THE SEO GLOSSARY
& PLAYBOOK
On-Page · Off-Page · Technical · Local · Content
AIO · AEO · GEO · LLMO · SXO · Metrics · Tools
465 defined terms across 14 sections
46 core terms with full playbook entries
174 referenced sources, primary documentation first
Every entry verified against published sources, August 2026
Reference edition · August 2026
How to use this playbook
This document is two things at once. It is a glossary — 465 terms defined precisely enough to settle an argument — and it is a playbook, because the 46 terms that carry the most operational weight get a full entry rather than a definition.
Those core entries follow a fixed structure: what the term means, why it matters commercially, how to actually execute it, and the mistake practitioners most often make. Everything else is defined in two or three lines in the tables that follow each section.
Sections 1 to 6 cover the classical disciplines: foundations, on-page, content, technical, off-page, and local. Sections 7 to 10 cover the AI-era disciplines — AI Overviews, answer engines, generative engines, and search experience. Sections 11 and 12 cover measurement and tooling. Section 13 is the one most glossaries omit and this one leads with: terms that are dead, and claims that were never true.
Framing throughout is B2B technology and services rather than ecommerce, since that is where the commercial questions are hardest.
Rows and passages marked CHANGED. These flag something that has been deprecated, renamed, removed, or that is asserted in common practice beyond what the evidence supports. They are the most valuable part of the document — roughly a fifth of what practitioners still repeat about SEO expired at some point in the last three years, and most published glossaries have not caught up. Section 13 collects them all in one place.
On contested vocabulary. The boundaries between SEO, AEO, GEO, and LLMO are genuinely unsettled — Google says AI-search optimisation is "still SEO," Ahrefs says the three acronyms name one concept, and only GEO has a defensible academic origin. This glossary defines each on its own terms and states plainly where the industry has not agreed, rather than inventing a tidy taxonomy that does not exist. See the note at the end of Section 9.
Contents
Section 1 Foundations & Search Fundamentals............................... Core vocabulary
Section 2 On-Page SEO..................................................... Title, headings, links, images
Section 3 Content SEO & Search Intent.................................. Clusters, briefs, refresh
Section 4 Technical SEO................................................... Crawl, index, render, schema
Section 5 Off-Page SEO........................................................ Links, digital PR, authority
Section 6 Local SEO................................................................... GBP, citations, reviews
Section 7 AIO — AI Overviews & Google AI Search. AI Mode, fan-out, GSC AI reports
Section 8 AEO — Answer Engine Optimization............... Snippets, PAA, extractability
Section 9 GEO & LLMO................................................... Citations, crawlers, AI visibility
Section 10 SXO — Search Experience Optimization...... Engagement and conversion
Section 11 Metrics, KPIs & Measurement..................................... GSC, GA4, pipeline
Section 12 Tools.......................................................................... The practitioner stack
Section 13 Deprecated Terms & Myths....................................... What to stop saying
Section 14 References & Bibliography.................................................... 174 sources
Detailed contents
The index below is a live field. In Microsoft Word it fills automatically on open (or select all and press F9 to refresh). In other viewers it may show blank — use the section index above.
FOUNDATIONS & SEARCH FUNDAMENTALS...................................................... 7
CONTENT SEO & SEARCH INTENT...................................................................... 15
Standard terms — content and intent.......................................................... 16
Standard terms — crawling and indexing..................................................... 21
Standard terms — rendering and architecture............................................ 23
Standard terms — performance.................................................................... 25
Standard terms — structured data............................................................... 26
Standard terms — status codes, redirects, and infrastructure.................. 27
AIO: AI OVERVIEWS & GOOGLE'S AI SEARCH SURFACES................................. 38
AEO: ANSWER ENGINE OPTIMIZATION............................................................. 41
GEO & LLMO: GENERATIVE ENGINE & LLM OPTIMIZATION........................... 43
Standard terms — GEO and LLMO................................................................ 45
A note on GEO, AEO, LLMO, and AIO............................................................ 48
SXO: SEARCH EXPERIENCE OPTIMIZATION....................................................... 49
METRICS, KPIs & MEASUREMENT..................................................................... 51
Standard terms — Search Console................................................................ 52
Standard terms — GA4 and analytics........................................................... 53
Standard terms — ranking, visibility, and third-party metrics.................... 54
Myths — never were what they are claimed to be..................................... 59
A. Google — Search Central documentation............................................... 62
B. Google — crawling infrastructure (relocated late 2025)........................ 63
D. Google Search Central blog — dated announcements........................... 63
E. Google — Search Console, Analytics & Business Profile help................. 64
F. Web platform & performance................................................................... 64
G. AI platform documentation — crawlers and agents............................... 65
H. Infrastructure and crawler policy............................................................. 65
I. Academic and independent research........................................................ 65
M. Semrush, Moz, Yoast & other vendors.................................................... 67
N. Trade press and secondary reporting...................................................... 68
Section 1
FOUNDATIONS & SEARCH FUNDAMENTALS
The vocabulary everything else is built on. If a term appears in more than one discipline, it is defined here once and cross-referenced later.
CORE TERM
Search Engine Optimization (SEO)
Definition. The practice of improving a website's visibility in the unpaid results of search engines, by making content discoverable, understandable, and demonstrably more useful than the alternatives for a given query.
Why it matters. For B2B technology and services companies, organic search is typically the highest-intent, lowest-marginal-cost acquisition channel. Unlike paid media, its output compounds — a page that ranks continues to produce pipeline without incremental spend.
How to execute. SEO decomposes into four classical disciplines — technical (can the engine reach and understand the site), on-page (does this page match the query), content (do we cover the topic credibly and completely), and off-page (does the rest of the web treat us as authoritative) — plus the newer AI-search disciplines covered in Sections 7 to 10.
Common mistake. Treating SEO as a traffic function. Traffic without commercial intent is a cost, not a result. The measurable objective for a B2B services site is qualified pipeline, not sessions.
CORE TERM
Crawling, Indexing, and Ranking
Definition. The three sequential stages of how a search engine serves a result. Crawling is discovering and fetching URLs. Indexing is analysing the fetched content and storing it in a retrievable structure. Ranking is selecting and ordering indexed documents in response to a query.
Why it matters. These stages are strictly sequential, and diagnosis must follow the same order. A page that is not crawled cannot be indexed; a page that is not indexed cannot rank. Optimising the content of a page that Google has declined to index is wasted work.
How to execute. Diagnose in order using three independent data sources that routinely disagree: a crawler (what a bot can reach), Google Search Console's Page Indexing report (what Google chose to do), and server log files (what Googlebot actually did).
Common mistake. Jumping straight to content quality when the real problem is upstream. Always confirm indexation before rewriting anything.
CORE TERM
Search Intent
Definition. What a searcher is actually trying to accomplish with a query. Conventionally divided into four types: informational (seeking knowledge), navigational (seeking a specific site), commercial investigation (comparing options pre-purchase), and transactional (ready to act).
Why it matters. Intent determines format, depth, and conversion objective. It is also the single biggest determinant of whether traffic converts. A B2B services site ranking overwhelmingly for informational terms will generate impressive session counts and almost no pipeline.
How to execute. Never classify intent from the words alone — read the SERP. What ranks is Google's judgment about intent. A SERP full of vendor service pages means commercial intent regardless of how the query is phrased; a SERP full of forums means the query is unresolved.
Common mistake. Assuming "what is X" is informational. In mature B2B categories, definitional queries frequently return commercial SERPs, and the reverse is equally common.
CORE TERM
E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness)
Definition. The quality framework in Google's Search Quality Rater Guidelines, describing the characteristics of content Google's systems attempt to reward. The fourth "E" — Experience, meaning first-hand involvement with the subject — was added in December 2022 to the prior E-A-T.
Why it matters. For B2B technology firms selling expertise, E-E-A-T is the strategic bridge between what the company actually knows and what search systems can perceive. Trustworthiness is the dominant component; Google states the others contribute to trust rather than standing alone.
How to execute. Evidence it structurally: named authors with verifiable credentials and Person schema, an About page stating who the company is, first-hand delivery detail rather than aggregated public knowledge, sourced claims, and corroboration on third-party platforms.
Common mistake. Treating E-E-A-T as a ranking factor with a score. Google states plainly that "E-E-A-T itself isn't a specific ranking factor." It describes an objective Google's systems approximate through many signals — it is not a dial you can turn.
CORE TERM
SERP (Search Engine Results Page)
Definition. The page returned for a query. In 2026 it is a composite surface: AI Overviews, ads, local packs, featured snippets, People Also Ask, video carousels, discussion forums, and — usually well below all of it — the ten classical organic results.
Why it matters. "Position 1" no longer means what it did. The value of a ranking is a function of what else occupies the page. A first-position result beneath a full AI Overview may earn fewer clicks than a third-position result on a clean SERP.
How to execute. Audit SERP composition as part of keyword prioritisation. Export your ranking keywords with the SERP features each triggers and aggregate — the proportion of your keyword set carrying AI Overviews tells you how much of your visibility is exposed to answer-substitution.
Common mistake. Prioritising keywords by search volume without checking whether the clicks still exist.
Standard terms — foundations
|
Term |
Definition |
|
Algorithm |
The system of ranking signals and models a search engine uses to order results. Google's is not a single algorithm but many interacting ranking systems, several of which are publicly documented. |
|
Core update |
A broad, periodic change to Google's core ranking systems, announced publicly and typically taking one to two weeks to roll out fully. Recoveries are usually gradual and require substantive improvement, not tweaks. |
|
Ranking factor |
Any signal contributing to a document's position. Google has confirmed relatively few explicitly; most published "ranking factor" lists are correlation studies or inference. |
|
Ranking systems |
Google's own term for the named components of its ranking stack (e.g. helpful content signals, freshness systems, link analysis, page experience). Documented in the Ranking Systems Guide, which also lists retired systems. |
|
Organic results |
Unpaid search listings ordered by relevance and quality rather than by bid. |
|
Paid results / SERP ads |
Sponsored placements sold by auction, labelled as ads, occupying the most prominent SERP real estate on commercial queries. |
|
Query |
The text a user enters. Distinguished from keyword, which is the term an SEO targets — many queries can map to one keyword target. |
|
Keyword |
A term or phrase an SEO deliberately targets with a page. In practice a cluster label, not a literal string to be repeated. |
|
Long tail |
Low-volume, high-specificity queries which collectively represent the majority of all searches and typically convert at far higher rates than head terms. |
|
Head term |
A short, high-volume, broadly-defined query, usually highly competitive and ambiguous in intent. |
|
SERP feature |
Any non-standard result element — AI Overview, featured snippet, People Also Ask, local pack, video carousel, knowledge panel, discussion forum block. |
|
Knowledge Graph |
Google's database of entities and the relationships between them. The substrate behind knowledge panels and, increasingly, behind AI answer grounding. |
|
Knowledge panel |
The SERP box summarising a recognised entity. Cannot be created directly; influenced through Organization schema, sameAs references, and third-party corroboration. |
|
Entity |
A uniquely identifiable thing — a person, organisation, product, place, or concept — modelled independently of any particular string used to name it. |
|
Index |
The stored, retrievable representation of the web a search engine builds from crawled content. |
|
Googlebot |
Google's primary crawler, with distinct smartphone and desktop variants. Part of a broader crawler family including GoogleOther, Google-InspectionTool, Googlebot-News, and Storebot-Google. |
|
YMYL ("Your Money or Your Life") |
Topics that could materially affect health, financial stability, safety, or wellbeing, held to a higher quality bar. Definitions were clarified in the September 2025 Quality Rater Guidelines update. |
|
Search Quality Rater Guidelines (QRG) |
Google's published manual for the human raters who evaluate whether algorithm changes produce good results. Rater scores do not feed rankings directly — Google's own analogy is feedback cards at a restaurant. Current version: 11 September 2025. |
|
White hat / black hat / grey hat |
Informal descriptors for tactics that comply with, violate, or ambiguously skirt search engine guidelines. |
|
Google Search Essentials |
The current name for what were formerly the Webmaster Guidelines — Google's stated technical, spam, and quality requirements. |
|
Spam policies |
Google's enumerated prohibited practices, including cloaking, doorway abuse, keyword stuffing, link spam, scaled content abuse, site reputation abuse, expired domain abuse, and (from April 2026) back button hijacking. |
|
Manual action |
A human-applied penalty issued by Google's search quality team, reported in Search Console, requiring a documented fix and reconsideration request. |
|
Algorithmic demotion |
A ranking suppression applied automatically by a ranking system rather than by a human reviewer. No notification is given; recovery follows genuine improvement. |
|
Seasonality |
Predictable cyclical variation in search demand. A frequent cause of misattributed "declines" in month-on-month reporting; year-on-year comparison is the corrective. |
|
Search demand curve |
The distribution of query volume across a topic, from a few high-volume head terms to a very long tail of specific queries. |
Section 2
ON-PAGE SEO
Everything that happens within the four corners of a single page: the elements you control directly in the HTML and copy.
CORE TERM
Title element and title link
Definition. The <title> element is the HTML tag in the page head stating the page's topic. The title link is Google's own term for the clickable headline shown in a search result — which Google generates, sometimes from the <title> and sometimes from H1s, anchor text, or on-page copy.
Why it matters. It remains one of the strongest on-page relevance signals and the primary driver of click-through rate. On a SERP crowded with AI Overviews and ads, the title is often the only thing distinguishing your listing.
How to execute. Lead with the primary term, keep it under roughly 60 characters to avoid truncation, and — for B2B — replace superlatives with specifics. "Blockchain Development Company | 1,500+ Projects Delivered" outperforms "#1 Best Blockchain Development Company" because the claim is verifiable.
Common mistake. Treating the <title> as a guarantee of what displays. Google rewrites titles frequently, particularly when they are keyword-stuffed, boilerplate-heavy, or a poor match for the query. Rewrite rates are a useful diagnostic — if Google is consistently replacing yours, yours is worse than the alternatives it found on the page.
CORE TERM
Heading hierarchy (H1–H6)
Definition. HTML heading elements conveying document structure and topical outline.
Why it matters. Headings are how both readers and extraction systems navigate a long page. In an AI-search context, they take on a second job: a heading phrased as the question a user would actually ask makes the passage beneath it far more likely to be retrieved as an answer.
How to execute. Use one descriptive H1 matching the page's core subject, then nest logically. Phrase H2s and H3s as real questions or concrete statements ("How much does an AI development project cost?") rather than as labels ("Cost"). Never skip levels for styling reasons.
Common mistake. Enforcing a rigid "one H1 only, never more" rule as though it were a ranking factor. Google uses headings to understand structure; the requirement is logical nesting and descriptive text, not tag arithmetic.
CORE TERM
Internal linking
Definition. Hyperlinks between pages on the same domain. They drive crawl discovery, distribute link equity, and signal topical relationships between pages.
Why it matters. This is the largest lever a site owner fully controls, and on most large B2B sites it is badly misallocated. Authority typically concentrates in a handful of URLs — the homepage and two or three link-earning posts — while the commercial pages that need it sit four clicks deep with a handful of inbound internal links.
How to execute. Audit four things: crawl depth (are money pages within three clicks of the homepage), inlink count relative to commercial importance, anchor text descriptiveness, and link direction from high-authority pages down to pages that need authority. Then redistribute rather than simply adding more links.
Common mistake. Linking sideways between blog posts while never linking down into service pages. A high-authority blog post with no contextual link to the relevant service page is wasted equity.
CORE TERM
Image optimisation
Definition. The set of practices making images fast, accessible, and understandable — file format and compression, dimensions, descriptive filenames, alt attributes, and loading behaviour.
Why it matters. Images are usually the largest payload on a page and frequently the Largest Contentful Paint element, making them the dominant lever on load performance. Alt text is simultaneously an accessibility requirement and the primary way image content is understood.
How to execute. Serve WebP (or AVIF with a WebP fallback) at the displayed size; write alt text that describes the image in context; use descriptive filenames; lazy-load below-the-fold images and never the LCP image; set fetchpriority="high" on the hero image.
Common mistake. Keyword-stuffing alt text. Google's guidance is explicit that this may cause a site to be seen as spam — and it degrades the experience for screen reader users, which is what the attribute exists for.
