The SEO & AI-search glossary

An AI Overview is Google's generative answer that appears above the blue links for informational queries. Powered by Gemini, it summarises 3-8 cited sources into a synthesized answer, often with follow-up questions attached. AI Overviews now trigger on ~30% of US queries.

Generative Engine Optimization (GEO) is the practice of optimising content so it gets cited by generative AI engines — ChatGPT, Perplexity, Gemini, Claude, Google AI Overview, and Bing Copilot. Unlike classic SEO, the goal is citation and brand mention, not a blue-link click.

LLM Share of Voice is the percentage of relevant AI prompts across ChatGPT, Perplexity, Gemini, and Claude that cite or mention your brand versus your competitors. It's the AI-search equivalent of organic ranking share and is fast becoming the most predictive top-of-funnel metric for B2B SaaS.

A CTR curve is the expected click-through rate at each organic Google position — typically ~30% at #1, ~15% at #2, ~10% at #3, dropping to under 2% by position 10. Modern CTR curves have flattened due to AI Overviews and SERP features consuming clicks before the blue links.

SERP volatility is a measure of how much Google's rankings are shifting across a keyword set on a given day. High volatility = many pages moving up or down = likely algorithm update or index refresh. Low volatility = stable SERPs. It's the single best real-time indicator of Google-side changes.

Share of Voice in SEO is your domain's estimated percentage of total organic clicks or impressions across a fixed keyword set, compared to competitors ranking on the same terms. It's the single best metric for tracking category dominance because it accounts for both position AND search volume.

Anchor text is the visible, clickable text of a hyperlink. It tells Google what the linked page is about — making it one of the most powerful ranking signals — and simultaneously one of the most abused, which is why over-optimized anchors trigger penalties.

A backlink profile is the complete set of external links pointing to a website, plus their attributes — source domain authority, anchor text, follow vs nofollow, topical relevance, and geographic distribution. It's the single strongest off-page ranking signal after content quality.

Domain Authority (DA) is Moz's 0-100 score predicting how well a website will rank in search. Ahrefs uses Domain Rating (DR), Semrush uses Authority Score (AS) — all three estimate the strength of a site's backlink profile relative to the whole web. Higher = stronger authority = easier ranking.

Referring domains is the count of unique websites (root domains) that link to your site — different from total backlinks, because one site can link 100 times but only counts as one referring domain. It's the single strongest predictor of ranking in competitive SERPs.

Entity SEO is the practice of optimising for the concepts, people, places, products, and organisations Google's Knowledge Graph recognises — instead of just keywords. Google's modern ranking is entity-first: it understands "Apple" (company) is different from "apple" (fruit), and it ranks pages that clearly describe entities.

Keyword cannibalization happens when multiple pages on your site target the same query and compete against each other in Google's rankings — splitting authority, diluting CTR, and often causing all competing pages to rank lower than a single consolidated page would.

A topic cluster is a content model where one long-form "pillar" page targets a broad head term, surrounded by multiple "spoke" pages targeting related long-tail sub-topics, all interlinked. It signals topical authority to Google's NLP models and dominates a category more effectively than isolated blog posts.

Topical authority is Google's assessment that your site is a comprehensive, trustworthy source on a specific subject. It's built by covering a topic deeply (breadth × depth × freshness) rather than chasing unrelated high-volume keywords. Sites with strong topical authority rank new pages faster and higher in their niche.

Keyword difficulty is a 0-100 score estimating how hard it will be to rank in Google's top 10 for a given query. It blends the authority of the sites already ranking, backlink strength, on-page depth, and SERP-feature crowding into a single number.

The Keyword Golden Ratio (KGR) is a long-tail keyword formula: divide the number of Google results with the phrase in their title (allintitle) by the monthly search volume. If the ratio is under 0.25 and volume is below 250/month, the keyword is "golden" — usually rankable inside 30 days.

Long-tail keywords are specific, multi-word search queries — usually 4+ words — with lower individual search volume but sharper intent and less competition. Collectively they drive 70%+ of all Google searches and convert at 2-5x the rate of head terms.

