SEO Glossary
SEO Glossary: Terms & Acronyms Explained
AIO, GEO, LLMO, E-E-A-T - explained simply, with hands-on context
SEO is full of acronyms and buzzwords, and every provider explains them a little differently. Here you get the most important terms explained briefly - with the point most definitions leave out: what really counts for rankings in 2026, and where the hype begins.
No term found. Try a different search word — or write to me which term you’re missing.
Topic
AI Search: AIO, GEO, AEO & LLMO
AI Overviews (AIO)
Google’s AI-generated answer boxes above the organic results. They assemble their answer from several sources (query fan-out) - being cited brings visibility, but depending on search intent and presentation it can lead to fewer clicks than classic organic results.
AI Mode
Google’s conversational search mode with its own interface. The difference to AI Overviews: AI Mode is a separate search experience, while AIO is just a box within the classic SERP.
GEO / AEO
Generative and Answer Engine Optimization respectively - optimizing for AI and answer engines. In practice, GEO/AEO largely matches classic SEO but often gets a new label: clean structure, entities, citability.
LLMO
Large Language Model Optimization - the discipline of showing up in the answers of ChatGPT, Gemini & co. In my view LLMO is not a replacement for SEO, but complements it as an additional layer of visibility.
Grounding
Anchoring AI answers in verifiable sources to reduce hallucinations. For SEO, grounding is an important mechanism that, by current observation, is often a prerequisite for becoming citable as a source.
Citation (AI citation)
Your page being named as a source inside an AI answer. Unlike a ranking it is not a place in a list but a yes-or-no decision per answer. So far it is measurable mainly in Bing Webmaster Tools, which report citations, grounding queries and citation share. Google’s GenAI reports have shown only impressions since June 2026, no citations.
Zero-Click Search
Searches that end without a click because the answer sits directly in the SERP. The consequence: think of impressions and clicks separately - visibility today does not automatically equal traffic.
llms.txt
A proposed Markdown file in the site root meant to give LLMs structured access to your core content. Adoption by the major AI providers is, by current observation, not yet widely confirmed as of 2026.
Preferred Sources
A feature that lets users set preferred sources for AI Overviews and AI Mode. For publishers, Preferred Sources can be a lever to show up more visibly in AI answers shown to their own regular readers.
Google AI Max
Google’s AI-powered campaign feature in Ads that automates keywords, creatives and placements. For organic SEO it mainly matters as context: where paid AI surfaces grow, organic space can tend to shrink - a common observation from practice.
Open Knowledge Format (OKF)
A format proposal for structuring knowledge in a machine-readable, AI-friendly way. As of 2026 an early concept - relevant as a signal of where structured content for LLMs might be heading.
SEO in AI browsers
AI browsers summarize page content and answer questions instead of just showing links. That shifts the click logic again - your content has to carry even when nobody opens it in full.
Topic
Google Algorithm & Updates
Google Core Update
A broad adjustment of the core ranking systems, several times a year. Not a penalty but a re-evaluation of relevance and quality - those who drop often have a substance problem, not a technical one.
Query Processing / Query Fan-Out
Google’s breaking down of a query into several sub-queries, especially in AI Mode. It explains why a page can be cited for an entire topic field - or not.
Crawling & Indexing
The two steps before any ranking: Google first has to find a page (crawling) and add it to the index. What is not indexed does not rank - no matter how good the text is.
Quality Rater
Human reviewers who assess search results against the Quality Rater Guidelines. They do not influence any single ranking directly, but train and calibrate the systems.
Ranking System
The interplay of many individual systems - not „the one algorithm“ - and increasingly ML-driven. This makes monocausal explanations like „factor X costs you the ranking“ mostly useless.
Semantic Search / Knowledge Graph
Google’s understanding of meaning and relationships between entities instead of plain keyword matches. The Knowledge Graph is the database behind it - those captured cleanly as an entity are better understood and cited.