Standard terms — on-page
|
Term |
Definition |
|
Meta description |
An HTML meta element giving a short summary of the page. Not a ranking factor; Google uses it only when it judges it a better match than on-page text. A click-through-rate lever. |
|
Snippet |
Google's term for the description portion of a search result. Generated dynamically and often varying by query, from either the meta description or the page body. |
|
URL slug |
The readable path segment identifying a page. Google recommends words over ID numbers and hyphens over underscores. |
|
Keyword placement |
Positioning the target term and its variants in high-signal locations — title, H1, opening paragraph, subheadings, URL, alt text. A practice, not a documented metric. |
|
Semantic keywords |
Terms and concepts topically related to the primary query that naturally co-occur in genuine coverage of a subject. The legitimate successor concept to "LSI keywords." |
|
Alt text |
The alt attribute describing an image for accessibility and image search. Should describe the image in the context of the page. |
|
Open Graph (OG) |
Meta protocol (og:title, og:description, og:image, og:url, og:type) controlling link previews on social platforms. Not a Google ranking signal. |
|
Twitter/X Cards |
twitter:card meta tags for X link previews. Now degraded — X strips headline text and removed its Card Validator; Open Graph tags are the primary requirement. |
|
Anchor text |
The visible clickable text of a hyperlink, providing context about the destination to both readers and search engines. Descriptive anchors outperform "click here." |
|
Descriptive anchor |
An anchor that states what the destination page is about. The default standard for internal linking. |
|
Content freshness |
Recency and currency of content. Google operates documented freshness systems that boost recency for queries where it matters — query-dependent, not a universal boost. |
|
Table of contents (TOC) |
An in-page index of section links. Improves navigation on long pages and can generate additional link options within the organic snippet. |
|
Jump links |
Fragment links (#section-id) to named anchors within a page, which Google may surface in a snippet so users land directly on the relevant section. |
|
Readability |
How easily the target audience can parse the text, measured by proxies such as sentence length, passive voice, and word complexity. Not a direct ranking factor; it affects engagement and comprehension. |
|
Thin content |
Pages lacking substantive value relative to their topic. The defect is absent value, not low word count — a 400-word page that fully answers a narrow question is not thin. |
|
Above the fold |
The viewport region visible without scrolling. Relevant via intrusive-layout demotions and via LCP, not as a standalone factor. |
|
Boilerplate |
Repeated template content (navigation, footers, standard CTAs) appearing across many pages. Excessive boilerplate relative to unique content is a duplication risk on templated B2B sites. |
|
Canonical content |
The single authoritative version of a piece of content, as distinct from duplicates, syndicated copies, or parameter variants. |
|
Breadcrumbs |
A navigational trail showing a page's position in the site hierarchy, markable with BreadcrumbList schema. Note: breadcrumb rich results have been desktop-only since January 2025. |
|
Schema markup (on-page) |
Structured data embedded in the page describing its entities and relationships. See Section 4 for the full treatment. |
|
CTA (call to action) |
The action a page asks the reader to take. On B2B service pages, the presence, specificity, and friction of the CTA is usually a larger conversion lever than any copy change. |
|
Conversion-focused content optimisation |
Optimising a ranking page for the action it should produce rather than only for the traffic it attracts — pricing transparency, proof, objection handling, and a low-friction next step. |
Section 3
CONTENT SEO & SEARCH INTENT
The discipline of deciding what to publish, in what format, at what depth, and when to retire it.
CORE TERM
Topic cluster (pillar and spoke)
Definition. A group of interconnected, thematically related pages: a pillar page covering a broad topic at overview depth, and cluster pages covering each subtopic in detail, all interlinked.
Why it matters. It is the practical mechanism for building topical authority. Search engines infer subject-matter competence from the completeness of coverage and the legibility of the structure connecting it — not from individual page quality alone.
How to execute. Bound the topic tightly enough to actually own it. A mid-sized firm cannot own "artificial intelligence"; it can own "RAG systems for regulated industries." Map every sub-entity and question inside that boundary, publish the whole map, and wire it with contextual internal links in both directions.
Common mistake. Building clusters around keyword volume rather than around a bounded subject. Two hundred articles spread across forty unrelated topics produce breadth without depth and convince no one of anything.
CORE TERM
Keyword cannibalisation
Definition. Multiple pages on one site competing for the same query, splitting relevance signals, links, and clicks so that no page performs as well as a single consolidated page would.
Why it matters. It is one of the most common and most fixable causes of stalled rankings on large sites — and it is invisible unless you look for it specifically, because each individual page looks fine in isolation.
How to execute. Detect it in Search Console: pull the Performance report by query and page, and flag every query where more than one URL receives meaningful impressions, or where the ranking URL has changed repeatedly. Then choose deliberately between consolidating (301 the weaker into the stronger), differentiating (rewrite one for a genuinely different intent), or demoting (strip the targeting and internal links from the weaker page).
Common mistake. Defaulting to consolidation. Two pages ranking at 4 and 8 sometimes hold more SERP real estate than one page at 3 would.
CORE TERM
Content refresh, consolidation, and pruning
Definition. The three interventions for existing content. Refresh updates a page in place, keeping its URL and history. Consolidation merges overlapping pages into one via 301 redirect. Pruning removes pages with no value, via redirect or 410.
Why it matters. On a large site, this work usually returns faster than new publishing, because the pages already have crawl history, links, and index presence. It also reclaims crawl budget and editorial capacity from content that is actively diluting site quality.
How to execute. Score every page on clicks, impressions, referring domains, internal inlinks, conversions, topical relevance to services sold, and factual currency. Refresh pages ranking 5–20 whose thesis still holds. Rewrite pages where the SERP format has moved. Retire pages failing every axis. Execute in waves of 200–300 URLs and measure between waves.
Common mistake. Pruning on a single metric and in one large batch. A page with no traffic but forty referring domains should have its links preserved via redirect, not deleted.
CORE TERM
Content brief
Definition. A specification handed to a writer covering target query, intent, format, required depth, subtopics, internal links, sources, proof points, FAQs, and conversion objective.
Why it matters. It is the mechanism that makes generic content impossible to submit. Most weak B2B content is weak because the writer was given a topic and a word count rather than a specification and research access.
How to execute. Include two fields most templates omit: the specific claims the piece must support with named sources or internal data, and the single action the reader should take with the section it appears in. Justify the format and depth with evidence from the current SERP rather than convention.
Common mistake. Specifying word count as the depth requirement. Depth is a function of what the query needs and what competitors already cover; word count is a symptom of that, not a target.
CORE TERM
Information gain
Definition. The concept that a document's value depends partly on the new information it adds relative to what the reader has already seen. Google holds a patent — "Contextual Estimation of Link Information Gain," granted June 2024 — describing such a mechanism.
Why it matters. In categories where every competitor has covered the same topics, the only durable differentiator is information nobody else has: proprietary delivery data, named case studies with real numbers, specific failure modes, and opinions that take a position.
How to execute. Build content on things competitors cannot copy quickly. Instead of "benefits of RAG systems," publish what broke across your last four regulated-industry RAG deployments and what it cost to fix. Instead of aggregating public statistics, publish the medians from your own project portfolio.
Common mistake. Treating "information gain score" as a documented Google ranking signal. The patent is real; its use in organic ranking is unconfirmed and the patent itself is framed around automated assistants. Treat it as a sound content heuristic, not a metric.
Standard terms — content and intent
|
Term |
Definition |
|
Informational intent |
The searcher wants knowledge or an answer to a question. |
|
Navigational intent |
The searcher wants a specific website or page. |
|
Commercial investigation intent |
The searcher is comparing options before deciding — reviews, comparisons, "best X," "X vs Y." The highest-value intent class for most B2B sites. |
|
Transactional intent |
The searcher is ready to act — buy, book, request a quote, start a trial. |
|
Mixed intent |
A query whose SERP shows several intent types simultaneously, indicating Google is hedging. Often an opportunity, since intent is unsettled. |
|
SERP format matching |
Aligning your page to the content type and format that already dominates the SERP, on the evidence that Google has established what satisfies the intent. |
|
Content hub |
A curated collection of content on one subject with a hub landing page. Overlaps with pillar/cluster; "hub" emphasises the navigational container. |
|
Content gap |
Topics or queries competitors rank for that you do not cover, identified by comparing keyword footprints. |
|
Content decay |
The gradual decline of a page's rankings and traffic over time as information ages, competitors improve, and intent shifts. Distinct from a sudden algorithmic drop. |
|
Striking distance |
Keywords already ranking just below meaningful traffic, conventionally positions 11–20. Usually the highest-return near-term opportunity on a large site. |
|
Topical authority |
A search engine's inferred confidence that a domain is a legitimate source on a subject, based on coverage completeness, internal structure, external signals, and demonstrable expertise. |
|
Editorial calendar |
The planned sequence of content production. Best prioritised by pipeline potential rather than publishing cadence. |
|
Content velocity |
The rate of publishing. A diagnostic input, never a KPI — it is the metric that produces large blogs and few rankings. |
|
Evergreen content |
Content whose relevance does not decay quickly. Distinguished from news or trend content requiring constant refresh. |
|
Skyscraper technique |
Identifying heavily-linked content, producing a substantially better version, and pitching it to the original linkers. |
|
People-first content |
Google's current framing for quality content guidance, replacing the retired Helpful Content System language. Emphasises E-E-A-T with trust as the dominant component. |
|
Scaled content abuse |
Google spam policy covering the mass generation of pages primarily to manipulate rankings. Method-agnostic — AI, human, or hybrid production all qualify. |
|
Site reputation abuse |
Google spam policy covering third-party content published on an established host site mainly to exploit that host's existing ranking signals. Commonly known in industry as "parasite SEO." |
|
Expired domain abuse |
Google spam policy covering the purchase of expired domains to repurpose their accumulated authority for low-value content. |
|
Doorway pages |
Google spam policy covering sets of near-identical pages built to funnel traffic from similar queries rather than to serve as destinations. The main compliance risk in scaled location-page programmes. |
|
Keyword stuffing |
Named Google spam policy: filling a page with keywords or numbers to manipulate rankings. |
|
Cloaking |
Named Google spam policy: presenting different content to users and search engines with intent to manipulate rankings. |
|
Back button hijacking |
Newest named Google spam policy (announced April 2026, enforced from June 2026): interfering with browser back navigation. Site owners are responsible for third-party ad code that causes it. |
|
Content QA |
Pre-publication review covering factual accuracy, originality, intent match, structure, on-page elements, internal links, and conversion path. |
|
SME interview |
A recorded conversation with an internal subject-matter expert used as the raw material for content. The most reliable source of differentiated technical detail in a services business. |
Section 4
TECHNICAL SEO
The infrastructure layer: whether search engines can reach, render, understand, and efficiently process the site. Note that Google moved its crawling documentation to a new location (developers.google.com/crawling) in late 2025 — older bookmarks and citations will 404.
CORE TERM
Crawl budget
Definition. Google's term for "the set of URLs that Google can and wants to crawl" — the product of crawl capacity limit (how much your server can take without degrading) and crawl demand (how much Google wants to crawl, based on inventory, popularity, and staleness).
Why it matters. It is a genuine constraint on sites with roughly 10,000+ URLs or heavy URL churn — which describes most enterprise B2B sites with large blogs. When crawl budget is consumed by low-value URLs, genuinely important pages are crawled less often and updated more slowly.
How to execute. Diagnose from server logs, not from crawl simulators — logs record what Googlebot did, simulators infer what it might do. Look for crawl waste on parameter URLs, faceted combinations, thin archives, and internal search results, then eliminate the URL patterns rather than the individual URLs.
Common mistake. Assuming 404s waste crawl budget. Google has stated explicitly that 4xx responses — with the single exception of 429 — do not. The real consumers are infinite URL spaces and low-value indexable pages.
CORE TERM
Indexation diagnosis
Definition. Determining how many of a site's URLs are indexed, and why the rest are not. Google's Page Indexing report (formerly called "Coverage") is the authoritative source.
Why it matters. Two of Google's status labels look similar and mean entirely different things. "Discovered — currently not indexed" means Google knows the URL exists but chose not to crawl it, usually signalling crawl capacity or low perceived value — a structural problem. "Crawled — currently not indexed" means Google looked and declined — an editorial problem.
How to execute. For the first, fix discoverability and value signals: internal links, reduced URL bloat, clean priority sitemaps. For the second, fix the content: it is thin, duplicative, or adds nothing over what is already indexed. Segment sitemaps by page type so the report tells you which templates are failing rather than giving a single site-wide percentage.
Common mistake. Responding to "Discovered — currently not indexed" by resubmitting URLs. Resubmission does not change Google's judgment about whether the URL is worth crawling.
CORE TERM
robots.txt
Definition. A file at the host root controlling crawler traffic — which user agents may request which URL patterns.
Why it matters. It is the most consequential single file on a site and the most commonly misunderstood. Google is explicit that robots.txt "is not a mechanism for keeping a web page out of Google" — a blocked URL can still be indexed without a snippet if it is linked externally.
How to execute. Use it to prevent crawl waste on URL patterns you do not want fetched at all. Use noindex (which requires the page to remain crawlable) to keep pages out of the index. Never combine the two on the same URL — if the page is blocked, Google can never crawl it to see the noindex.
Common mistake. Two classics. First, blocking pages to remove them from the index, then finding them still listed months later as "No information is available for this page." Second, shipping a staging environment's Disallow: / to production. The second is the single most damaging SEO incident that routinely occurs, and it is preventable with one automated daily check.
CORE TERM
Canonicalisation
Definition. The process of designating one preferred URL among duplicates and consolidating their signals into it. rel="canonical" is the primary annotation, alongside redirects, sitemaps, internal linking, and HTTPS preference.
Why it matters. Duplicate content is not a penalty — it is a dilution problem. When signals split across multiple URLs serving the same content, none accumulates enough to rank well, and Google's canonical choice may not be yours.
How to execute. Set the canonical in the original HTML source, not via JavaScript — Google's December 2025 guidance is explicit on this. Ensure self-referencing canonicals on paginated pages (never canonicalise page 2 to page 1). Check Search Console for "Google chose different canonical than user," which tells you your hint was overruled and why.
Common mistake. Treating rel="canonical" as a directive. It is a hint, weighed against every other canonicalisation signal on the site. If your internal links, sitemap, and redirects all point elsewhere, the canonical tag loses.
CORE TERM
Core Web Vitals
Definition. Google's three field-measured page experience metrics, stable since 2024 and unchanged in 2026: LCP (Largest Contentful Paint — perceived load speed, good ≤ 2.5s), INP (Interaction to Next Paint — responsiveness, good ≤ 200ms), and CLS (Cumulative Layout Shift — visual stability, good ≤ 0.1). Assessed at the 75th percentile of page loads, segmented by device.
Why it matters. They are a genuine but modest ranking input, and a substantial conversion input. On B2B sites the commercial case is usually the stronger one — a service page that takes five seconds to render loses buyers before it loses rankings.
How to execute. Work from field data (CrUX), not lab scores — Lighthouse tells you what could be slow, CrUX tells you what is slow for real users, and that is what Google uses. Segment by template. The usual wins: serve the LCP image in a modern format at the right size with fetchpriority="high" and no lazy-loading, defer non-essential third-party scripts, and reserve explicit dimensions for late-loading elements.
Common mistake. Chasing a Lighthouse score of 100. Lab performance scores are not Core Web Vitals assessment and do not determine the page experience signal. Also note that INP replaced First Input Delay (FID) on 12 March 2024 — any glossary still listing FID as a Core Web Vital is out of date.
CORE TERM
Structured data and rich results
Definition. Machine-readable markup (preferably JSON-LD, using the schema.org vocabulary) describing a page's entities and their relationships. A rich result is an enhanced search listing driven by that markup.
Why it matters. Google removed more rich result types between 2023 and 2026 than in the preceding five years, which has changed the value proposition. Structured data is now less about SERP decoration and more about entity clarity — making it unambiguous to both classical retrieval and AI systems what your organisation is, who your experts are, and how your content relates.
How to execute. For a B2B technology firm, prioritise: Organization (with sameAs pointing to LinkedIn, Crunchbase, GitHub, and review platforms), Person/ProfilePage for named experts, Article with real authors and dates, BreadcrumbList, and SoftwareApplication for productised offerings. Validate eligibility with the Rich Results Test and syntax with the Schema Markup Validator — they answer different questions.
Common mistake. Two specific and widespread errors. `Service` schema produces no Google rich result — it appears nowhere in the search gallery, and promising clients SERP features from it is an overpromise. And self-serving review markup is prohibited: a business marking up reviews of itself, including via embedded third-party widgets, is ineligible for star ratings.
