Say hello to the new Saazy! See what’s new ✨

Glossary

LLMTel Glossary

A/B Testing

Running controlled experiments (e.g., 10% see Variant B) to compare versions of a page or element. Treat Googlebot like any other user group; don’t special‑case it or you risk cloaking. Use canonicals if separate URLs exist.

AI‑Powered Channels

Places where AI systems surface or generate content (chatbots, AI Overviews, answer engines). Optimize both on‑page and off‑page signals so AIs can retrieve and cite your brand.

AI Overviews (Search)

Summaries generated by search engines that blend results and directly answer queries. Strong structured data, factual snippets, and authority increase inclusion.

Alt Text (Images)

Descriptive text within the alt attribute that explains an image’s content for accessibility and for machines that can’t “see” images. Keep concise and specific.

Answer Engine Optimization (AEO)

Optimizing content so that AI assistants and search “answer boxes” can confidently quote, cite, and include your brand in direct answers. See Model Tuning / LLMO.

Backlink Authority

The quality and quantity of external sites linking to you. High‑authority, relevant links strengthen signals used by AI‑augmented search and increase brand mentions.

Breadcrumbs (Structured Data)

Schema that helps search engines understand page position within a site’s hierarchy; can improve rich results.

Burger Menu (Hamburger Navigation)

Collapsed mobile navigation pattern. Acceptable for SEO when loaded in the DOM at page load; still ensure key paths are discoverable and UX‑friendly.

Build Topical Authority

Create pillar pages and deep “spoke” articles on core themes to signal expertise. Reinforce with FAQ schema and strong internal linking.

Canonical URL

Preferred URL for duplicate or near‑duplicate pages, declared via <link rel=”canonical”> to consolidate signals.

Common Crawl Presence

Verification that your site’s URLs appear in the Common Crawl index. Many AI systems use it to map the web; ensure robots and sitemaps enable crawling.

Content Freshness

Updating content doesn’t automatically improve rankings. Update when it benefits users or truthfulness; expose “Last Updated” where appropriate.

Crawlability

How easily bots can discover, fetch, and render pages. Large sites must manage internal links, sitemaps, robots rules, performance, JS rendering, and duplication.

Crawl Budget

The volume of URLs search engines will crawl on your site within a time period. Improve with clean architecture, fast performance, and fewer duplicates.

Directory Listings (NAP Consistency)

Claim and standardize business profiles (Google Business, Apple Maps, Bing Places, Yelp). Keep Name, Address, Phone identical to reduce confusion.

Domain History / Reputation

There’s no “reset button” for a domain with a poor past. You must clean it up over time or consider moving; quality and trust signals need to be rebuilt.

Duplicate Content

Substantially similar pages that split signals. Use canonicals, consolidate thin pages, and differentiate content.

E‑E‑A‑T (Experience, Expertise, Authoritativeness, Trustworthiness)

A framework used in quality rater guidelines and reflected in many ranking and retrieval signals; show real experience, credentials, and sources.

Entity (Knowledge Graph)

A machine‑understandable object like a brand, person, or product. Clear entity markup and corroborated facts improve how AIs connect your content to queries.

Featured Snippets / Direct Answers

Compact answers pulled into SERPs and AI Overviews. Use question‑style H2/H3 headings and concise (<50 words) definitions, plus lists/tables where helpful.

Fine‑Tuning (Model Training Step 2)

Model stage after pretraining that specializes skills and alignment before public release; contributes to the delay between data ingestion and what models “know.”

Generated Prompt Results (GPR)

The answers you get from an AI when you enter a prompt (analogous to search results for a query). Used as a lens for brand presence across AI systems.

GPR Score (Total)

Sum of your LLM Recognition Score and LLM Mention Score; indicates overall AI presence from static model memory plus live, citation‑driven mentions.

Hreflang / x‑default

Signals language‑/region‑specific versions. Useful for multi‑country same‑language sites to serve the right URL; not a ranking boost but aids correct URL swapping.

High‑Crawl Sites (Features On)

Placements on major, frequently crawled domains (news, industry leaders) that reinforce authority and increase discovery.

Image Indexing (SVG Considerations)

Inline SVGs may not be picked up for image search; consider separate SVG files when image traffic matters. Provide descriptive filenames and alt text.

Internal Linking

Logical, shallow paths and contextual links that surface priority pages, distribute PageRank/authority, and aid crawling.

JSON‑LD (Structured Data)

Preferred format for adding schema.org markup (Organization, Product, Article, FAQ, Breadcrumbs) to make facts machine‑readable.