Search intent is the goal behind a query — what the user actually wants to accomplish. The four canonical intents are informational (learn), navigational (find a specific site), commercial (research before buying), and transactional (buy or convert). Matching intent is now the #1 ranking factor after content quality.

Search volume is the estimated number of times a keyword is searched per month, usually reported per country. Most tools derive it from clickstream panels + Google Ads Keyword Planner, so the number is a smoothed estimate — treat it as a band, not a precise count.

Internal linking is the practice of linking from one page on your site to another. It distributes ranking authority across your domain, helps crawlers discover new content, defines topical relationships, and signals to Google which pages are most important — all of it fully under your control.

The meta description is the ~155-character snippet under your title tag in Google search results. It's not a direct ranking factor, but it heavily influences click-through rate — a compelling description at position 5 can out-earn a bland one at position 3.

The title tag is the HTML <title> element that becomes your page's clickable headline in Google search results and browser tabs. It's the single strongest on-page relevance signal and the biggest lever on CTR at your current ranking position.

A featured snippet is the answer box Google shows at the top of the SERP ("position zero"), extracted from a page already ranking in the top 10. Formats include paragraph, list, table, and video. Winning one can 2-5x your CTR for the query.

People Also Ask is Google's expandable question box on the SERP, showing 4 related questions with click-to-reveal answers. Each expanded question loads 2-4 more, so a single PAA box can surface 20+ questions. Winning a PAA slot drives clicks even from position 30+.

SERP features are the non-blue-link elements on a Google results page — AI Overviews, featured snippets, People Also Ask, image packs, video carousels, local packs, knowledge panels, shopping results, and site links. They now occupy 40-70% of a typical SERP and are ranked separately from organic blue links.

A Google algorithm update is a change to the ranking system that reshapes SERPs across large swathes of the web. Google runs thousands of changes per year; only a few are "core updates" — broad, named, multi-week rollouts that shift 5-30% of rankings across most niches.

SERP stands for Search Engine Results Page — the page Google (or Bing, DuckDuckGo, etc.) returns after a query. Modern SERPs mix classic blue links with AI Overviews, featured snippets, People Also Ask boxes, image packs, videos, and shopping carousels.

A canonical URL is the "master" version of a page you tell Google to index when the same or similar content is reachable from multiple URLs. Declared via <link rel="canonical" href="…" /> in the <head>, it consolidates ranking signals and prevents duplicate-content dilution.

Core Web Vitals are Google's three page-experience metrics: LCP (loading), INP (interactivity — replaced FID in 2024), and CLS (visual stability). They're a confirmed Google ranking factor, though a small one — a tiebreaker between otherwise-equal pages more than a primary lever.

Crawl budget is the number of URLs Googlebot will request on your site in a given period. It's a function of your site's host load capacity and Google's "crawl demand" (how important + fresh your content appears). Small sites rarely hit the limit; sites over ~10k URLs must manage it actively.

Indexation is the process of Google adding your page to its searchable index. A page must be crawled, deemed valuable, and pass duplicate-content checks before it's indexed. GSC's Coverage report shows every indexation state — indexed, discovered-not-indexed, crawled-not-indexed, and excluded.

A noindex meta tag or X-Robots-Tag HTTP header tells search engines to crawl a page but exclude it from their index. It's the correct way to remove pages from Google — unlike robots.txt (which blocks crawling but not indexation of already-known URLs).

robots.txt is a plain-text file at your domain root (yoursite.com/robots.txt) that tells crawlers which URLs they can and cannot request. It controls crawl budget, blocks staging paths, and points crawlers at your sitemap — but it does not prevent indexation.

Schema markup (structured data) is standardised JSON-LD you embed in your page's HTML to describe its content to search engines and LLMs — articles, FAQs, products, reviews, events, and more. It powers rich results in Google, better AI-Overview citations, and cleaner LLM answers.

An XML sitemap is a machine-readable list of the URLs on your site that you want search engines to crawl and index. Submitted via Google Search Console (or discovered through robots.txt), it accelerates discovery of new pages and communicates freshness with per-URL lastmod dates.