CORE TERM
JavaScript rendering
Definition. The process by which Google executes a page's JavaScript before indexing it. Google crawls raw HTML first, extracts href links, then queues the page for rendering in its Web Rendering Service — an evergreen headless Chromium.
Why it matters. Content that exists only after hydration is indexed on a delay, inconsistently, and at crawl-budget cost. More importantly in 2026: most AI crawlers do not execute JavaScript at all, so client-side-rendered content is effectively invisible to AI retrieval even when Google eventually indexes it.
How to execute. Establish whether it is actually a problem by comparing the raw HTML against the rendered DOM. If the H1, body copy, internal links, canonical, and structured data are all in the source, rendering is not your issue. If not, prefer server-side rendering or static generation; ensure internal links are real <a href> elements, since Googlebot does not click buttons or scroll.
Common mistake. Relying on dynamic rendering — serving prerendered HTML to bots and JavaScript to users. Google formally labelled this a deprecated workaround in early 2024. Use SSR or SSG instead.
Standard terms — crawling and indexing
|
Term |
Definition |
|
Crawl capacity limit |
Also called hostload: the ceiling on how much time your server will hold connections open for Googlebot. Current Google terminology; "crawl rate limit" is legacy. |
|
Crawl demand |
How much Google wants to crawl a host, driven by inventory size, URL popularity, and how often content genuinely changes. Fake freshness signals do not raise it. |
|
Crawl depth (click depth) |
The number of internal clicks from the homepage needed to reach a URL. Deeper pages are crawled less often and receive less internal equity. |
|
Orphan page |
A page with no internal links pointing to it. Found by diffing a crawl against sitemaps, logs, analytics, and Search Console data. |
|
Log file analysis |
Parsing raw server access logs to see actual bot behaviour — which URLs Googlebot hit and how often, crawl waste, response codes encountered, and URLs never crawled. The only true record of crawling. |
|
Index bloat |
A large share of low-value pages in the index (parameter URLs, thin tag archives, internal search results), diluting quality signals and wasting crawl resources. |
|
Indexation rate |
The ratio of indexed URLs to submitted or crawlable URLs. A diagnostic ratio, not a Google metric. |
|
Soft 404 |
An error or empty page served with a 200 status code. Google detects these heuristically and excludes them from the index. |
|
noindex |
A robots meta tag or X-Robots-Tag rule preventing indexing. The page must remain crawlable for it to work. Note: Google states JavaScript-injected noindex has undefined behaviour and should be avoided. |
|
X-Robots-Tag |
An HTTP response header carrying the same rules as the robots meta tag. Essential for non-HTML resources like PDFs and feeds. |
|
Robots directives |
The current supported set: noindex, nofollow, none, nosnippet, max-snippet, max-image-preview, max-video-preview, notranslate, noimageindex, unavailable_after, indexifembedded. Note: noarchive, nocache, and nositelinkssearchbox are no longer used by Google. |
|
XML sitemap |
A machine-readable list of URLs. Hard limits of 50MB uncompressed or 50,000 URLs per file. Google explicitly ignores <priority> and <changefreq>. |
|
Sitemap index |
A sitemap of sitemaps, used above the 50,000-URL limit. Segmenting by page type turns sitemap reporting into a per-template indexation diagnostic. |
|
lastmod |
The timestamp of a page's last significant update. Google uses it only if it is consistently and verifiably accurate — touching it site-wide nightly teaches Google to ignore the field. |
|
IndexNow |
An open protocol for instantly notifying search engines of URL changes. Supported by Bing, Yandex, and Naver. Google has never adopted it. |
|
Duplicate content |
Multiple URLs serving identical or near-identical content. Not a penalty; it causes signal dilution and unpredictable canonical selection. |
|
Near-duplicate content |
Pages that are substantially but not entirely identical — the normal outcome of heavy template reuse on B2B industry and location pages. |
|
Parameter handling |
Management of query-string URLs. Search Console's URL Parameters tool was retired in 2022; current options are robots.txt patterns, canonicals, or not generating the parameters. |
|
Faceted navigation |
Filter and sort interfaces generating a URL per attribute combination. Google names two harms: overcrawling of effectively infinite URL spaces, and slower discovery of genuinely new content. |
|
Crawl trap / infinite space |
Any URL pattern generating unlimited unique URLs — faceted combinations, calendar links, session IDs, relative-path loops. |
|
Pagination |
Splitting a sequence across URLs. Current rules: unique crawlable URL per page, self-referencing canonicals, no URL fragments for page numbers. |
|
rel="next" / rel="prev" |
Obsolete. Google's documentation states plainly that it no longer uses these tags, though other engines may. |
|
Infinite scroll |
Content loaded on scroll or button click. Because Googlebot follows href attributes rather than scrolling or clicking, JS-only loading hides content from discovery unless a paginated URL fallback exists. |
Standard terms — rendering and architecture
|
Term |
Definition |
|
Crawl → render → index pipeline |
Google's three sequential phases: fetch HTML and extract links, queue for rendering, index the rendered result. |
|
Two-wave indexing |
The informal name for that pipeline — a first pass on raw HTML, a second after JavaScript execution. Still directionally accurate, though the render delay is now typically seconds to minutes. |
|
Web Rendering Service (WRS) |
Google's headless Chromium rendering infrastructure, kept current with stable Chrome. Note that pages returning non-200 status codes may not be sent to rendering at all. |
|
Client-side rendering (CSR) |
An empty HTML shell with content assembled in the browser. The highest-risk pattern for both classical and AI search. |
|
Server-side rendering (SSR) |
HTML generated per request on the server, so crawlers receive complete markup in the first response. Google's recommended approach. |
|
Static site generation (SSG) |
HTML generated at build time and served as static files. Fastest TTFB and fully crawlable. |
|
Hydration |
Attaching JavaScript behaviour to server-rendered HTML in the browser. A leading cause of poor INP, since it blocks the main thread. |
|
Dynamic rendering |
Serving prerendered HTML to bots and JavaScript to users. Formally deprecated by Google as a workaround; use SSR or SSG. |
|
Flat vs deep architecture |
Flat keeps most pages within about three clicks of the homepage, aiding discovery and equity distribution. Deep nests content in long hierarchies, reducing crawl frequency to leaf pages. |
|
Siloing |
Grouping topically related pages via internal links to create coherent clusters that concentrate relevance within a subject area. |
|
Hub and spoke |
An architectural pattern where a central hub links to detail pages and each links back. The structural implementation of a topic cluster. |
|
PageRank |
The formula judging a page's importance by the quantity and quality of pages linking to it. The public toolbar metric was retired in 2016; a heavily evolved version remains internal. |
|
PageRank sculpting |
The obsolete practice of using nofollow on internal links to funnel equity. Broken since Google's 2009 change (equity evaporates rather than redirecting) and further undermined when nofollow became a hint in 2020. Historical literacy only. |
|
Subdomain vs subdirectory |
blog.example.com versus example.com/blog. Both are valid; the practical argument favours subdirectories because signals consolidate on one host and crawl capacity is assessed per host. |
|
hreflang |
Annotation declaring language and regional variants of a page, implementable via HTML <link>, HTTP header, or sitemap. |
|
hreflang return links |
The requirement that every variant links back to every other. Google is unambiguous: if two pages do not both point to each other, the tags are ignored. The most common hreflang failure. |
|
x-default |
The reserved hreflang value used when no other language or region matches — typically a global landing page or language selector. |
|
International SEO |
Targeting multiple languages and markets via hreflang, URL structure, localised content, and regional infrastructure. |
|
ccTLD |
A country-code top-level domain (.de, .in, .co.uk). The strongest geo-targeting signal, at the cost of splitting authority across domains. |
Standard terms — performance
|
Term |
Definition |
|
LCP — Largest Contentful Paint |
Time until the largest text block or image in the viewport renders. Good ≤ 2.5s, poor > 4.0s. |
|
INP — Interaction to Next Paint |
Responsiveness across all qualifying interactions, reporting roughly the worst input-to-paint latency. Good ≤ 200ms, poor > 500ms. Replaced FID in March 2024. |
|
CLS — Cumulative Layout Shift |
A unitless score for unexpected visual movement during the page lifespan. Good ≤ 0.1, poor > 0.25. |
|
TTFB — Time to First Byte |
Time from request to first response byte. Not a Core Web Vital; a key diagnostic for LCP capturing server, redirect, and network cost. |
|
FCP — First Contentful Paint |
Time until any content first renders. Largely governed by render-blocking resources. |
|
TBT — Total Blocking Time |
Total main-thread blocking between FCP and interactivity. A lab-only metric — which is why it is not a Core Web Vital — but the best lab proxy for INP. |
|
Field data (RUM) |
Real User Monitoring: performance as actually experienced by visitors. This is what Google uses. |
|
Lab data (synthetic) |
Performance measured in a controlled simulation. Reproducible and useful for debugging; does not reflect the real user distribution. |
|
CrUX (Chrome User Experience Report) |
Google's public dataset of real Chrome user experience. The field-data source behind Core Web Vitals. Requires sufficient traffic and public discoverability to qualify. |
|
Render-blocking resources |
CSS and synchronous JavaScript in the <head> that must be processed before first paint. The dominant cause of poor FCP and LCP. |
|
Critical CSS |
The minimal CSS needed to style above-the-fold content, inlined so first paint does not wait on an external stylesheet. |
|
Lazy loading |
Deferring off-screen resource loading via loading="lazy". Critical caveat: never lazy-load the LCP image. |
|
fetchpriority |
An HTML hint setting a resource's relative download priority. The standard LCP fix is fetchpriority="high" on the hero image, which otherwise starts at low priority. |
|
preload |
Lets the browser discover a resource early, but still fetches at default priority. Complementary to fetchpriority, not an alternative. |
|
preconnect |
Performs DNS, TCP, and TLS setup to a third-party origin ahead of time, removing connection latency from the critical path. Use sparingly. |
|
WebP |
Image format roughly 25–35% smaller than JPEG at equivalent quality, with universal browser support. The safe default. |
|
AVIF |
AV1-based format delivering a further 20–30% saving over WebP. Best served via <picture> with a WebP fallback. |
|
HTTP/2 |
Multiplexed protocol removing HTTP-layer head-of-line blocking. Googlebot crawls over HTTP/1.1 and HTTP/2, choosing whichever performs better. |
|
HTTP/3 (QUIC) |
UDP-based protocol with faster connection establishment. Googlebot does not crawl over HTTP/3 as of 2026 — it benefits users, not crawlers. |
Standard terms — structured data
|
Term |
Definition |
|
schema.org |
The cross-engine vocabulary defining types and properties for describing entities. Note: schema.org validity does not equal Google rich result eligibility. |
|
JSON-LD |
JavaScript Object Notation for Linked Data — a script block carrying structured data separately from visible markup. Google's recommended format. |
|
Microdata |
Structured data expressed via inline HTML attributes. Supported but brittle under template changes. |
|
RDFa |
An HTML5 attribute-based linked data format. Supported by Google, rarely used. |
|
Organization schema |
Markup disambiguating a business entity, supporting name, url, logo, sameAs, address, contactPoint, foundingDate, numberOfEmployees, and identifiers. The highest-value schema type for B2B. |
|
sameAs |
The property linking an entity to its authoritative profiles elsewhere. The primary mechanism for entity disambiguation. |
|
Person / ProfilePage |
Markup describing an individual — author, executive, engineer. Valuable for author entity building and evidencing expertise. |
|
Article schema |
Markup for editorial content supplying headline, images, author, and dates. |
|
BreadcrumbList |
Markup replacing the URL in the SERP with a breadcrumb trail. Desktop-only since January 2025. |
|
SoftwareApplication |
Markup for apps and software including category, operating system, offers, and rating. Relevant for productised B2B offerings. |
|
VideoObject |
Markup enabling video rich results, key moments, and video indexing. |
|
Dataset |
Markup describing a structured data collection, surfacing in Google Dataset Search. |
|
Rich Results Test |
Google's tool testing whether a URL is eligible for a Google rich result, using live rendering. Only reports on types Google currently supports. |
|
Schema Markup Validator |
schema.org's own validator, checking markup against the full vocabulary regardless of Google support. Use it for syntax; use the Rich Results Test for eligibility. |
|
FAQPage |
CHANGED · Rich result retired. FAQ rich results stopped appearing in May 2026, with reporting and Rich Results Test support dropped in June 2026. The schema.org type remains valid and may aid entity and LLM comprehension, but it produces zero Google SERP enhancement. |
|
HowTo |
CHANGED · Removed. HowTo rich results were dropped in September 2023 and the documentation deleted. |
|
WebSite / SearchAction |
CHANGED · Deprecated. Previously powered the sitelinks search box, which Google dropped in November 2024. Harmless for entity definition; produces no SERP feature. |
|
Service schema |
CHANGED · Valid schema.org vocabulary, but Google has no `Service` rich result. Useful for entity clarity and AI parsing only. |
Standard terms — status codes, redirects, and infrastructure
|
Term |
Definition |
|
200 OK |
Success. Content passes to the indexing pipeline, though indexing is never guaranteed. Only 200-status pages are reliably queued for rendering. |
|
301 Moved Permanently |
Permanent redirect and a strong canonicalisation signal. The standard method for consolidating URLs. |
|
302 Found |
Temporary redirect and a weak canonicalisation signal. Misuse in permanent migrations is a classic error. |
|
307 / 308 |
Method-preserving temporary and permanent redirects respectively. 308 is functionally equivalent to 301 for Google's purposes. |
|
404 Not Found |
Resource absent. Previously indexed URLs are dropped and crawl frequency decays gradually. |
|
410 Gone |
Explicit permanent removal. Handled like 404 but can trigger slightly faster de-indexing. The correct code for deliberate deletion. |
|
429 Too Many Requests |
Rate limiting. The one 4xx code Google treats as a server overload signal, causing it to slow crawling. |
|
451 Unavailable For Legal Reasons |
Content blocked for legal reasons. Handled within Google's 4xx family. |
|
500 / 503 |
Server error and temporary unavailability. 503 with a Retry-After header is the correct response for planned maintenance — it signals back-off without de-indexing. |
|
Redirect chain |
Two or more sequential redirects. Google follows up to ten hops; chains waste crawl capacity and add user latency. |
|
Redirect loop |
A cycle of redirects that never resolves. The URL is abandoned and nothing is indexed. |
|
Meta refresh |
A client-side redirect via meta tag. Google can follow it but treats it as weak; a server-side 301 is always preferable. |
|
HTTPS |
TLS-encrypted HTTP. A confirmed lightweight ranking signal since 2014 and a default canonicalisation preference. |
|
HSTS |
A response header instructing browsers to use HTTPS for a domain. Removes the redirect hop for repeat visitors; browser-enforced, so it does not change crawler behaviour. |
|
Mixed content |
Subresources loaded over HTTP on an HTTPS page. Active mixed content is blocked by modern browsers, which can break rendering and therefore indexing. |
|
CDN |
Geographically distributed edge servers caching content close to users. The primary SEO effect is improved TTFB and LCP. |
|
Edge SEO |
Implementing technical SEO changes through serverless code at the CDN layer — redirects, header injection, robots.txt, hreflang, meta overwrites — rather than in the origin codebase. Typical latency cost is around 10ms. |
|
Server-side testing |
Running A/B tests at the server or edge so crawlers receive fully-formed HTML. Avoids the rendering and cloaking risks of client-side testing tools; requires stable bucketing for Googlebot. |
|
Bot verification |
Confirming that a request claiming to be Googlebot genuinely originates from Google, via reverse DNS or published IP ranges. Essential before acting on log data or applying bot-specific edge rules. |
|
Crawl-diff testing |
Comparing a staging crawl against production before release to surface unintended changes to titles, canonicals, meta robots, and internal links. The highest-value regression control available. |
Section 5
OFF-PAGE SEO
Signals originating outside your domain: what the rest of the web says about you, links to you, and how it corroborates who you are.
CORE TERM
Backlinks and link equity
Definition. A backlink is a link from one site to another. Link equity is the authority transferred through it, modulated by the linking page's own authority, its placement (editorial main-content links transfer more than footer links), the anchor context, and crawlability.
Why it matters. Links remain one of the strongest external signals, and for competitive B2B commercial terms they are frequently the deciding factor between two otherwise comparable pages.
