AI marketing glossary

HTTP redirect codes — 301 (Moved Permanently, passes SEO authority), 302 (Found, temporary, does not pass authority) — critical for URL changes.

Experimental method comparing two variants (A and B) of an element on randomized audiences to statistically measure which one converts better.

A B2B strategy that flips the funnel: you first identify a target account list (50–500 companies), then deploy coordinated marketing and sales efforts toward each account individually.

The share of new users who reach a predefined "aha moment" within their first days — the pivotal PLG metric, predictive of long-term retention.

The discipline of optimizing to be cited by AI engines (ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews) rather than ranking in traditional Google search results.

An AI system capable of breaking down a goal into a plan, executing tools, observing results, and iterating until the objective is reached — without step-by-step human supervision.

Hyperlink from an external site pointing to a page — a historically major SEO factor, weighted by the referring domain's authority and topical relevance.

A coherent set of editorial choices (tone, vocabulary, rhythm, register) that make a brand instantly recognizable across all its content.

Semi-fictional representation of a customer type based on real data — role, pain points, motivations, channels, vocabulary — used to guide messaging and product decisions.

Total marketing + sales spend divided by the number of new customers acquired in a period — the pivotal SaaS unit economics metric, always compared against LTV.

The HTML tag `<link rel="canonical">` that tells search engines which URL is the authoritative version of a page when multiple URLs serve identical or very similar content.

A prompting technique that asks the LLM to spell out its reasoning step by step before producing the final answer — dramatically improves accuracy on reasoning tasks.

Percentage of customers or revenue lost in a period — the critical distinction between logo churn (customers) and revenue churn (€) is often overlooked.

An analytical method that groups users by a common characteristic (typically their signup month) to observe how their behavior evolves over time.

The practice of contacting prospects who have never interacted with your brand — via email, LinkedIn, or phone — with a personalized message proposing a conversation.

Three web performance metrics (LCP, INP, CLS) that Google has used as a ranking factor since 2021 — measured under real-world conditions (CrUX).

An ultra-comprehensive article (3,000–10,000 words) serving as the definitive reference on a strategic topic — the anchor of a topic cluster, the primary target for backlinks and internal links.

The volume of URLs Googlebot is willing to crawl on a site within a given period — determined by the site's crawl capacity and Google's estimated crawl demand.

Interface element (button, link) that prompts the user to take a specific action — its copy, visual contrast, and placement determine the conversion rate.

The clicks / impressions ratio expressed as a percentage — measures how compelling a headline, thumbnail, CTA, or ad is to the audience that was exposed to it.

A sequential representation of all the interactions a prospect or customer has with a brand — from discovery to loyalty — broken down by stage, channel, emotion, and friction point.

Google's quality framework (since 2022, with the addition of E for Experience) that evaluates pages based on the author's experience, their expertise, the site's authority, and content trustworthiness.

Ratio (interactions / reach or followers) × 100 — a normalized measure of social content performance, calculated differently depending on the platform and tool.

Content whose value does not decline over time — as opposed to news content — the preferred content marketing investment for its long-term ROI.

Answer block displayed at the top of the Google SERP (before the organic #1) that directly quotes a web page excerpt to respond to a query — nicknamed "position zero".

A subset of AEO focused on generative engines (ChatGPT, Claude, Gemini, Perplexity) that produce synthesized text rather than just extracting snippets.

A data-driven marketing discipline combining rapid experimentation, unconventional techniques, and product loop exploitation to accelerate user growth.

The phenomenon by which an LLM generates a statement that appears plausible and confident but is factually false or fabricated — the number-one reliability risk in AI systems.

Structured description of the market segment that extracts the most value from a product — firmographics, technographics, triggers, and intent signals.

A quantified metric tied to a business objective — must be specific, measurable, attributable to an action, and linked to a decision threshold.

Standalone web page designed to convert targeted traffic on a single action — distinct from the main site by its single-objective focus and minimal navigation.