Embeddings
The translation of text into numerical vectors whose closeness to each other measures similarity in meaning. The maths behind it is cosine similarity, a value near 1 means very similar. This is the mechanism behind semantic search and the reason concepts count rather than repeated keywords.
Google API Leak / User Signals
A content-warehouse API dataset allegedly leaked in 2024, holding thousands of internal ranking attributes including hints at click and user signals. NavBoost, which appears in it, had already been confirmed by Google in the US antitrust trial in 2023, together with its click memory covering the past 13 months, and before 2017 the window was 18. The individual click attributes from the schema, by contrast, remain officially undocumented. It is no proof of live ranking, but one of the deepest looks into Google’s possible signal world.
Google Discover
Google’s personalized content feed outside of classic search. Traffic without keyword intent - volatile, but a visibility channel of its own for visually strong, timely content.
Topic
Technical SEO
Googlebot
Google’s crawler, which fetches pages and prepares them for indexing. According to the documentation it stops fetching a supported file after the first 2 MB and passes on only the downloaded part for indexing. That affects almost nobody, the median sits at around 33 KB of HTML. You still get no warning in Search Console.
Crawl Budget
The resources Google spends on crawling a domain. Rarely a real problem for small sites - relevant only for large, dynamic or poorly structured sites.
Core Web Vitals
Google’s performance metrics LCP, INP and CLS. A tie-breaker signal: poor values hurt, perfect values alone won’t push you to the top.
JavaScript SEO / Rendering
The question of whether Google sees your JS-generated content. Rendering happens, but can be delayed or incomplete - server-side delivered content remains the safest route.
Canonical tag
The pointer to which URL is the authoritative one when several identical or similar versions exist. Google treats it as a signal, not as an instruction, and may decide otherwise. Position is what matters: if the tag sits in the body instead of the head, Google ignores it. That happens silently when a script in the head injects an iframe and closes the head early.
robots.txt
Controls which crawlers may access which paths (RFC 9309). Important in 2026: AI crawlers like GPTBot or Google-Extended can be addressed here specifically.
noindex
The instruction not to take a page into the index. It only works if the page may also be crawled, because a path blocked in robots.txt is never read. Do not confuse it with its neighbours: the AI Overviews switch, Google-Extended and nosnippet do three different jobs. Mixing them up blocks either too much or nothing at all.
hreflang
The markup that tells Google which language version is meant for which audience. Errors stay silent here: since the International Targeting report was retired in 2022, Google no longer reports them centrally in Search Console. Leaving out x-default gives away the safety net for every user none of your versions fits.
Breadcrumb Navigation
Path navigation that shows hierarchy - for users and for Google. With BreadcrumbList schema it also appears in the SERP snippet.
Structured Data / Schema
Machine-readable markup (JSON-LD) of your content. It brings no rankings directly, but helps Google and AI systems capture entities and context cleanly.
Rich Results
Enriched search results that Google builds from structured data. Google decides which types get shown and also withdraws them again: FAQ Rich Results were discontinued in May 2026. Deleting the schema for that reason would be premature, because it remains machine-readable context.
AI Crawlers / User-Agents for AI Agents
Bots like GPTBot, ClaudeBot, PerplexityBot or Google-Extended that fetch content for AI training and answers. Controllable via robots.txt - you decide who may read you for the AI era.
Topic
Content & Quality
E-E-A-T
Experience, Expertise, Authoritativeness, Trust - a framework from the Quality Rater Guidelines. Not a direct ranking factor, but the bundle of signals that explains why thin content doesn’t rank despite keywords.
Topical Authority
Authority for a topic, not for a domain. In a Semrush analysis across 1,094 categories the topic owner’s Authority Score came out ahead in only 52.5 percent of head-to-head comparisons, which is coin-flip territory. Whoever holds a topic usually keeps it: monthly stability was 90.4 percent.