How to execute. Measure referring domains, not total backlinks — fifteen thousand backlinks may come from five thousand domains, and the domain count is the meaningful figure. Assess prospective linking domains on topical relevance, genuine organic traffic, and editorial standards rather than on a single vendor authority score.
Common mistake. Building links to the homepage. On a B2B site the pages that need external authority are the commercial service pages, and they are usually the ones with the fewest.
CORE TERM
rel attributes: nofollow, sponsored, ugc
Definition. Link qualifiers telling Google how to treat a link. rel="nofollow" (2005) is the general-purpose "don't associate this with us." rel="sponsored" and rel="ugc" were introduced in September 2019 for paid placements and user-generated content respectively.
Why it matters. Since March 2020, Google treats all three as hints, not directives — it may still crawl and count them. Industry treatment of nofollow as an absolute block is outdated by six years.
How to execute. Use sponsored for anything compensated — Google prefers it over nofollow for paid content. Use ugc for comments and forum posts. Values are combinable (rel="ugc nofollow").
Common mistake. Writing rel="dofollow". It is not a real attribute and never has been in any specification — an ordinary link is followed by default.
CORE TERM
Digital PR and linkable assets
Definition. Earning editorial coverage and links by giving publishers something genuinely worth covering — original research, proprietary data, tools, or expert commentary. A linkable asset is content built specifically to attract links.
Why it matters. It is the dominant white-hat link acquisition method in 2026, and it has become doubly valuable: the third-party coverage it produces is also what AI answer engines retrieve when asked category-level questions.
How to execute. For a services business, the highest-yield asset is almost always original data from your own delivery portfolio — median timelines by project type, the most common causes of overrun with frequencies, cost ranges. Nobody else has it, and journalists need citable statistics.
Common mistake. Relying on HARO-style journalist matching without checking the landscape. HARO, rebranded as Connectively, shut down on 9 December 2024 with no single official successor. Current alternatives include Featured, Qwoted, ProfNet, and Muck Rack — any 2026 guide recommending "HARO link building" is recycling stale material.
CORE TERM
The disavow tool
Definition. A Search Console feature asking Google to ignore specified inbound links.
Why it matters. This is one of the widest gaps between Google's stated position and common agency practice. Google's own guidance: "In most cases, Google can assess which links to trust without additional guidance, so most sites will not need to use this tool," and "if used incorrectly, this feature can potentially harm your site's performance."
How to execute. Google restricts appropriate use to sites with both a substantial volume of spammy inbound links and a manual action or imminent risk of one — and advises attempting physical link removal first.
Common mistake. Routine monthly "disavow hygiene" driven by a tool's toxicity score. Google does not recognise the concept of "toxic links" — toxicity scores are vendor heuristics with no Google counterpart, and acting on them disavows links that were doing no harm and may have been helping.
Standard terms — links
|
Term |
Definition |
|
Referring domain |
The count of unique domains linking to a site, as opposed to total backlink count. The more meaningful of the two. |
|
Editorial link |
A link given without being requested or paid for. The only category Google's guidelines fully endorse. |
|
Link profile |
The full assessment of a site's backlinks across quantity, quality, diversity, anchor distribution, and topical relevance. |
|
Link velocity |
The rate at which a site gains or loses backlinks. Not a confirmed positive ranking factor — velocity patterns feed spam detection rather than acting as a signal of quality. |
|
Link decay / link rot |
Progressive loss of existing backlinks as linking pages are deleted or restructured without redirects. |
|
Link reclamation |
Converting unlinked brand mentions into links and repairing lost or broken inbound links. Usually the cheapest link source available. |
|
Unlinked brand mention |
A reference to a brand, product, or person in online content without a backlink. Never confirmed as a classical ranking signal, but increasingly associated with AI-search visibility. |
|
Exact match anchor |
Anchor text identical to the target keyword. By definition can only target one term. |
|
Phrase match anchor |
The target keyword appearing verbatim within a longer anchor. |
|
Partial match anchor |
All query words present but not as a contiguous phrase. |
|
Branded anchor |
Anchor text that is the brand or company name. The dominant natural anchor type in a healthy profile. |
|
Naked URL anchor |
The raw URL string used as anchor text. |
|
Generic anchor |
Non-descriptive phrasing — "click here," "this site," "read more." |
|
Image anchor |
Where the link is an image, its alt text serves as the anchor. |
|
Anchor text over-optimisation |
An unnaturally high concentration of exact-match anchors. A primary Penguin target — and note that Ahrefs' correlation work found only weak association between exact-match anchors and rankings, so engineering anchor ratios is risk without demonstrated upside. |
|
Broken link building |
Finding dead outbound links on relevant pages and offering your live resource as a replacement. |
|
Resource page link building |
Pitching inclusion on curated "useful resources" pages within your topic area. |
|
Guest posting |
Publishing on a third-party site. Legitimate when genuinely editorial; a spam-policy violation when done at scale for links with optimised anchors. |
|
Link spam |
Google's definition: creating links to or from a site primarily to manipulate rankings. |
|
Link scheme |
Umbrella term for prohibited practices: buying or selling links for ranking purposes, excessive exchanges, automated link creation, and unqualified text ads. |
|
Paid links |
CHANGED · Nuance frequently misreported. Google states buying and selling links is a normal part of the economy — the violation is failing to qualify the link with sponsored or nofollow. Paid links are permitted; unqualified paid links are not. |
|
PBN (private blog network) |
A network of sites created solely to link out and improve another site's rankings. A link scheme under Google's policies. |
|
Link exchange |
Reciprocal linking. Prohibited only when excessive — Google's wording targets "excessive link exchanges" and partner pages created solely for cross-linking. |
|
Advertorial / native advertising |
Paid articles carrying ranking-credit links. Prohibited unless the links are qualified. |
|
Domain Authority (DA) |
Moz's 1–100 logarithmic score predicting relative ranking likelihood. Not a Google metric and not visible to Google. Comparative only. |
|
Domain Rating (DR) |
Ahrefs' 0–100 logarithmic measure of backlink profile strength only. Nofollow links do not contribute. The page-level equivalent is URL Rating (UR). |
|
Authority Score (AS) |
Semrush's 1–100 compound metric combining link power, estimated organic traffic, and spam-factor checks. Broader than DA or DR because it includes traffic. |
|
Brand signals |
Aggregate indicators of brand recognition — branded search volume, branded anchors, linked and unlinked mentions, entity corroboration. |
|
Co-citation |
Two entities being mentioned by the same third-party sources without linking to each other, interpreted as an associative relevance signal. Never confirmed by Google. |
|
Co-occurrence |
Related terms appearing in proximity across many sites covering a topic. Never confirmed by Google. |
|
Entity building |
Establishing an unambiguous, machine-readable identity for a brand or person across the web — consistent naming, Organization schema with sameAs, and corroborating third-party profiles. |
|
Digital PR metrics |
Placements secured, referring domains earned, link-to-placement ratio, publication reach, branded search lift, and referral traffic. |
Section 6
LOCAL SEO
Optimising for queries with geographic intent and for the map-based surfaces that serve them. For a B2B services firm with multiple delivery offices, this discipline governs how those locations appear — and it changed substantially in 2025–26.
CORE TERM
Google Business Profile (GBP)
Definition. Google's free business listing product powering presence in Search, Maps, and the local pack. Renamed from Google My Business in November 2021; "Google Business Profile" remains the correct name in 2026.
Why it matters. For any query with local intent, the profile — not the website — is the primary ranking asset. Google's own stated ranking factors for local results are relevance, distance, and prominence, and the profile is the main lever on the first and third.
How to execute. Get the primary category exactly right — it is the single strongest relevance lever, and Google advises choosing the most specific option available. Complete every applicable field, keep hours accurate, publish Posts, and maintain a steady review cadence.
Common mistake. Following advice to "seed your GBP Q&A." The Q&A feature was discontinued — Google announced it in September 2025, the API was cut off in November 2025, and it was removed from profiles by early 2026. Any 2026 checklist including it is recycling old material.
CORE TERM
Local ranking factors: relevance, distance, prominence
Definition. Google's three publicly stated local ranking factors. Relevance is how well a profile matches the search. Distance is how far the business is from the searcher (or from the location term used). Prominence is how well-known the business is — Google explicitly names review count and rating as inputs.
Why it matters. Only two of the three are meaningfully influenceable. Distance is largely fixed, which sets a hard ceiling on how far a single location can rank and explains why grid-based rank tracking shows a business at position 3 at its own address and outside the top 10 a short drive away.
How to execute. Optimise relevance through categories, services, and profile content; optimise prominence through reviews, local links, and citations. Accept distance as a constraint and plan location coverage accordingly rather than trying to defeat it.
Common mistake. Chasing the "centroid" — the geographic centre of a city — as a ranking anchor. Google has never documented it, and modern evidence favours searcher-device proximity.
CORE TERM
NAP consistency and citations
Definition. NAP is Name, Address, Phone — the identity triplet in a local listing. A citation is any online mention of that data. Structured citations appear in formal directories with dedicated NAP fields; unstructured citations appear in free-form content such as local news or sponsorship pages.
Why it matters. Consistency lets search engines confidently resolve every mention to one entity. Inconsistency fragments the entity, which weakens prominence and can produce duplicate listings.
How to execute. Establish the canonical NAP format first, then audit and correct across the major directories and data aggregators. Active US aggregators are Foursquare Places, Data Axle, TransUnion Digital Business Profile (formerly Neustar Localeze), GPS Network, and YP Network.
Common mistake. Working from a stale aggregator checklist. Data Axle's Express Update is defunct, and Localeze has been renamed — both still appear in widely-circulated guides.
CORE TERM
Location pages and service area pages
Definition. A location page represents a physical branch on your own site, carrying that branch's NAP, hours, staff, and local content. A service area page targets a locality the business travels to rather than occupies.
Why it matters. They are the main mechanism for ranking in the organic results beneath the local pack, and the main compliance risk in local SEO.
How to execute. Each page must contain information that only applies to that location — local team, local case studies, local regulatory context, local partners. A store locator must expose crawlable HTML links to each page, since JavaScript-only locators routinely leave location pages undiscovered.
Common mistake. Generating hundreds of near-identical city pages by find-and-replace. These meet Google's definition of doorway pages — a named spam policy — and are the single largest compliance exposure in scaled local programmes.
CORE TERM
Review policy (2026 changes)
Definition. Google's Rating Manipulation policy governs how businesses may solicit and handle reviews.
Why it matters. Effective 17 April 2026, Google added two prohibitions that invalidate near-universal industry practices: merchants may not ask staff to solicit a set number of reviews (review quotas), and may not request reviews containing specific content, including naming a staff member.
How to execute. Genuine, un-incentivised review requests by email, WhatsApp, or QR code remain permitted. Remove technician review bonuses and quotas, and remove "please mention your technician by name" from request scripts.
Common mistake. Continuing to run staff review incentive programmes because they were standard practice for a decade. They are now explicitly prohibited, alongside the longer-standing ban on offering payment, discounts, or free goods in exchange for reviews.
Standard terms — local
|
Term |
Definition |
|
Local pack / map pack / 3-pack |
The set of local business listings (usually three) shown above organic results when Google detects local intent. In 2026 it appears under two labels — "Places" and the newer "Businesses" variant. |
|
Local Finder |
The expanded local results interface reached by clicking through from the local pack. Shows a longer ranked list with filters; distinct from Google Maps proper. |
|
Local justification |
The snippet Google appends to a local pack result explaining why it matched — a quoted review phrase, "Their website mentions…", or a matching service. |
|
Proximity bias |
The disproportionate weight distance carries in local pack ranking. Grid testing consistently shows dramatic rank decay over short distances. |
|
Local intent |
Query intent implying a nearby result is wanted. Explicit intent carries a geo-modifier; implicit intent is a category alone, where Google infers location from IP, account data, GPS, or history. |
|
Geo-modifier |
A location term appended to a query — city, neighbourhood, region, or "near me." |
|
"Near me" search |
A query with an explicit proximity modifier. Users add it when Google's default radius is too wide or when travelling. |
|
Localised SERP |
A results page whose composition changes with searcher location for an identical query string. |
|
GBP primary category |
The single category best describing the business's core offering. The strongest single relevance lever in the profile. |
|
GBP additional categories |
Secondary categories describing distinct departments or services. Google explicitly warns against selecting a category for every service. Custom categories cannot be created. |
|
GBP Posts (Updates) |
Short-form content published to a profile in three types — Updates, Offers, and Events. Posts auto-archive after six months unless a date range is set. |
|
GBP Attributes |
Structured descriptors covering accessibility and business identity (women-owned, veteran-owned, LGBTQ+ owned). Some are owner-editable; others are populated from customer input and cannot be changed. |
|
GBP Products / Services |
Itemised catalogue modules. Services are category-gated from a predefined list plus custom entries; Products render as a browsable carousel. |
|
Service Area Business (SAB) |
A business travelling to customers rather than serving them at a storefront. Must hide its address if operating from a residence, and Google states the total service area should not extend beyond about two hours' driving time from base. |
|
Hybrid business |
A location both serving walk-in customers at its address and travelling to customers. Must be staffed during posted hours. |
|
Multi-location SEO |
Optimising many location or service-area pages at scale without cannibalisation or duplication. |
|
Store locator |
A search or map interface routing users to individual location pages. Must expose crawlable HTML links. |
|
LocalBusiness schema |
Schema.org type for a physical business. Google documents only two required properties: address and name. |
|
GeoCoordinates |
Recommended schema property carrying latitude and longitude. Google specifies a minimum of five decimal places of precision. |
|
openingHoursSpecification |
Recommended schema property expressing hours as day plus open and close times, with optional validity dates for seasonal hours. |
|
areaServed |
CHANGED · Valid schema.org property for expressing a service territory, but not listed in Google's LocalBusiness rich-result documentation. Legitimate for entity clarity; widely over-promised as a rich-result driver. |
|
Review velocity |
The rate at which new reviews arrive. Case study evidence associates sustained velocity with ranking stability; evidential rather than Google-confirmed. |
|
Review recency |
How recently reviews were posted. A steady cadence appears to outperform sporadic bursts, measured relative to local competitors. |
|
Review responses |
Owner replies to reviews. Google states responding demonstrates that you value customers and that positive reviews with helpful replies help a business stand out. |
|
Rating manipulation |
Google's policy category covering fake, incentivised, or conflicted reviews — including reviews from undisclosed employment, contractual, family, or competitor relationships. |
|
Apple Business Connect |
Apple's free listing platform (launched January 2023, superseding Apple Maps Connect) managing the place card shown in Apple Maps, Siri, Spotlight, Wallet, and Messages. Adds Showcases and Brand Pages. |
|
Bing Places for Business |
CHANGED · Microsoft's free local listing platform. Relaunched October 2025 and migrated from bingplaces.com to bing.com/forbusiness, with improved Google-listing import and bulk editing. |
|
GBP Performance report |
The current name for what was "GBP Insights." Documented metrics include Interactions, Searches, Views, Directions, Calls, Website clicks, Bookings, Products, and Offers. |
|
Direction requests |
Count of users requesting driving directions — the closest available proxy for physical visit intent. |
|
Discovery vs Direct vs Branded searches |
CHANGED · Deprecated segmentation. This breakdown is absent from Google's current Performance documentation, replaced by a raw Searches term list. Still heavily cited in blog posts recycling pre-2022 material. |
|
GBP Chat and Call history |
CHANGED · Removed 31 July 2024. Messaging and call-history tracking are no longer available in Business Profile. |
|
Local Search Ranking Factors |
CHANGED · The industry's largest annual expert survey on local ranking. Now published by Whitespark, not Moz. The 2026 edition surveyed 47 experts across 187 factors. |
|
Local link building |
Earning links from geographically relevant sources — chambers of commerce, local news, sponsorships, community organisations. Contributes to both link signals and unstructured citations. |
Section 7
AIO: AI OVERVIEWS & GOOGLE'S AI SEARCH SURFACES
Google's own generative surfaces. This is the area where 2026 brought the most consequential changes to the practitioner's toolkit.
CORE TERM
AI Overview (AIO)
Definition. Google's AI-generated summary appearing at the top of some results pages, with links out to supporting sources. Google states they appear "only when our systems determine that it is additive to classic Search." Launched broadly in the US in May 2024 as the successor to SGE.
Why it matters. They materially change what a ranking is worth. Pew Research found roughly 1% of users clicked a source link directly from an AI Overview, and SparkToro's mid-2026 study measured 68% of Google searches ending without any click — up from 60% in 2024.
