, covering input delay, processing time and presentation delay. It replaced First Input Delay in March 2024 because FID measured only the first interaction and only its input delay, and therefore looked good on pages that were in practice sluggish. Common causes of poor INP are long JavaScript tasks, heavy event handlers, large DOM sizes and third-party scripts.", "example": "A filter panel runs a 500 ms synchronous sort on every checkbox click; yielding to the main thread and debouncing brings INP under 200 ms.", "related": [ "core-web-vitals", "first-input-delay", "third-party-scripts", "hydration", "crux" ], "sourceUrl": "https://web.dev/articles/inp" }, { "id": "internal-linking", "slug": "internal-linking", "term": "Internal linking", "aka": [ "Internal links" ], "category": "links", "shortDefinition": "Links between pages on the same site, which control crawl paths, distribute link equity, and communicate topical relationships and page importance.", "definition": "Internal links are the only link type you fully control, and on large sites they usually have more impact than any realistic external campaign. Effective work means reducing click depth to important pages, linking contextually within body content rather than only from navigation, using descriptive anchor text, and ensuring links are real `` elements crawlers can follow. Automated internal linking should be constrained by relevance rules or it produces noise.", "example": "Adding contextual links from 40 informational articles to a commercial category page lifts its impressions substantially over a quarter.", "related": [ "anchor-text", "link-equity", "site-architecture", "orphan-page", "topic-cluster" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/links-crawlable" }, { "id": "international-seo", "slug": "international-seo", "term": "International SEO", "aka": [ "Global SEO" ], "category": "international", "shortDefinition": "Structuring and optimising a site so the right language and country version is served to the right audience in each market's search results.", "definition": "The discipline has three decisions at its core: URL structure (ccTLD, subdirectory or subdomain), targeting signals (hreflang and geo-targeting), and content strategy (translation, transcreation or independent local content). Getting the structure wrong is expensive to reverse, so it should be settled before scale. Local competitiveness varies enormously by market, and search engine share differs — Baidu, Yandex, Naver and Seznam matter in their territories.", "example": "A retailer chooses `/de-ch/` and `/fr-ch/` subdirectories for Switzerland rather than separate domains, to share domain authority.", "related": [ "hreflang", "cctld", "subfolder-vs-subdomain", "geo-targeting", "transcreation" ], "sourceUrl": "https://developers.google.com/search/docs/specialty/international" }, { "id": "internationalized-domain-name", "slug": "internationalized-domain-name", "term": "Internationalised domain name", "aka": [ "IDN", "Punycode" ], "category": "international", "shortDefinition": "A domain name containing non-ASCII characters, encoded for DNS as ASCII Punycode beginning with `xn--`.", "definition": "IDNs let brands use native scripts — Cyrillic, Arabic, Chinese, accented Latin — and Google handles them, but the Punycode form appears in logs, analytics, some tools and occasionally in browsers, which complicates reporting and link building. They also carry a homograph-spoofing risk that leads some browsers and security tools to display the Punycode form. Non-ASCII paths follow similar rules and should be consistently percent-encoded.", "example": "`münchen-hotels.de` resolves as `xn--mnchen-hotels-ilb.de` in DNS and appears that way in server logs.", "related": [ "cctld", "url-structure", "international-seo", "multilingual-seo", "geo-targeting" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/url-structure" }, { "id": "intrusive-interstitial", "slug": "intrusive-interstitial", "term": "Intrusive interstitial", "aka": [ "Pop-up penalty" ], "category": "technical", "shortDefinition": "A pop-up or overlay that obscures the main content immediately after a user arrives from search, which Google's guidance discourages on mobile.", "definition": "Google distinguishes intrusive interstitials — full-screen overlays and standalone gateway pages covering content on entry — from acceptable ones such as legally required cookie or age notices, login walls for gated content, and reasonably sized banners. The concern is that users tapping a result cannot see the content they came for. Beyond the guidance, interstitials also degrade CLS and INP.", "example": "A newsletter overlay covering the whole viewport 200 ms after a mobile visitor lands is replaced with a slim bottom bar.", "related": [ "page-experience", "cumulative-layout-shift", "mobile-first-indexing", "core-web-vitals", "helpful-content" ], "sourceUrl": "https://developers.google.com/search/docs/appearance/page-experience" }, { "id": "javascript-rendering", "slug": "javascript-rendering", "term": "JavaScript rendering", "aka": [ "JavaScript SEO" ], "category": "crawling-indexing", "shortDefinition": "The execution of a page's client-side JavaScript by a search engine so that content and links generated at runtime become indexable.", "definition": "Google renders with an evergreen Chromium and indexes the resulting DOM, but rendering is deferred, resource-limited and unforgiving: it does not click, scroll indefinitely, accept cookies, or wait for user interaction. Content behind interactions, requests blocked by robots.txt, or errors that throw during hydration simply will not be indexed. Other engines and most AI crawlers render far less reliably than Google, which is why server-rendered HTML remains the safest baseline.", "example": "A site loads reviews only after a 'Show reviews' click; the reviews never appear in Google's rendered HTML, so review keywords never rank.", "related": [ "render-queue", "dom", "server-side-rendering", "hydration", "spa" ], "sourceUrl": "https://developers.google.com/search/docs/crawling-indexing/javascript/javascript-seo-basics" }, { "id": "haro-sourcing", "slug": "haro-sourcing", "term": "Journalist sourcing platforms", "aka": [ "HARO", "Connectively", "Qwoted", "Featured" ], "category": "links", "shortDefinition": "Services that connect journalists seeking expert sources with people willing to comment, historically a route to earning media links and citations.", "definition": "Help a Reporter Out was the best-known service; it was rebranded as Connectively by Cision and then shut down in December 2024, with Qwoted, Featured, SourceBottle and journalist requests on social platforms absorbing the demand. The mechanic is unchanged: fast, specific, genuinely expert responses win placements, while templated pitches do not. Volume-driven abuse of these platforms, including AI-generated fake expertise, has made editors far more sceptical.", "example": "A veterinarian answers a national newspaper query about pet insurance within twenty minutes and is quoted with a link to her practice.", "related": [ "digital-pr", "link-building", "unlinked-mentions", "author-authority", "brand-mention-monitoring" ], "sourceUrl": null }, { "id": "json-ld", "slug": "json-ld", "term": "JSON-LD", "aka": [ "JavaScript Object Notation for Linked Data" ], "category": "technical", "shortDefinition": "A W3C-standard JSON syntax for linked data, embedded in a `