Commodity Content / contentEffort
Interchangeable content that any AI delivers in seconds. The contentEffort hint surfaced in the 2024 Google API leak suggests that effort and substance could be measurable - so differentiation is advisable.
Featured Snippet
The highlighted answer box above the organic results, which Google pulls from an existing page. It brings visibility and often costs the click, because the answer already sits in the SERP. It is favoured by short, directly phrased answers high up in the text, such as a summary of 40 to 50 words right below the H1.
AI Texts & Google
Google judges not the method of production but the quality. By Google’s current statements, AI text is allowed - thin, redundant AI text is the problem, no matter who wrote it.
Content Pruning
The targeted removal or merging of dead, weak content. Fewer but stronger pages often beat a bloated archive.
Topic
Links & Authority
Link Juice / Link Equity
The ranking value passed on through links. It is split across all links on a page - which is why internal linking is not decoration but active steering of your authority.
Hub & Spoke
An architecture of a central pillar page (hub) and in-depth articles (spokes) that link to each other. It bundles topical authority and directs the internal link flow.
External Links
Outbound links to other domains. Set correctly, a quality and context signal - not a „giving away“ of authority, as the myth claims.
Link attributes: nofollow, sponsored, ugc
rel values that tell Google how to treat an outgoing link: nofollow as general distancing, sponsored for paid placements, ugc for user-generated content. Google reads them as a hint, not as a command. Paid links without such a marker breach the spam policies.
Backlinks / Link Building
Incoming links from other sites - still a strong signal, but quality beats quantity. For small domains, relevance counts more than volume.
Negative SEO / Spam Backlinks
The attempt to harm someone else’s site with toxic links. By Google’s own statement it usually ignores such links on its own - the Disavow tool is only rarely needed now.
Domain Authority
A third-party metric (Moz), not a Google factor. Useful as a rough comparison value, dangerous as a target KPI.
Topic
Spam, Black Hat & Manipulation
Parasite SEO
Exploiting the authority of strong third-party domains to rank your own - often low-quality - content. Google has been cracking down on this since 2024: a short-term lever with growing risk.
Black Hat SEO
Manipulation techniques against Google’s guidelines - from cloaking to link spam. Sometimes works briefly, costs visibility long-term; in the gambling space a permanent phenomenon.
CTR Manipulation
The artificial generation of clicks to fake better user behavior to Google. Whether and how strongly CTR ranks at all is contested - the manipulation stays risky and mostly detectable.
Site Reputation Abuse
Placing off-topic content on an authoritative domain to hijack its reputation (e.g. iGaming on news sites). Google has sanctioned this as a standalone violation since 2024.
Topic
Measurement & Tools
Google Search Console (GSC)
Google’s free data tool for your own domain: impressions, clicks, position, indexing status. The source that gives you Google’s own numbers instead of third-party estimates. RegEx filters in the performance report turn it from a table of numbers into a diagnostic instrument.
Impressions
How often a link to your site appeared in the search results. According to Google it counts as soon as the element is present on the current results page, whether or not anyone scrolls to it. In carousels and expandable widgets it only counts after scrolling or expanding. A drop while clicks stay stable therefore points to a measurement effect rather than lost rankings.
SERP volatility
How strongly rankings move from day to day, measured by sensors such as Semrush Sensor on a scale from 0 to 10. From the “High” range upwards an algorithm update becomes likely. The reading says nothing about whether your site gained or lost.
Want this not just explained, but actually implemented?
I help you translate these concepts into measurable rankings - data-driven and without the hype, tailored to your project, market and constraints. Prefer to dig in yourself? The whole system is in my book Das SEO-Kreativ Handbuch (in German) - 296 pages, 20 chapters, from the foundations to a 90-day plan.
The assessments in this glossary are based on my own experience and current observations, with no claim to completeness or official confirmation by Google. Brands mentioned (e.g. Google, Gemini, ChatGPT) belong to their respective owners; Google is a trademark of Google LLC.