How to execute. Change what you optimise for by intent class. On informational queries where Overviews saturate, optimise for citation — a direct 50 to 70 word answer under the H1, tabular comparisons, ordered process lists, sourced statistics. On commercial and transactional queries where Overview presence is far lower, keep optimising for the click.
Common mistake. Reading falling clicks with flat or rising impressions as a ranking loss. That pattern means the rankings survived and the clicks were absorbed by the answer layer — a completely different problem with a completely different fix.
CORE TERM
AI Mode
Definition. Google's dedicated conversational search surface built on Gemini, designed for queries requiring further exploration, reasoning, or complex comparison, and supporting multi-turn follow-ups.
Why it matters. Google reported one billion monthly active users at the one-year mark in May 2026, with average queries roughly three times longer than traditional search, and follow-up query volume growing over 40% month on month in the US.
How to execute. Longer, first-person, natural-language queries are a different keyword universe from the two-word head terms most B2B keyword research is built on. Mine sales call transcripts and support tickets for the full-sentence questions buyers actually ask, and build content that answers them in their phrasing.
Common mistake. Assuming classical keyword tools capture this demand. They do not — no platform publishes AI query volume, and every "prompt volume" figure in a vendor tool is modelled, not measured.
CORE TERM
Search Console generative AI reporting
Definition. Search Console reports launched on 3 June 2026 showing impressions from AI Overviews and AI Mode, segmentable by page, country, device, and date.
Why it matters. It is the first first-party visibility any site has had into Google's AI surfaces. Google defines an impression here as an instance in which a link to a site is shown within a generative AI feature.
How to execute. Use it to size exposure by template and by market. Note its hard limitation: it reports no queries and no clicks, so AI-surface click-through rate cannot be calculated from it. Google also confirmed that a URL appearing in both an AI Overview and the blue links counts as a single impression.
Common mistake. Confusing the three distinct AI controls Google now offers. `Google-Extended` governs Gemini training and does not affect AI Overviews. `nosnippet` governs snippet display. The Search Console generative AI control, rolled out June–July 2026, is the property-level setting that actually excludes content from AI Overviews and AI Mode without affecting normal Search ranking. Secondary sources conflate these constantly.
Standard terms — AIO
|
Term |
Definition |
|
SGE (Search Generative Experience) |
The Search Labs experiment launched in May 2023 that placed generative summaries into results. Renamed and productised as AI Overviews in May 2024; the acronym is now historical only. |
|
Query fan-out |
The technique where a single query is decomposed into multiple related sub-queries issued in parallel across subtopics and data sources, with results synthesised into one answer. Google confirms both AI Overviews and AI Mode may use it; the internals are undocumented, so third-party "fan-out simulators" are inference. |
|
Query decomposition |
Breaking a complex question into smaller answerable components — a sub-step of fan-out. |
|
Query variant generation |
Producing multiple rephrasings or related queries from one input, then issuing each separately and combining results. |
|
Grounding |
Connecting a model's output to verifiable external sources so responses are anchored in retrievable data rather than parametric memory. Google's stated purpose is reducing hallucination and providing auditability. |
|
Retrieval-Augmented Generation (RAG) |
An architecture retrieving relevant documents at query time and supplying them to a model as context before generation. In a search context, RAG is the mechanism that makes a page eligible to be cited. |
|
AI Overview citation |
A link to a web page displayed within or alongside an AI Overview or AI Mode answer as a supporting source. |
|
Zero-click search |
A search ending without a click to any external site, resolved on the results page itself. SparkToro measured 68.01% of US Google searches in early 2026. |
|
Answer substitution |
The effect of an AI answer satisfying the query so the user does not visit any source. The mechanism behind falling clicks on stable rankings. |
|
Multi-turn search |
A session where the user refines through successive conversational turns with context retained across them. |
|
Multimodal search |
Queries combining or using voice, image, or video. Google reported more than one in six AI Mode searches are multimodal. |
|
Conversational query patterns |
The observed shift toward longer, natural-language, first-person queries. Google reports common AI Mode opening words are "what," "how," "I," "is," and "can." |
Section 8
AEO: ANSWER ENGINE OPTIMIZATION
Optimising to be the answer rather than to be a link near it. Note that the boundary between AEO, GEO, and SEO is genuinely contested — see the note at the end of Section 9.
CORE TERM
Answer Engine Optimization (AEO)
Definition. Optimising content to be selected as the direct answer in systems that answer rather than list — featured snippets, knowledge panels, voice assistants, and AI answer surfaces.
Why it matters. The unit of success changes from a ranked position to an extracted passage. You are competing against any page containing a better-extractable answer, including pages that would never outrank you classically.
How to execute. Structure for extraction: the direct answer in the first 40 to 80 words of a section, self-contained passages that make sense lifted out of the page, comparisons as tables, processes as ordered lists, and definitions stated plainly rather than embedded mid-paragraph.
Common mistake. Treating AEO as distinct enough from SEO to warrant a separate strategy. Google's own stated position is that "optimizing for generative AI search is optimizing for the search experience, and thus still SEO." The tactical overlap is very large; what genuinely differs is measurement.
CORE TERM
Answer-first content structure
Definition. Structuring a page so the direct answer to its core question appears immediately, before context, narrative, or brand framing.
Why it matters. It is the highest-leverage, lowest-cost content intervention available on most B2B sites, because it improves featured snippet capture, AI citation eligibility, and reader experience simultaneously — using content that already ranks.
How to execute. Lead each section with a self-contained declarative sentence that would make sense if lifted out entirely. Avoid opening with "In today's rapidly evolving landscape." Avoid pronouns needing the previous paragraph to resolve. Test it: pull any random 60-word passage and ask whether it stands alone as a correct, attributable statement.
Common mistake. Presenting this as a documented ranking factor. It is a well-founded heuristic inferred from how passages are extracted — no search engine documents it as a selection signal.
Standard terms — AEO
|
Term |
Definition |
|
Answer engine |
A system returning a synthesised direct answer rather than a ranked list. The term predates LLMs; no authoritative body defines its current boundaries. |
|
Featured snippet |
A block at the top of Google's organic results extracting an answer from a ranking page, with attribution and a link. Standard types: paragraph, list, table, and video. |
|
Position zero |
Informal name for the featured snippet slot. Increasingly a misnomer, since AI Overviews now occupy the position above it. |
|
People Also Ask (PAA) |
A SERP feature showing expandable related questions with short extracted answers. Each answer is generally a featured snippet for that sub-question. |
|
Passage-level extraction |
The mechanism by which search and AI systems retrieve the specific segment of a page that best answers a query, rather than evaluating the page as a whole. |
|
Content chunking |
Structuring content into small, self-contained sections that can be retrieved independently. Note: the retrieval-side chunking done by AI systems is real and technical; authoring-side "chunk optimisation" is an inferred practice, and publishers do not control how their content is chunked. |
|
Extractability |
How readily a passage can be lifted from a page and still be correct, complete, and attributable. The practical objective of AEO work. |
|
Self-contained passage |
A block of text that does not depend on surrounding context to be understood. |
|
FAQ block |
CHANGED · A question-and-answer section on a page. Note that FAQ rich results were retired in 2026 — the block retains user and extraction value, but produces no Google SERP enhancement. |
|
Voice search optimisation |
Optimising for spoken queries, which skew longer, more conversational, and more question-formed. Largely subsumed into general AEO practice. |
|
Definition block |
A short, plainly worded definition placed near the top of a page. The single most reliably extracted content structure. |
|
Comparison table |
A structured tabular comparison. Strongly favoured by both snippet extraction and AI retrieval for commercial-investigation queries. |
Section 9
GEO & LLMO: GENERATIVE ENGINE & LLM OPTIMIZATION
Optimising to be retrieved, quoted, and cited inside generated responses from systems such as ChatGPT, Perplexity, Gemini, and Claude — and the crawler-access decisions that make it possible.
CORE TERM
Generative Engine Optimization (GEO)
Definition. The practice of improving content visibility in generative engine responses. The term originates in a peer-reviewed paper — GEO: Generative Engine Optimization (Aggarwal et al., arXiv November 2023, published at KDD 2024) — which defined it as a black-box optimisation framework and reported visibility gains of up to 40%.
Why it matters. Retrieval-based answers now sit between a large share of buyers and any website. For a B2B services firm, being absent from the answer to "best generative AI development companies" is a commercial problem no amount of classical ranking fixes.
How to execute. The paper tested nine content modifications. The best performer by a wide margin was Quotation Addition (+41% on its position-adjusted word count metric), followed by citing sources and adding statistics. The worst performers were keyword stuffing and unique words — the tactics that most resemble classical manipulation.
Common mistake. Citing those results as though they describe live Google or ChatGPT behaviour today. They come from 2023–24 model generations tested on a synthetic benchmark. The directional finding — that citations, quotations, and statistics increase inclusion — is robust and consistent with later practitioner evidence; the specific percentages are not portable.
CORE TERM
GEO is largely a third-party game
Definition. The observation that AI answers to category-level questions draw predominantly from sources other than vendor websites — review platforms, directories, publisher listicles, and community threads.
Why it matters. This inverts the classical model. In SEO you compete by improving your own pages; in GEO, a substantial share of brand references originate on third-party sites you do not control. If you are absent from the sources a system retrieves, your own content quality is irrelevant to that answer.
How to execute. Map which sources are actually cited across a set of category prompts, then work the ones that recur. Typically fewer than ten sources account for the majority of citations — usually two or three review platforms, several publisher roundups, and a community forum. Maintain complete, current profiles on the review platforms; pitch inclusion in the roundups with a differentiated angle; participate genuinely, as a named expert, in the communities.
Common mistake. Building an on-site GEO programme and measuring no movement on category prompts, because the answer was never going to come from your domain.
CORE TERM
AI crawler taxonomy: training vs retrieval vs user-triggered
Definition. The now-standard three-way distinction. Training crawlers collect corpora for model pretraining. Retrieval or search crawlers build an index used to answer queries and cite sources. User-triggered fetchers load a single URL in real time because a user asked something right now.
Why it matters. This is the distinction that makes robots.txt decisions rational. Blocking the training bot while allowing the search bot is the standard "stay visible, don't feed the model" posture — and it is only possible because the major operators now document their agents separately.
How to execute. For most B2B companies whose content is marketing rather than a product, allow retrieval and user-triggered agents; make a separate, deliberate decision about training crawlers. Blocking retrieval agents removes you from live answers where a citation is genuinely possible.
Common mistake. Blocking all AI user agents on principle, then discovering a year later that competitors are cited in the majority of category answers and you appear in none. Note also that Perplexity documents that its user-triggered agent generally ignores robots.txt, on the reasoning that a human requested the specific fetch — a contested stance, and a reason robots.txt alone is insufficient for access control.
CORE TERM
Measuring AI visibility
Definition. Tracking brand presence in AI answers by repeatedly submitting a fixed prompt set across platforms and recording mentions, citations, competitor presence, and sentiment.
Why it matters. No rank tracker covers this comprehensively and no platform publishes query data, so a self-maintained prompt set is the only observable measurement available.
How to execute. Define 60 to 150 prompts representing real buyer questions across the funnel. Run them on a fixed schedule across the platforms that matter. Record: were you mentioned, were you cited with a link, which URL, what position in the answer, who else was cited, and what the answer said about you. That yields four trackable metrics — mention rate, citation rate, share of citations against named competitors, and sentiment.
Common mistake. Comparing scores between vendors. Every AI-visibility metric — citation rate, share of voice, visibility score, prompt volume — is sampled from a vendor-chosen prompt set against a vendor-chosen competitor list. They are directionally useful within one tool over time, and not comparable across tools.
CORE TERM
llms.txt
Definition. A proposed convention: a Markdown file at the domain root giving AI systems a curated map of a site's most important content. Proposed by Jeremy Howard in September 2024; version 2 published in August 2026.
Why it matters. Mostly as a case study in separating specification from adoption. It has a real spec and real uptake among developer-tool companies — and no major AI search platform has publicly committed to consuming it as a retrieval or ranking input.
How to execute. Google's position is explicit: "You don't need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search," and such files "neither harm nor help your site's visibility or rankings in Google Search." Treat it as a cheap hedge with no downside, keep it genuinely curated rather than a URL dump, and do not attribute outcomes to it.
Common mistake. Presenting llms.txt as an AI visibility strategy. The things that demonstrably move citations — extractable content, credible sourcing, entity clarity, third-party presence — are where the effort belongs.
Standard terms — GEO and LLMO
|
Term |
Definition |
|
Generative engine |
A system using generative models to gather and summarise information into a synthesised response rather than a ranked list. |
|
LLM Optimization (LLMO) / LLM SEO |
Optimising a brand's presence in large language model outputs, covering both retrieval-time citation and representation in training data. |
|
AI SEO |
A loose umbrella term used variously for optimising for AI search and for using AI in SEO workflows. Ambiguous — worth defining explicitly whenever used. |
|
GEO-BENCH |
The benchmark introduced alongside the GEO paper: a large-scale set of diverse user queries across domains with relevant web sources, used to evaluate optimisation strategies systematically. |
|
Position-Adjusted Word Count |
One of two visibility metrics defined in the GEO paper: the normalised word count of response sentences attributable to a citation, with exponential decay so earlier citations weigh more. |
|
Subjective Impression |
The GEO paper's second metric — an LLM-scored composite across relevance, influence, uniqueness, position, perceived volume, click likelihood, and diversity. |
|
Prompt set |
The fixed list of queries a visibility tool submits repeatedly to AI platforms to sample brand presence. Every AI-visibility metric is defined relative to this set. |
|
Citation rate |
CHANGED · How often a brand or domain is cited as a source in AI answers across a tracked prompt set. Vendor definitions differ materially — some count linked citations only, others any brand mention. |
|
Brand mention rate |
The percentage of tracked prompts whose AI responses contain the brand name, linked or not. |
|
Share of Voice (in AI) |
The percentage of AI-answer mentions accruing to your brand versus a defined competitor set, over a defined prompt set. Not an absolute market measure. |
|
Share of Model (SoM) |
CHANGED · How often and how favourably a brand appears in LLM outputs relative to competitors. Coined and trademarked by the agency Jellyfish — a proprietary commercial framing rather than a neutral industry metric. |
|
Citation share |
The proportion of all citations in a topic or prompt set pointing to your domain versus competitors'. |
|
AI visibility score |
CHANGED · A composite 0–100 measure of brand presence across AI answers. Proprietary and computed differently by every vendor; never compare across tools. |
|
Source opportunities / missing sources |
Third-party sites that AI answers cite when mentioning competitors but not your brand. Operationally the most actionable AI-visibility output, since it converts directly to a digital PR target list. |
|
AI referral traffic |
CHANGED · Sessions arriving from an AI assistant, identified by referrer and usually isolated with a GA4 custom channel group. Treat all figures as floors — platforms withhold referral data, so measured AI traffic is systematically understated. |
|
Crawl-to-refer ratio |
Cloudflare Radar's metric for the value exchange between AI platforms and publishers: how often a platform sends traffic to a site relative to how often it crawls it. Reported ratios have run into the tens of thousands to one. |
|
Prompt volume / AI topic volume |
CHANGED · An estimate of how frequently users ask about a topic across AI platforms. No AI platform publishes query volume — all such figures are modelled from panels or extrapolation and are not equivalent to keyword search volume. |
|
Sentiment in AI answers |
The tone with which a brand is described inside AI responses, typically scored positive, neutral, or negative and decomposed into drivers. |
|
GPTBot |
OpenAI's training crawler, collecting content that may be used to train foundation models. Blocking it does not affect ChatGPT search visibility. |
|
OAI-SearchBot |
OpenAI's search-indexing crawler, used to surface sites in ChatGPT's search features. Blocking it removes you from ChatGPT search answers. |
|
ChatGPT-User |
OpenAI's user-triggered fetcher, used when a user's question requires visiting a page. Not used for automatic crawling and does not inform search indexing. |
|
ClaudeBot |
Anthropic's training crawler, collecting public web content that may be used to train and improve its models. |
|
Claude-SearchBot |
Anthropic's retrieval crawler, crawling content to improve the quality and relevance of Claude's search results. |
|
Claude-User |
Anthropic's user-triggered fetcher, retrieving pages when a user asks a question requiring a specific URL. Anthropic documented these three agents separately in February 2026, enabling granular robots.txt decisions. |
|
PerplexityBot |
Perplexity's search-index crawler, used to surface and link sites in Perplexity results. Not used for foundation model training; respects robots.txt. |
|
Perplexity-User |
CHANGED · Perplexity's user-triggered fetcher. Perplexity documents that it generally ignores robots.txt on the basis that a human requested the fetch. |
|
CCBot / Common Crawl |
The crawler of the Common Crawl Foundation, whose open archives are a major upstream training-data source for many models. Blocking it is a broad indirect training opt-out affecting future crawls only. |
|
Google-Extended |
CHANGED · Not a crawler but a robots.txt token controlling whether Google-crawled content may train future Gemini models and ground Gemini apps. Does not control AI Overviews or AI Mode, and does not affect Search inclusion or ranking. |
|
Applebot-Extended |
Apple's training opt-out token. Apple states it does not crawl — it signals that content should not train Apple's foundation models, while pages disallowing it can still appear in Siri, Spotlight, and Safari results. |
|
Meta-ExternalAgent / Meta-WebIndexer / Meta-ExternalFetcher |
Meta's training crawler, retrieval crawler, and user-triggered fetcher respectively. |
|
Bytespider / Amazonbot |
CHANGED · ByteDance's and Amazon's crawlers. Neither operator publishes a training-versus-retrieval breakdown comparable to OpenAI's or Anthropic's, so the categorisations circulating in crawler directories are third-party inference. |
|
Content Signals Policy |
Cloudflare's robots.txt extension adding declarations for search, ai-input, and ai-train. Cloudflare acknowledges these are not legally binding, and Google has not committed to honouring them. |
|
Pay per crawl |
Cloudflare's mechanism letting site owners charge AI crawlers for access, returning HTTP 402 to unpaid requests. From 15 September 2026 Cloudflare blocks training and agent bots by default on new ad-monetised domains. |
|
RSL (Really Simple Licensing) |
An open standard letting publishers declare machine-readable licensing terms — attribution, pay per crawl, pay per inference — via robots.txt, headers, or feeds. Backed by Creative Commons, Cloudflare, and Akamai. |
|
Agentic browsing |
A browser or browser mode where an AI agent reads, summarises, and acts on pages on the user's behalf. Named examples include Perplexity Comet, OpenAI Atlas, and Dia. |
|
AI agents as visitors |
CHANGED · The treatment of autonomous agents as a distinct traffic class alongside humans and classic bots. A real phenomenon with no standard measurement — distinguishing agentic from human traffic is currently unsolved. |
|
MCP (Model Context Protocol) |
An open standard, originated by Anthropic, for connecting AI applications to external data, tools, and workflows. Relevant to search as a route by which agents access structured business data directly, bypassing HTML. |
|
Agentic commerce |
Transactions where AI agents complete purchases autonomously on a user's behalf, with LLM platforms acting as gatekeepers for both discovery and checkout. |
|
ACP (Agentic Commerce Protocol) |
An open standard from OpenAI and Stripe defining interactions between agents, merchants, and payment providers. |
|
UCP (Universal Commerce Protocol) |
Google's broader agentic-commerce framework spanning discovery through post-purchase across AI surfaces, incorporating MCP and Agent2Agent. |
A note on GEO, AEO, LLMO, and AIO
CHANGED · The industry has not settled these definitions, and any glossary claiming otherwise is overstating the case. Ahrefs' published position is that GEO, LLMO, and AEO are "three names for the same concept, not distinct disciplines." Google states that optimising for generative AI search is "still SEO." Wikipedia notes that consensus "remains uncertain."