Free resource (ebook, template, audit, calculator) exchanged for an email address — the most widely used entry point for converting anonymous traffic into a named lead.

A method that assigns a numerical score to each lead based on firmographic attributes and behavioral signals (intent) to help sales teams prioritize their outreach.

A text file at the root of a site (https://example.com/llms.txt) that presents the brand to AI crawlers in a format they natively consume.

Total net revenue a customer generates over the entire length of their commercial relationship — for SaaS: ARPU × gross margin / monthly churn.

A rule that assigns credit for a conversion across multiple marketing touchpoints — first-touch, last-touch, linear, time-decay, position-based, or data-driven.

The platforms and practices that automate repetitive marketing workflows — triggered emails, nurturing sequences, lead scoring, segmentation — based on user behavior.

Sequential model of prospect progression — awareness, interest, consideration, conversion, retention — used to measure and optimize the pass-through rate at each stage.

An open protocol from Anthropic that standardizes how an AI agent connects to external tools or data sources — the USB-C equivalent for LLMs.

The three standard lead qualification statuses: MQL (marketing qualified), SQL (sales qualified after BANT confirmation), PQL (qualified by product usage — the pivotal PLG metric).

Monthly recurring revenue of a SaaS, normalized excluding one-shot payments — the pivot metric for SaaS management, broken down into new MRR, expansion MRR, churn MRR, and net new MRR.

The single metric that best captures the value a product delivers to its users — guides all product and marketing trade-offs, popularized by Sean Ellis and Facebook.

Customer satisfaction score from −100 to +100 based on the question "would you recommend our product?" rated 0 to 10 — % promoters (9-10) − % detractors (0-6).

Automated generation of SEO pages from templates × data (e.g., "best tool X for Y" scaled across hundreds of combinations), used by Airbnb, Zapier, and G2.

The discipline of designing instructions given to an LLM (ChatGPT, Claude, Gemini) to maximize the quality, reliability, and reproducibility of the generated responses.

An architecture that pairs an LLM with an upfront semantic retrieval step (vector DB) to ground generation in fresh or private data.

A text file at the root of a site that tells crawlers (Googlebot, GPTBot, etc.) which parts may be indexed.

The (gains − costs) / costs × 100 ratio measuring marketing profitability — distinct from ROAS, which counts revenue without deducting costs.

The discipline of optimizing to rank in organic search results (Google, Bing) — the upper tier of the traditional blue links.

Results page displayed by a search engine in response to a query — composed of organic links, ads, featured snippets, AI Overviews, knowledge panels, and more.

A brand's share of mentions or visibility relative to competitors on a given channel — broken down into social SOV, search SOV, and AEO SOV (AI citations).

Psychological principle by which individuals look to others' choices to make decisions — on the web: testimonials, logos, usage metrics, reviews, badges, case studies.

The schema.org vocabulary rendered as inline JSON-LD in an HTML page to describe its content in a machine-readable way to search and AI engines.

A data structure that represents entities and relations in a graph, with validity timestamps (valid_from / valid_to) to track how facts evolve over time.

A RAG technique applied to tool selection for an AI agent: embed the tool catalogue and dynamically retrieve the k most relevant tools per conversational turn.

Content architecture organized around broad "pillar" pages surrounded by specialized "cluster" pages that link back to the pillar — a topical authority signal for Google.

The perceived credibility of a site on a specific topic, measured by Google through the density and quality of content published on that topic — the dominant SEO factor in 2026, surpassing overall domain authority.

An AI agent memory system where each block is semantically typed (brand_voice, audience, user_prefs, etc.) rather than stored as free-form text.

Numerical representations of text units (word, sentence, document) in a multi-dimensional vector space where geometric proximity reflects semantic similarity.

Average number of new users an existing user brings in — an epidemiological formula. K > 1 = exponential viral growth, K < 1 = dampened growth.