Knowledge Cutoff

The most recent date of data included in a model’s training snapshot. New facts added to the web may not appear until a future model release.

Knowledge Graph

Graph of entities and their relationships used to interpret pages and queries; sources include Wikipedia/Wikidata among others.

Large Language Model Optimization (LLMO) / Model Tuning / GEO

Practices to influence inclusion in AI‑generated answers (beyond classic SEO). Focus on structured facts, authority, freshness, and corroboration so AIs cite you.

Last Updated (Timestamps)

Visible update dates on articles and key pages that signal recency to users and systems.

LLM Mention Score

The count of models that mention your brand in generated answers based on live, web‑augmented retrieval; can be improved quickly with stronger signals.

LLM Recognition Score

The count of models that recognize your brand in their base knowledge; largely static until retraining, but improvable long‑term via better web signals.

Mobile & Performance (PageSpeed)

Aim for ~90+ on Mobile and Desktop in PageSpeed Insights. Fix opportunities (compression, caching, minification). Re‑test regularly to maintain fast loads.

Model Release Lag

Practical delay between when the web is crawled and when models expose that knowledge: pretraining → fine‑tuning → public release.

Off‑Page Signals

External corroboration that builds authority and entity clarity: Wikipedia/Wikidata, high‑quality backlinks, directory listings, social mentions, features on major sites.

On‑Page Signals

Content and technical practices on your site: strong titles/meta, headings, alt text, structured data, clear “Quick Facts,” freshness, and performance.

Organization Schema

JSON‑LD block on your homepage with name, URL, logo, sameAs social links, contact points, address, and founding date to help AIs map your entity.

Pillar / Cluster Content Strategy

Create comprehensive pillar guides plus deep spokes to cover subtopics; add FAQ schema and track engagement (time on page, scroll depth).

Pretraining (Model Training Step 1)

Initial unsupervised training on massive corpora to learn language patterns; typically months long and defines the knowledge cutoff.

Press Releases & News

Regular, substantive updates (monthly cadence is fine) that feed discovery and corroborate facts for AI systems.

Quick Facts (Fact Cards)

Visible box with 3–5 core data points (e.g., founded, HQ, CEO). Mirrors machine‑readable schema and improves retrieval into snippets/overviews.

RAG (Retrieval‑Augmented Generation)

Technique where an AI retrieves fresh documents at answer time to ground responses. Your content must be crawlable, authoritative, and clearly structured to be selected.

Redirects (301 vs. 302)

– 301 = permanent signal consolidation
– 302 = temporary. Use correctly to preserve equity and avoid confusion.

Robots.txt

File that controls crawler access. Avoid blocking key pages; reference your sitemap.

Schema‑Linked Profiles

Exposing official profiles via Organization schema sameAs (LinkedIn, X, YouTube, etc.) to connect your entity across the web.

Search‑Augmented Systems

AIs that blend internal knowledge with web retrieval (e.g., AI Overviews). Ranking well in classic search often improves AI inclusion.

SEO Basics

Front‑loaded keywords in titles/meta, logical H1–H3 usage, and readable structure aid both users and machines.

Sitemap.xml

Machine list of important URLs; reference it in robots.txt and submit to GSC/Bing to speed discovery.

Social Mentions (Monitoring)

Track brand mentions across the web and social to measure sentiment; respond to praise and concerns to build goodwill.

SVG Animation

Great for UX and performance, but avoid relying on inline SVG alone for image search visibility; supplement with crawlable text content.

Timely Updates & Fresh Content

Show “Last Updated,” publish at a sustainable cadence, and refresh evergreen pages quarterly where value exists.

Title Tag & Meta Description

Front‑load primary terms; write human‑readable, high‑intent copy. These are scanned first by search and by many AI systems.

Topical Coverage

Depth across related subtopics that demonstrates expertise and improves retrieval odds for answer engines.

UGC (User‑Generated Content)

Comments/reviews can add helpful text signals when pages are otherwise light, but require moderation.

URL Structure

Clean, descriptive slugs improve UX and crawling; avoid excessive parameters for indexable pages.

Voice SEO

Optimizing for spoken answers where clicks are fewer. Concise, trustworthy facts and schema help; explore business models beyond purely click‑driven attribution.

Wikipedia Page

Draft in a sandbox with neutral tone and reliable independent sources (3–5+). Submit via Articles for Creation and add citations throughout.

Wikidata Item

Create and source statements (instance of, country, founding date, website, logo). Link to Wikipedia via Sitelinks and reference reliable URLs for each claim.