Only GEO has a defensible origin — the KDD 2024 paper — and even there, the paper's meaning (black-box content optimisation measured on a synthetic benchmark) is narrower than the industry's current usage.
The practical resolution used in this glossary: treat them as different measurement frames over a largely shared tactical base. The tactics that serve all of them are the same — crawlable content, extractable structure, factual density, credible sourcing, entity clarity, third-party corroboration. What differs is what you count. Classical SEO counts positions and clicks; AEO counts extracted answers; GEO counts citations and share of model. Choosing a label matters far less than being explicit about which metric you are accountable for.
Section 10
SXO: SEARCH EXPERIENCE OPTIMIZATION
The discipline that treats the ranking as the beginning of the job rather than the end.
CORE TERM
Search Experience Optimization (SXO)
Definition. The integration of SEO with user experience and conversion design — optimising the complete journey from query through landing to conversion, rather than optimising visibility alone.
Why it matters. It closes the gap that produces the most common failure mode in B2B SEO: a page that ranks well, receives traffic, and converts nobody. SEO owns the arrival; SXO owns whether the arrival was worth anything.
How to execute. Four components in practice: search fundamentals (the page must be found), experience (navigation, accessibility, speed, readability, layout stability), engagement (does the content hold the reader and answer what they came for), and conversion (is there a relevant, low-friction next step relevant to their stage).
Common mistake. Measuring SXO work with SEO metrics. Rankings and sessions will not move; engagement rate, scroll depth, key event rate, and qualified lead rate will. Instrument for the second set before starting.
CORE TERM
Intent satisfaction
Definition. Whether the page actually resolves what the searcher came for — the underlying quality SXO optimises and that engagement metrics only proxy.
Why it matters. Dwell time, bounce rate, and time on page are all ambiguous. A short visit can mean the page failed, or that it answered the question perfectly in ten seconds. Intent satisfaction is what you actually care about and cannot measure directly.
How to execute. Triangulate. Combine engagement signals with on-page surveys ("did this answer your question?"), search-within-site behaviour, subsequent page paths, and — most usefully in B2B — what prospects say in sales calls about what they read. Return-to-SERP behaviour is the strongest negative signal available, if you can approximate it.
Common mistake. Treating time on page as a quality metric. It is a diagnostic input whose meaning depends entirely on the intent of the query.
Standard terms — SXO
|
Term |
Definition |
|
Conversion Rate Optimization (CRO) |
The discipline of increasing the share of visitors who complete a desired action. SXO is broadly the intersection of CRO and SEO, applied specifically to search-arriving traffic. |
|
Engagement rate |
GA4's metric: engaged sessions divided by total sessions. The successor to bounce rate, which in GA4 is simply its inverse. |
|
Engaged session |
A session lasting longer than ten seconds, or containing a key event, or containing two or more page views. |
|
Dwell time |
The interval between a user clicking a result and returning to the SERP. Not directly measurable in your own analytics and not a documented ranking factor. |
|
Pogo-sticking |
A user clicking a result, returning quickly to the SERP, and choosing a different result. Widely interpreted as a dissatisfaction signal; never confirmed by Google. |
|
Return to SERP |
The general pattern of a user going back to search after a visit. The behavioural shape behind pogo-sticking. |
|
Scroll depth |
How far down a page users travel. Useful for diagnosing whether long-form content is actually being read or abandoned at the introduction. |
|
Above-the-fold experience |
What a visitor sees before scrolling. On B2B service pages, whether the value proposition and a next step are visible without scrolling is a measurable conversion lever. |
|
Friction |
Anything increasing the effort required to complete the desired action — form field count, mandatory phone numbers, unclear CTAs, forced account creation. |
|
Form conversion rate |
The share of form views resulting in a submission. Usually the highest-leverage single metric on a B2B service page. |
|
Micro-conversion |
A smaller committed action short of the primary goal — a resource download, a pricing page view, a video watch. Useful as a leading indicator on long B2B sales cycles. |
|
Accessibility (a11y) |
Designing so people with disabilities can use the site. Overlaps heavily with SXO and SEO — semantic HTML, descriptive alt text, keyboard navigation, and colour contrast serve both. |
|
Core Web Vitals (as UX) |
The same three metrics covered in Section 4, framed as experience rather than ranking. The commercial case for fixing them is usually stronger than the ranking case. |
|
Interstitials |
Overlays appearing before or over content. Intrusive interstitials on mobile are subject to demotion and are a substantial conversion drag. |
|
Message match |
The consistency between the query, the title link, and the page's opening content. A mismatch is the most common cause of immediate abandonment. |
|
Trust signals |
On-page elements evidencing credibility — named case studies, certifications, client logos, review scores, named team members, compliance statements. |
Section 11
METRICS, KPIs & MEASUREMENT
The vocabulary of proving SEO worked — and of avoiding the reporting errors that make it look like it did when it didn't.
CORE TERM
Brand vs non-brand segmentation
Definition. Separating queries containing your brand name from those that do not, and reporting them independently.
Why it matters. Brand traffic is a function of your marketing spend, PR, and existing customer base — a demand-capture metric, not an SEO metric. Mixing it into organic reporting means a funding announcement or an ad campaign will inflate your SEO numbers, and a quiet quarter will make good work look like failure.
How to execute. Build a regex covering the brand name plus every misspelling, spacing variant, and product-name combination appearing in Search Console. Apply it as a filter and track non-brand separately. Review the regex quarterly, since new misspellings appear as awareness grows.
Common mistake. Reporting total organic growth to leadership without segmenting. It is the single most common way an SEO programme conceals a real decline behind a healthy headline.
CORE TERM
Segmentation by page type, intent, and position band
Definition. Reporting performance separately by template (service, industry, product, blog, case study), by search intent, and by ranking band (1–3, 4–10, 11–20, 21–30, 31+).
Why it matters. Aggregate numbers hide divergence, and divergence is where the insight is. Flat total traffic can conceal blog traffic falling 34% while service-page conversions rise 78% — a headline that would nearly justify a budget cut is actually the best commercial year the channel has had.
How to execute. Tag every URL with its template and cluster at the outset. The position-band cut is the most actionable: keywords in 11–20 have already demonstrated relevance and need a push, not a rebuild, and band migration is a leading indicator that shows strategy is working months before traffic reflects it.
Common mistake. Reporting "average position" across an entire keyword set. On a large site it mixes a position-2 head term with a position-90 irrelevant one and is close to meaningless. Report band distribution instead.
CORE TERM
Connecting SEO to pipeline
Definition. Joining organic landing pages to leads, leads to opportunities, and opportunities to revenue, so SEO can be reported in business terms rather than session counts.
Why it matters. For a B2B services company, lead count is a weak proxy and traffic is no proxy at all. The metric that survives contact with a CFO is organic-sourced and organic-influenced pipeline value.
How to execute. Three joins. Capture first-touch and last-touch organic landing pages as hidden fields on every form. Tag opportunities with the originating content cluster. Report at cluster level, not page level — a B2B buyer reads four or five pages over eleven weeks, so "this blog post generated $0" is technically true and strategically useless.
Common mistake. Relying on GA4's last-click attribution alone. It credits the branded search or direct visit at the end of the journey and zeroes out the content that started it.
CORE TERM
Reading Search Console correctly
Definition. Interpreting Google's first-party performance data with its limitations understood.
Why it matters. Search Console is the most reliable SEO data source available and the most frequently misread. Its metrics have precise definitions that differ from intuition.
How to execute. Understand four things. Average position is averaged across impressions and is not a rank tracker equivalent. Query data is anonymised and thresholded, so query-level totals never sum to property totals and long-tail performance is systematically understated. The Page indexing report is the current name — "Coverage" is legacy. And for unsampled analysis, use the bulk data export to BigQuery, which bypasses the UI's row limits and is the only way to do genuine long-tail or cannibalisation analysis at scale.
Common mistake. Diagnosing a click decline without checking impressions. Falling clicks with flat or rising impressions is SERP-layout erosion, not ranking loss — and in 2026 that is usually an AI Overview.
Standard terms — Search Console
|
Term |
Definition |
|
Impressions |
How frequently a site appears in results for a query, page, or date. Counted when a link is seen or would be seen. |
|
Clicks |
Instances where a user selects your site from search results. |
|
CTR |
Clicks divided by impressions. |
|
Average position |
The average position of the topmost result from your site, averaged across impressions. Heavily distorted by long-tail volume. |
|
Query dimension |
Performance grouped by search term. Subject to anonymisation and privacy filtering. |
|
Page dimension |
Performance grouped by the canonical URL receiving the impression or click. |
|
Page indexing report |
The report showing which URLs are indexed and why others are not. Current name; "Index Coverage" is legacy. |
|
URL Inspection |
The tool showing Google's indexed version, canonical selection, coverage status, and rendered HTML for a single URL. |
|
Generative AI performance report |
Reporting launched June 2026 for AI Overviews, AI Mode, and Discover AI features. Provides impressions, pages, countries, devices, and dates — but no queries and no clicks. |
|
Bulk data export |
Daily automated export of Search Console performance data to BigQuery, unsampled and unconstrained by UI row limits. |
|
Enhancements reports |
Search Console reporting on structured data validity by type. A template bug produces hundreds of identical errors overnight, making this a useful regression alarm. |
Standard terms — GA4 and analytics
|
Term |
Definition |
|
Key events |
CHANGED · GA4's term for events measuring actions important to the business. Renamed from "conversions" in March 2024 — "conversion" now refers specifically to the Google Ads object created from a GA4 key event. |
|
Sessions |
A group of user interactions within a timeframe, started by a session_start event. GA4 sessions do not break on campaign change, unlike Universal Analytics. |
|
Engagement rate |
Engaged sessions divided by total sessions. GA4's replacement for bounce rate. |
|
Key event rate |
Sessions or users with a given key event divided by the total. GA4's replacement for conversion rate. |
|
Attribution models |
CHANGED · GA4 currently offers only three: data-driven, paid and organic last click, and Google paid channels last click. First click, linear, time decay, and position-based were removed in November 2023 — first-touch analysis now requires BigQuery or a CRM. |
|
Default channel group |
GA4's rule-based classification of traffic sources. |
|
Organic Search channel |
CHANGED · GA4's rule: source matches a recognised search site list, or medium exactly matches "organic." Note that AI assistant referrals fall outside this list and land in Referral unless a custom channel group is configured. |
|
Custom channel group |
A user-defined channel classification. The standard mechanism for isolating AI assistant traffic in GA4. |
|
Assisted conversion |
CHANGED · A conversion where a channel appeared in the path but was not the final touch. Not a standalone GA4 metric as it was in Universal Analytics; it must be reconstructed from path or attribution reporting. |
|
First-touch vs last-touch |
First-touch credits the initial interaction, favouring discovery channels like content; last-touch credits the final one, favouring capture channels like brand search. |
|
MQL / SQL |
Marketing Qualified Lead and Sales Qualified Lead — the standard handoff points for attributing organic search to pipeline. Definitions are organisation-specific. |
|
Sales acceptance rate |
The share of marketing-passed leads that sales accepts as worth pursuing. The single most useful quality filter in SEO reporting. |
|
Conversions per 1,000 sessions |
A normalised conversion metric enabling fair comparison between clusters of very different traffic volumes. |
Standard terms — ranking, visibility, and third-party metrics
|
Term |
Definition |
|
SERP position |
The ordinal rank of a result for a given query, location, and device. Increasingly ambiguous as AI Overviews and packs occupy the page. |
|
Position band |
A bucketed grouping of rankings used to report distribution shifts rather than noisy individual movements. |
|
Striking distance |
Keywords ranking just below meaningful traffic, conventionally positions 11–20, where modest improvement produces disproportionate click gains. |
|
Visibility score |
A vendor index combining a keyword set's rankings weighted by volume and estimated CTR. Comparable over time within one tool; never across tools. |
|
Share of voice (SOV) |
Your share of total available organic clicks or impressions across a defined keyword set, relative to competitors. |
|
Estimated traffic value |
The modelled cost of buying, via paid search, the traffic a page earns organically. A rough commercial proxy, not revenue. |
|
Keyword difficulty (KD) |
A vendor 0–100 estimate of ranking difficulty, in most tools derived chiefly from the backlink profiles of current top-ranking pages. Not cross-comparable between vendors. |
|
CPC as commercial proxy |
Using a keyword's paid cost-per-click as a signal of commercial value, on the reasoning that advertisers bid up queries that convert. Unreliable where paid competition is thin. |
|
Search volume |
The estimated monthly search count for a term. An input to prioritisation, never a ranking — misleading when the SERP is saturated with features, when intent does not match the business, or when the volume aggregates many distinct intents. |
|
Expected qualified clicks |
A more defensible prioritisation metric than volume: volume × realistic CTR at the achievable position given SERP layout × an intent-quality factor from your own historical data. |
|
Crawl stats report |
Search Console reporting on Googlebot's crawl requests over time, by response, file type, and purpose. The lightweight alternative to full log analysis. |
Section 12
TOOLS
The standard practitioner stack, with what each is actually for.
|
Tool |
What it is and what it's for |
|
Google Search Console |
Google's free first-party property tool: Performance, Page indexing, URL Inspection, sitemaps, Core Web Vitals, structured data, manual actions, and the 2026 generative AI reports. The only source of Google's own view of your site — always the first place to verify a third-party claim. |
|
Google Analytics 4 (GA4) |
Event-based analytics measuring sessions, engaged sessions, engagement rate, key events, attribution, and channel grouping, with free BigQuery export. |
|
BigQuery |
Google Cloud data warehouse. The destination for Search Console bulk export and GA4 raw event export, enabling unsampled long-tail, cannibalisation, and cohort analysis beyond UI row limits. |
|
Looker Studio |
Free dashboarding tool with native Search Console, GA4, and BigQuery connectors. The standard layer for blended SEO reporting. |
|
Screaming Frog SEO Spider |
Desktop crawler for technical audits — titles, meta, headings, status codes, redirects, canonicals, hreflang, directives — with JavaScript rendering and Search Console, GA4, and PageSpeed API integrations. The default technical audit tool. |
|
Screaming Frog Log File Analyser |
Parses server access logs to reveal actual Googlebot behaviour. The only tool class that records crawling rather than simulating it. |
|
Sitebulb |
Desktop and cloud crawler emphasising prioritised, explained audit hints and crawl visualisations over raw data output. |
|
Ahrefs |
Backlink-index-first platform: Site Explorer, Keywords Explorer, Content Gap, Site Audit, Rank Tracker, plus DR/UR metrics and Brand Radar for AI-surface visibility. |
|
Semrush |
All-in-one SEO, PPC, and competitive platform: Domain Overview, Keyword Magic, Position Tracking, Site Audit, Authority Score, and an AI visibility toolkit. |
|
Moz |
The platform behind Domain Authority, Page Authority, Spam Score, Link Explorer, and the Moz SEO Learning Center. |
|
PageSpeed Insights |
Presents CrUX field data alongside a Lighthouse lab audit for the same URL — the standard way to see both data types side by side. |
|
Lighthouse |
Open-source lab auditing engine in Chrome DevTools, CLI, and CI, scoring performance, accessibility, best practices, and SEO. Lab scores are not Core Web Vitals assessment. |
|
CrUX (Chrome UX Report) |
The public dataset of real Chrome user experience, exposed via API, BigQuery, PageSpeed Insights, and Search Console. The field data behind Core Web Vitals. |
|
Rich Results Test |
Tests whether a URL is eligible for a Google rich result, using live rendering. |
|
Schema Markup Validator |
schema.org's validator, checking markup against the full vocabulary regardless of Google support. |
|
Rank trackers |
Tools sampling SERPs on a schedule by keyword, location, and device. A necessary complement to Search Console's average position, which averages across impressions. |
|
Local rank grid tools |
Tools sampling local rankings across a geographic grid rather than a single point, making proximity decay visible. Essential for any multi-location or service-area programme. |
|
AI visibility trackers |
CHANGED · An emerging category prompting AI surfaces on a fixed prompt set and recording mentions, citations, and share of voice. Every vendor computes these differently on a different prompt set — never compare scores across tools. |
|
Google Trends |
Relative search interest over time and by region. Useful for seasonality and for validating whether a term's volume is trend-driven. |
|
Google Business Profile Manager |
The interface for managing local listings, categories, posts, services, reviews, and the Performance report. |
Section 13
DEPRECATED TERMS & MYTHS
A working glossary is as useful for what it removes as for what it defines. Every item below appears routinely in current published SEO material and is either dead, never existed, or is asserted far beyond the evidence.
Deprecated — no longer exists
|
Term |
Status |
|
FID (First Input Delay) |
Replaced by INP as a Core Web Vital on 12 March 2024 and fully retired from Google's tooling in September 2024. |
|
FAQ rich results |
Stopped appearing 7 May 2026; reporting and Rich Results Test support dropped June 2026. The schema type remains valid; the SERP feature does not. |
|
HowTo rich results |
Removed September 2023 and the documentation deleted. |
|
Sitelinks search box |
Dropped 21 November 2024, along with the nositelinkssearchbox directive. |
|
Breadcrumb rich results on mobile |
Desktop-only since January 2025. |
|
Course info, estimated salary, learning video, special announcement, vehicle listing rich results |
Removed September 2025. |
|
Helpful Content System / HCU |
CHANGED · Retired as a standalone system. Google's ranking systems guide lists it under retired systems: it became part of core ranking in March 2024. "HCU recoveries" are core update recoveries. |
|
rel="next" / rel="prev" |
No longer used by Google, though other engines may still read them. |
|
Dynamic rendering |
Formally labelled a deprecated workaround by Google in early 2024. |
|
noarchive / nocache |
No longer used by Google Search. |
|
URL Parameters tool (Search Console) |
Retired in 2022. |
|
Search Console crawl rate limiter |
Retired January 2024. Signal overload with 429, 500, 503, or 504 instead. |
|
GA4 first-click, linear, time-decay, position-based attribution |
Removed November 2023. |
|
GA4 "conversions" |
Renamed to "key events" in March 2024. |
|
Search Console "Coverage" report |
Renamed "Page indexing." |
|
GBP Q&A |
Discontinued — announced September 2025, API cut November 2025, removed from profiles by early 2026. |
|
GBP Chat and Call history |
Removed 31 July 2024. |
|
GBP Discovery / Direct / Branded search breakdown |
Absent from current Performance documentation. |
|
HARO / Connectively |
Shut down 9 December 2024. No single official successor. |
|
Neustar Localeze |
Renamed TransUnion Digital Business Profile. |
|
Data Axle Express Update |
Defunct. |
|
Toolbar PageRank |
Removed 2016. |
|
SGE (Search Generative Experience) |
Renamed AI Overviews in May 2024. Historical use only. |
|
Moz-hosted Local Search Ranking Factors |
Now published by Whitespark. |
Myths — never were what they are claimed to be
|
Claim |
Reality |
|
LSI keywords |
CHANGED · Do not exist in the context of Google. Latent Semantic Indexing is genuine 1980s technology built for small static corpora, not the web. Google's John Mueller: "There's no such thing as LSI keywords." Use "semantically related terms." |
|
Keyword density |
Not a ranking factor, repeatedly confirmed. Optimising toward a density percentage trends directly into keyword stuffing. |
|
Content length as a ranking factor |
Not a ranking factor. Correlation studies confuse depth with word count. Length should follow what the query requires. |
|
Content-to-code ratio |
Never was an SEO factor, per Google. Retained only as a rough bloat diagnostic in some audit tools. |
|
Image title attribute |
Carries no documented ranking or accessibility value. Alt text is the meaningful attribute. |
|
rel="dofollow" |
Not a real attribute and never has been in any specification. An ordinary link is followed by default. |
|
nofollow as an absolute block |
CHANGED · Since March 2020, nofollow, sponsored, and ugc are hints, not directives. Google may still crawl and count them. |
|
"Toxic links" |
A vendor concept with no Google counterpart. Toxicity scores are heuristics, and acting on them routinely disavows links that were doing no harm. |
|
Routine disavow hygiene |
Contrary to Google's guidance, which states most sites will never need the tool and that incorrect use can harm performance. |
|
E-E-A-T as a ranking factor |
CHANGED · Google states plainly that E-E-A-T itself is not a specific ranking factor. There is no E-E-A-T score. |
|
Quality rater scores affect rankings |
They do not feed rankings directly. Google's own analogy is feedback cards at a restaurant. |
|
PageRank sculpting |
Broken since 2009 — nofollowed internal links evaporate equity rather than redirecting it. |
|
Link velocity as a positive signal |
Not a confirmed ranking factor. Velocity patterns feed spam detection, not quality assessment. |
|
Centroid targeting |
Industry folklore. Never documented by Google; evidence favours searcher-device proximity. |
|
"All paid links are prohibited" |
CHANGED · Google states buying and selling links is a normal part of the economy. The violation is failing to qualify them with sponsored or nofollow. |
|
areaServed produces a rich result |
Valid schema.org, but absent from Google's LocalBusiness rich-result documentation. |
|
Service schema produces a rich result |
It does not. Service appears nowhere in Google's search gallery. |
|
Self-serving review markup earns stars |
CHANGED · Explicitly ineligible. A business marking up reviews of itself — including via embedded third-party widgets — cannot earn star review features. |
|
llms.txt improves AI search visibility |
CHANGED · No major platform has committed to consuming it. Google states such files neither harm nor help visibility in Google Search. |
|
"Information gain score" |
The patent is real; its use in organic ranking is unconfirmed, and the patent is framed around automated assistants. A sound content heuristic, not a documented signal. |
|
Google-Extended controls AI Overviews |
CHANGED · It does not. It governs Gemini model training. AI Overview exclusion is a separate Search Console control introduced June–July 2026. |
|
Prompt volume equals search volume |
No AI platform publishes query volume. All prompt-volume figures are modelled from panels or extrapolation. |
|
AI visibility scores are comparable across tools |
They are not. Each is sampled from a vendor-chosen prompt set against a vendor-chosen competitor list. |
|
"Parasite SEO" is a viable tactic |
Largely the same behaviour Google now prosecutes as site reputation abuse. High risk in 2026. |
Section 14
REFERENCES & BIBLIOGRAPHY
Every definition in this glossary was checked against a published source in August 2026. Sources are grouped by type, with primary platform documentation first — where Google, OpenAI, Anthropic, Apple, Meta, or Cloudflare document their own behaviour, that documentation outranks any secondary interpretation of it.
Where this glossary flags a term as contested, deprecated, or mythical, the flag is traceable to a source below.
A. Google — Search Central documentation
1. Google Search Central — SEO Starter Guide — https://developers.google.com/search/docs/fundamentals/seo-starter-guide
2. Google Search Central — Google Search's Guide to Ranking Systems (includes retired systems) — https://developers.google.com/search/docs/appearance/ranking-systems-guide
3. Google Search Central — Spam Policies for Google Web Search — https://developers.google.com/search/docs/essentials/spam-policies
4. Google Search Central — Creating Helpful, Reliable, People-First Content (E-E-A-T) — https://developers.google.com/search/docs/fundamentals/creating-helpful-content
5. Google Search Central — Control Your Title Links in Search Results — https://developers.google.com/search/docs/appearance/title-link
6. Google Search Central — Control Your Snippets in Search Results — https://developers.google.com/search/docs/appearance/snippet
7. Google Search Central — Keep a Simple URL Structure — https://developers.google.com/search/docs/crawling-indexing/url-structure
8. Google Search Central — Google Images SEO Best Practices — https://developers.google.com/search/docs/appearance/google-images
9. Google Search Central — Understanding Page Experience in Google Search Results — https://developers.google.com/search/docs/appearance/page-experience
10. Google Search Central — Introduction to robots.txt — https://developers.google.com/search/docs/crawling-indexing/robots/intro
11. Google Search Central — Robots Meta Tags and X-Robots-Tag Specifications — https://developers.google.com/search/docs/crawling-indexing/robots-meta-tag
12. Google Search Central — Block Search Indexing with noindex — https://developers.google.com/search/docs/crawling-indexing/block-indexing
13. Google Search Central — How to Specify a Canonical URL with rel="canonical" — https://developers.google.com/search/docs/crawling-indexing/consolidate-duplicate-urls
14. Google Search Central — Build and Submit a Sitemap — https://developers.google.com/search/docs/crawling-indexing/sitemaps/build-sitemap
15. Google Search Central — How HTTP Status Codes, Network and DNS Errors Affect Google Search — https://developers.google.com/search/docs/crawling-indexing/http-network-errors
16. Google Search Central — Understand JavaScript SEO Basics — https://developers.google.com/search/docs/crawling-indexing/javascript/javascript-seo-basics
17. Google Search Central — Dynamic Rendering as a Workaround — https://developers.google.com/search/docs/crawling-indexing/javascript/dynamic-rendering
18. Google Search Central — Qualify Your Outbound Links to Google — https://developers.google.com/search/docs/crawling-indexing/qualify-outbound-links
19. Google Search Central — Localized Versions of Your Pages (hreflang) — https://developers.google.com/search/docs/specialty/international/localized-versions
20. Google Search Central — Pagination, Incremental Page Loading, and Search — https://developers.google.com/search/docs/specialty/ecommerce/pagination-and-incremental-page-loading
21. Google Search Central — AI Features and Your Website — https://developers.google.com/search/docs/appearance/ai-features
22. Google Search Central — Google's Guide to Optimizing for Generative AI Features — https://developers.google.com/search/docs/fundamentals/ai-optimization-guide
23. Google Search Central — Latest Google Search Documentation Updates — https://developers.google.com/search/updates
B. Google — crawling infrastructure (relocated late 2025)
24. Google — Crawl Budget Management for Large Sites — https://developers.google.com/crawling/docs/crawl-budget
25. Google — Myths About Crawling — https://developers.google.com/crawling/docs/myths-about-crawling
26. Google — Managing Crawling of Faceted Navigation URLs — https://developers.google.com/crawling/docs/faceted-navigation
27. Google — Google Crawlers and Fetchers Overview (incl. Google-Extended) — https://developers.google.com/crawling/docs/crawlers-fetchers/overview-google-crawlers
28. Google — Verifying Googlebot and Other Google Crawlers — https://developers.google.com/crawling/docs/crawlers-fetchers/verify-google-requests
29. Google — Reduce Googlebot Crawl Rate — https://developers.google.com/crawling/docs/crawlers-fetchers/reduce-crawl-rate
C. Google — structured data
30. Google Search Central — Structured Data Markup that Google Search Supports (Search Gallery) — https://developers.google.com/search/docs/appearance/structured-data/search-gallery
31. Google Search Central — Intro to Structured Data Markup — https://developers.google.com/search/docs/appearance/structured-data/intro-structured-data
32. Google Search Central — Organization Structured Data — https://developers.google.com/search/docs/appearance/structured-data/organization
33. Google Search Central — Local Business (LocalBusiness) Structured Data — https://developers.google.com/search/docs/appearance/structured-data/local-business
34. Google Search Central — Review Snippet Structured Data (self-serving review policy) — https://developers.google.com/search/docs/appearance/structured-data/review-snippet
35. Google Search Central — Breadcrumb Structured Data — https://developers.google.com/search/docs/appearance/structured-data/breadcrumb
36. Google Search Central — FAQPage Structured Data (deprecation notice) — https://developers.google.com/search/docs/appearance/structured-data/faqpage
37. Google Search Central — Profile Page Structured Data — https://developers.google.com/search/docs/appearance/structured-data/profile-page
38. Google — Rich Results Test — https://search.google.com/test/rich-results
39. schema.org — Vocabulary — https://schema.org/
40. schema.org — Schema Markup Validator — https://validator.schema.org/
D. Google Search Central blog — dated announcements
41. Google — Introducing INP to Core Web Vitals (May 2023) — https://developers.google.com/search/blog/2023/05/introducing-inp
42. Google — Bulk Data Export: A New Way to Access Your Search Console Data (Feb 2023) — https://developers.google.com/search/blog/2023/02/bulk-data-export
43. Google — Farewell, Sitelinks Search Box (Oct 2024) — https://developers.google.com/search/blog/2024/10/sitelinks-search-box
44. Google — Crawling December: Faceted Navigation (Dec 2024) — https://developers.google.com/search/blog/2024/12/crawling-december-faceted-nav
45. Google — E-E-A-T Gets an Extra E for Experience (Dec 2022) — https://developers.google.com/search/blog/2022/12/google-raters-guidelines-e-e-a-t
46. Google — Introducing a New Spam Policy for "Back Button Hijacking" (Apr 2026) — https://developers.google.com/search/blog/2026/04/back-button-hijacking
47. Google — Introducing Search Generative AI Performance Reports in Search Console (Jun 2026) — https://developers.google.com/search/blog/2026/06/gen-ai-performance-reports
E. Google — Search Console, Analytics & Business Profile help
48. Search Console Help — Search Performance Report — https://support.google.com/webmasters/answer/7576553
49. Search Console Help — Page Indexing Report — https://support.google.com/webmasters/answer/7440203
50. Search Console Help — Bulk Data Export Reference — https://support.google.com/webmasters/answer/12917991
51. Search Console Help — Disavow Links to Your Site — https://support.google.com/webmasters/answer/2648487
52. Google Analytics Help — Key Events — https://support.google.com/analytics/answer/9267568
53. Google Analytics Help — Attribution and Attribution Modeling — https://support.google.com/analytics/answer/10596866
54. Google Analytics Help — Default Channel Group Definitions — https://support.google.com/analytics/answer/9756891
55. Google Analytics Help — Engaged Sessions and Engagement Rate — https://support.google.com/analytics/answer/12195621
56. Google Analytics Help — Dimensions and Metrics Reference — https://support.google.com/analytics/answer/9143382
57. Google Business Profile Help — Tips to Improve Your Local Ranking on Google — https://support.google.com/business/answer/7091
58. Google Business Profile Help — Guidelines for Representing Your Business on Google — https://support.google.com/business/answer/3038177
59. Google Business Profile Help — Add, Edit, or Remove Categories — https://support.google.com/business/answer/7249669
60. Google Business Profile Help — Create, Edit & Manage Posts — https://support.google.com/business/answer/7662907
61. Google Business Profile Help — Manage Your Business Attributes — https://support.google.com/business/answer/9049526
62. Google Business Profile Help — Understand Your Business Profile Performance — https://support.google.com/business/answer/9918094
63. Google Business Profile Help — Changes to Business Profile Chat and Call History — https://support.google.com/business/answer/14919056
64. Google Business Profile Community — Upcoming Changes to the Q&A Feature — https://support.google.com/business/thread/392024106/a-heads-up-about-the-upcoming-changes-to-the-q-a-feature
65. Google Maps — Prohibited & Restricted Content Policy (Rating Manipulation) — https://support.google.com/contributionpolicy/answer/7400114
F. Web platform & performance
66. web.dev (Google) — Web Vitals — https://web.dev/articles/vitals
67. web.dev (Google) — Optimize Resource Loading with the Fetch Priority API — https://web.dev/articles/fetch-priority
68. web.dev (Google) — Why HTTPS Matters — https://web.dev/articles/why-https-matters
69. web.dev (Google) — Fixing Mixed Content — https://web.dev/articles/fixing-mixed-content
70. Chrome for Developers — Chrome UX Report (CrUX) — https://developer.chrome.com/docs/crux
71. Chrome for Developers — Lighthouse Documentation — https://developer.chrome.com/docs/lighthouse
72. Google — PageSpeed Insights — https://pagespeed.web.dev/
73. MDN Web Docs — Strict-Transport-Security — https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Strict-Transport-Security
74. Open Graph Protocol — The Open Graph Protocol — https://ogp.me/
G. AI platform documentation — crawlers and agents
75. OpenAI — Bots and Crawlers (GPTBot, OAI-SearchBot, ChatGPT-User, OAI-AdsBot) — https://developers.openai.com/api/docs/bots
76. Perplexity — PerplexityBot and Perplexity-User Documentation — https://docs.perplexity.ai/guides/bots
77. Apple Support — About Applebot and Applebot-Extended — https://support.apple.com/en-us/119829
78. Meta for Developers — Meta Web Crawlers — https://developers.facebook.com/docs/sharing/webmasters/web-crawlers/
79. Common Crawl — CCBot — https://commoncrawl.org/ccbot
80. Google Cloud — Grounding Overview, Vertex AI — https://docs.cloud.google.com/vertex-ai/generative-ai/docs/grounding/overview
81. Model Context Protocol — Introduction — https://modelcontextprotocol.io/docs/getting-started/intro
82. llmstxt.org — The /llms.txt File Specification (v2, Aug 2026) — https://llmstxt.org/
83. RSL Standard — Really Simple Licensing 1.0 — https://rslstandard.org/
H. Infrastructure and crawler policy
84. Cloudflare Blog — Your Site, Your Rules: New AI Traffic Options (Search / Agent / Training, pay per crawl) — https://blog.cloudflare.com/content-independence-day-ai-options/
85. Cloudflare Blog — The Crawl Before the Fall of Referrals (crawl-to-refer ratio) — https://blog.cloudflare.com/ai-search-crawl-refer-ratio-on-radar/
86. Cloudflare Blog — Control Content Use for AI Training with Managed robots.txt — https://blog.cloudflare.com/control-content-use-for-ai-training/
87. Microsoft Bing — IndexNow Protocol — https://www.bing.com/indexnow
88. Bing Blogs — Introducing the New Bing Places for Business (Oct 2025) — https://blogs.bing.com/search/October-2025/Introducing-the-New-Bing-Places-for-Business-Built-for-Business-Owners,-Powered-by-Research
89. Apple Newsroom — Introducing Apple Business Connect — https://www.apple.com/newsroom/2023/01/introducing-apple-business-connect/
I. Academic and independent research
90. Aggarwal, Murahari, Rajpurohit, Kalyan, Narasimhan & Deshpande — GEO: Generative Engine Optimization (arXiv 2311.09735 / KDD 2024) — https://arxiv.org/abs/2311.09735
91. Same — full PDF with metrics and method results — https://arxiv.org/pdf/2311.09735
92. Princeton University — GEO: Generative Engine Optimization (publication record) — https://collaborate.princeton.edu/en/publications/geo-generative-engine-optimization/
93. Pew Research Center — Google Users Are Less Likely to Click on Links When an AI Summary Appears (Jul 2025) — https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/
94. SparkToro (Rand Fishkin) — In 2026, Less than One Third of Google Searches Still Send a Click — https://sparktoro.com/blog/in-2026-less-than-one-third-of-google-searches-still-send-a-click/
95. Whitespark — Local Search Ranking Factors 2026 — https://whitespark.ca/local-search-ranking-factors/
96. Sterling Sky — Does Review Recency Impact Ranking? (Case Study) — https://www.sterlingsky.ca/google-review-recency-ranking/
J. Search Engine Land
97. What Is Local SEO? — https://searchengineland.com/guide/what-is-local-seo
98. What Is the Google Local Pack? — https://searchengineland.com/guide/google-local-pack
99. The Proximity Paradox: Beating Local SEO's Distance Bias — https://searchengineland.com/guide/google-proximity-bias-in-local-search
100. "Near Me" SEO — https://searchengineland.com/guide/near-me-search-optimization
101. Structured vs Unstructured Citations for Local SEO — https://searchengineland.com/guide/structured-vs-unstructured-citations
102. Location Page SEO — https://searchengineland.com/guide/location-pages-seo
103. Service Area Pages — https://searchengineland.com/guide/service-area-pages
104. Multi-Location SEO — https://searchengineland.com/guide/multi-location-seo
105. What Is Link Equity? — https://searchengineland.com/guide/link-equity
106. What Is Link Velocity? — https://searchengineland.com/guide/link-velocity
107. Unlinked Mentions — https://searchengineland.com/guide/unlinked-mentions
108. What Is Share of Voice? — https://searchengineland.com/guide/share-of-voice
109. Entity-First Content Optimization — https://searchengineland.com/guide/entity-first-content-optimization
110. Links and Brand Signals: The SEO Authority Model — https://searchengineland.com/links-brand-signals-seo-authority-model-475968
111. Query Fan-Out in AI Search: What It Is and How It Works — https://searchengineland.com/guide/query-fan-out
112. Anthropic Clarifies How Claude Bots Crawl Sites (Feb 2026) — https://searchengineland.com/anthropic-claude-bots-470171
113. Cloudflare Offers Way to Block AI Overviews — Will Google Comply? — https://searchengineland.com/cloudflare-content-signals-462538
114. Google Zero-Click Searches Reach 68% in Early 2026: Study — https://searchengineland.com/google-zero-click-searches-2026-study-479717
115. Google's AI Overviews Are Hurting Clicks: Pew Study — https://searchengineland.com/google-ai-overviews-hurting-clicks-study-459434
116. Google Updates Search Quality Raters Guidelines (Sep 2025) — https://searchengineland.com/google-updates-search-quality-raters-guidelines-adding-ai-overview-examples-ymyl-definitions-461908
117. Google to No Longer Support FAQ Rich Results — https://searchengineland.com/google-to-no-longer-support-faq-rich-results-476957
118. What Is Edge SEO? — https://searchengineland.com/edge-seo-447510
119. Information Gain in SEO: What It Is and Why It Matters — https://searchengineland.com/what-is-information-gain-seo-why-it-matters-429763
120. Table of Contents Can Provide Additional Link Options in Organic Search — https://searchengineland.com/pro-tip-table-of-contents-can-provide-additional-link-options-in-organic-search-336615
K. Search Engine Journal
121. Google Now Reports AI Search Impressions — How to Read Them (Jun 2026) — https://www.searchenginejournal.com/google-reports-ai-search-impressions-how-to-read-them/582824/
122. What Opting Out of Google's AI Search Features Means Now — https://www.searchenginejournal.com/what-opting-out-of-googles-ai-search-features-means-now/584321/
123. Google Reveals First AI Mode Usage Numbers After One Year (May 2026) — https://www.searchenginejournal.com/google-shares-first-ai-mode-usage-data-after-one-year/575443/
124. Agentic Commerce: What SEOs Need to Consider (ACP & UCP) — https://www.searchenginejournal.com/agentic-commerce-what-seos-need-to-consider-acp-ucp/563503/
125. A Complete List of Google's Featured Snippet Types — https://www.searchenginejournal.com/featured-snippets-types/219907/
126. Google's Information Gain Patent for Ranking Web Pages — https://www.searchenginejournal.com/googles-information-gain-patent-for-ranking-web-pages/524464/
127. New Google Spam Policy Targets Back Button Hijacking — https://www.searchenginejournal.com/new-google-spam-policy-targets-back-button-hijacking/571859/
128. Google Drops FAQ Rich Results From Search — https://www.searchenginejournal.com/google-drops-faq-rich-results-from-search/574429/
129. Google Phases Out Support for Noarchive Meta Tag — https://www.searchenginejournal.com/google-phases-out-support-for-noarchive-meta-tag/529021/
130. Link Building Terms You Should Know: The Ultimate Glossary — https://www.searchenginejournal.com/link-building-guide/link-building-terminology/
131. Co-Citation & Co-Occurrence: How Important Are They for SEO Today? — https://www.searchenginejournal.com/co-citation-co-occurrence-how-important-are-they-for-seo-today/370620/
L. Ahrefs
132. SEO Glossary — https://ahrefs.com/seo/glossary
133. GEO, LLMO, AEO… It's All Just SEO — https://ahrefs.com/blog/geo-is-just-seo/
134. How to Track and Analyze Your AI Traffic — https://ahrefs.com/blog/track-analyze-ai-traffic/
135. LSI Keywords: What Are They and Do They Matter? — https://ahrefs.com/blog/lsi-keywords/
136. What Is Search Intent? A Complete Guide — https://ahrefs.com/blog/search-intent/
137. What Is Content Decay? — https://ahrefs.com/blog/content-decay/
138. Internal Links for SEO: An Actionable Guide — https://ahrefs.com/blog/internal-links-for-seo/
139. What Is Anchor Text? — https://ahrefs.com/blog/anchor-text/
140. Link Building for SEO: The Beginner's Guide — https://ahrefs.com/seo/link-building
141. Content Gap Analysis — https://ahrefs.com/blog/content-gap-analysis/
142. On-Page SEO: The Beginner's Guide — https://ahrefs.com/blog/on-page-seo/
143. Glossary entries — Domain Rating, PageRank, People Also Ask, Orphan Page, Index Bloat, SEO Silo, Pillar Page, Internal Link, Co-Citation — https://ahrefs.com/seo/glossary
144. Ahrefs Help — How to Set Up Custom Prompts to Track Brand Visibility in AI Assistants — https://help.ahrefs.com/articles/13192745-how-to-set-up-custom-prompts-to-track-brand-visibility-in-ai-assistants
M. Semrush, Moz, Yoast & other vendors
145. Semrush Knowledge Base — AI Visibility Metrics — https://www.semrush.com/kb/1594-ai-seo-metrics
146. Semrush Knowledge Base — Authority Score and Backlink Scores — https://www.semrush.com/kb/747-authority-score-backlink-scores
147. Semrush Blog — How to Measure and Report on AI Search Visibility — https://www.semrush.com/blog/measure-ai-visibility/
148. Semrush Blog — Content Chunking — https://www.semrush.com/blog/content-chunking
149. Semrush Blog — Topic Clusters: What They Are and How to Build Them — https://www.semrush.com/blog/topic-clusters/
150. Semrush Blog — Content Pruning: A Step-by-Step Guide — https://www.semrush.com/blog/content-pruning/
151. Semrush Blog — Content Brief — https://www.semrush.com/blog/content-brief/
152. Semrush Blog — What Is Digital PR? — https://www.semrush.com/blog/digital-pr/
153. Semrush Blog — Link Building Strategies — https://www.semrush.com/blog/link-building-strategies/
154. Semrush Blog — Link Building Metrics — https://www.semrush.com/blog/link-building-metrics/
155. Moz — Domain Authority — https://moz.com/learn/seo/domain-authority
156. Yoast — What Is Search Experience Optimization (SXO)? — https://yoast.com/what-is-search-experience-optimization-sxo/
157. Yoast — Does Readability Rank? — https://yoast.com/does-readability-rank/
158. Yoast — Keyword Cannibalization: How to Identify and Fix It — https://yoast.com/keyword-cannibalization/
159. Screaming Frog — SEO Spider — https://www.screamingfrog.co.uk/seo-spider/
160. Screaming Frog — SEO Log File Analyser — https://www.screamingfrog.co.uk/log-file-analyser/
161. Sitebulb — https://sitebulb.com/
162. Clearscope — What Are Striking Distance Keywords? — https://www.clearscope.io/blog/what-are-striking-distance-keywords
163. BrightLocal — What Are Local Citations? — https://www.brightlocal.com/learn/local-citations/what-are-local-citations/
164. BrightLocal — Using Data Aggregators for Local Citations — https://www.brightlocal.com/learn/local-citations/data-aggregators/
165. Writer — GEO, AEO, and SEO in 2026: The Enterprise Guide to AI Visibility — https://writer.com/blog/geo-aeo-optimization/
N. Trade press and secondary reporting
166. Search Engine Roundtable — Google Says Keyword Density Still Not an SEO Ranking Factor — https://www.seroundtable.com/google-keyword-density-seo-32683.html
167. Search Engine Roundtable — Google Says Code to Text Ratio Has Never Been an SEO Factor — https://www.seroundtable.com/google-code-to-text-ratio-seo-factor-32889.html
168. PPC Land — Google Finally Gives Search Console Its Own Generative AI Visibility Reports — https://ppc.land/google-finally-gives-search-console-its-own-generative-ai-visibility-reports/
169. PPC Land — Google Tightens Maps Review Policy: Staff Names and Quotas Now Banned — https://ppc.land/google-tightens-maps-review-policy-staff-names-and-quotas-now-banned/
170. Marketing Week — "Share of Model" Is the New Marketing Measure for the AI Era — https://www.marketingweek.com/tom-roach-share-of-model-ai-era/
171. Digiday — No Playbook, Just Pressure: Publishers Eye the Rise of Agentic Browsers — https://digiday.com/media/no-playbook-just-pressure-publishers-eye-the-rise-of-agentic-browsers/
172. Octiv Digital — Connectively (Formerly HARO) to Shut Down on December 9, 2024 — https://www.octivdigital.com/ideas-and-advice/connectively-formerly-haro-to-shut-down-on-december-9-2024/
173. Wikipedia — Generative Engine Optimization — https://en.wikipedia.org/wiki/Generative_engine_optimization
174. Wikipedia — AI Overviews — https://en.wikipedia.org/wiki/AI_Overviews
How to keep this glossary current
Roughly a fifth of the entries in this document did not exist, or meant something different, two years ago — and Section 13 exists because a comparable proportion of what practitioners still repeat has quietly expired. Three habits keep a glossary like this from rotting:
Check the primary source before the summary. Where a platform documents its own behaviour, that documentation is authoritative and everything else is interpretation. A large share of persistent SEO myths survive because secondary coverage outlives the thing it described.
Watch the four sources that actually break news. Google Search Central's blog and its documentation-updates page announce changes first. Search Engine Land and Search Engine Journal cover them within a day. Everything else is downstream.
Re-audit the AI sections twice a year, and the classical sections annually. Sections 7 through 9 carry the most volatility: crawler taxonomies, Search Console AI reporting, and platform access policy all changed materially during 2026 alone. Sections 1 through 6 move far more slowly — but structured data has been the exception, losing more rich result types since 2023 than in the preceding five years combined.
No comments:
Post a Comment