AI Marketing + MCP Glossary 2026: 200+ Terms Defined
Plain-English definitions for every AI marketing and MCP term you'll encounter — from A/B testing to zero-shot prompting. No jargon, no fluff.
AI marketing has developed its own vocabulary fast — and the MCP ecosystem has added another layer on top of it. This glossary defines the terms you'll encounter when working with AI marketing tools, reading research, or setting up MCP connections.
Terms are listed alphabetically. Where a term is commonly misused or confused with another term, we've noted it. Where a term connects to Ooty's tools, we've linked to the relevant product page.
Numbers You'll See Throughout This Glossary
The most-cited AI marketing statistics -- each linked to its original source in the definitions above
88%
of marketers use AI tools daily
CoSchedule, 2025
87.4%
of AI referral traffic comes from ChatGPT
SE Ranking, 2025
15.9%
ChatGPT referral conversion rate (vs 1.76% Google)
SE Ranking, 2025
6x
higher transaction rates from personalised emails
Instapage/Experian, 2025
400%
content production speed increase with AI writing tools
Loopex Digital, 2025
Source: Multiple sources, 2025 | ooty.io
A
A/B Testing Comparing two versions of a piece of content, ad, or page to see which performs better with a real audience. You split your audience and show Group A version 1, Group B version 2, then measure a defined outcome (clicks, conversions, time on page). AI tools can now suggest which variants to test based on historical performance patterns.
Agentic AI An AI system that can take sequences of actions autonomously to complete a goal — rather than just answering a single question. An agentic AI might research a topic, write a draft, check it against your style guide, and send it to a teammate, all without being prompted at each step. Distinct from generative AI, which generates a single output per prompt.
AI Overview (Google) The AI-generated summary that appears at the top of some Google search results pages, above the traditional blue-link results. The brand appears in an AI Overview response only if Google's systems select it as a credible source — which is distinct from ranking on page one.
AI Search Search interfaces powered by large language models that generate answers rather than returning a list of ranked links. Examples: ChatGPT search, Perplexity, Google AI Mode, Microsoft Copilot search. Also called "generative search."
AI Visibility The degree to which a brand, product, or piece of content appears in AI-generated answers across platforms like ChatGPT, Perplexity, Google AI Overviews, and Gemini. A nascent metric distinct from search ranking positions.
Algorithm A set of rules or instructions that a system uses to make decisions. In marketing contexts: the rules a platform uses to decide who sees what content (social media algorithms), or which pages rank for which queries (search algorithms).
Anchor Text The clickable text in a hyperlink. Anchor text signals to search engines what the linked page is about. Relevant for SEO backlink strategy.
Answer Engine A search system that returns a direct answer rather than a list of links. Perplexity explicitly describes itself as an answer engine. Increasingly, traditional search engines are adding answer engine features (Google AI Overviews, Bing Answers).
API (Application Programming Interface) A defined way for one software system to communicate with another. When a marketing tool accesses your Google Analytics data, it's using Google's API. MCP is built on top of APIs — MCP servers typically wrap existing APIs to make them accessible to AI assistants.
Attribution Crediting the touchpoints that led to a conversion. In marketing analytics, attribution models decide which channel or campaign gets "credit" for a sale. Common models: last-click (credit to the final touchpoint before purchase), first-click (credit to the first touchpoint), linear (credit split across all touchpoints), data-driven (credit distributed based on statistical analysis).
Audience Segmentation Dividing a target market into distinct groups based on shared characteristics — demographics, behaviours, purchase history, etc. AI tools can identify segments that weren't obvious from manual analysis.
Automation Using software to perform repetitive tasks without manual input. In marketing: automated email sequences, automated bid management in paid ads, automated social media scheduling. Distinct from AI — automation follows fixed rules; AI adapts based on patterns.
Average Position (Search Console) The mean rank at which your page appears in Google search results for a given query. Average position 1.0 means you always appear first. An average position of 7.5 means you appear somewhere between 1st and 14th on average. Used in Ooty Octopus to identify ranking opportunities.
B
Backlink A link from one website to another. Backlinks from credible, relevant sites signal authority to search engines and are a core ranking factor for traditional SEO. Also relevant for GEO — AI systems tend to cite content that has external links pointing to it.
BERT (Bidirectional Encoder Representations from Transformers) A language model developed by Google, used in its search algorithm. Relevant because BERT was one of the first major shifts toward semantic understanding in search — understanding what a query means, not just which keywords it contains. A precursor to the current generation of LLMs.
Bid Strategy In paid advertising, the approach used to determine how much you pay per click, impression, or conversion. AI-powered bid strategies (Google's Smart Bidding, Meta's Advantage+ bidding) adjust bids in real time based on predicted conversion probability.
Bounce Rate The percentage of sessions where a user visits one page and leaves without interacting further. In GA4, the equivalent metric is "engagement rate" (sessions where the user stayed more than 10 seconds, viewed multiple pages, or completed a conversion). Used in Ooty Compass analysis.
Brand Safety Ensuring that your advertising doesn't appear alongside content that conflicts with your brand's values. Relevant for programmatic advertising and YouTube ads, where AI systems place ads automatically.
Broad Match (Keywords) A Google Ads keyword match type where your ad can show for searches related to your keyword — including synonyms, related concepts, and paraphrases. Google's AI expanded what "related" means significantly over the past few years.
C
Cache A temporary store of data. In MCP contexts: some MCP servers cache responses from external APIs to reduce latency and avoid hitting rate limits. Cached data may be slightly out of date. Reputable MCP servers indicate data freshness.
Call-to-Action (CTA) A prompt that asks a user to take a specific action: "Sign up," "Download the guide," "Start your free trial." AI tools are increasingly used to test CTA variants and personalise CTAs based on user behaviour.
ChatGPT OpenAI's AI assistant, the most widely used in its category. Relevant to marketing because: (a) it drives 87.4% of AI referral traffic across major industries, (b) brands want their products to be mentioned when users ask ChatGPT for recommendations, and (c) ChatGPT supports MCP via its Agents SDK and Responses API.
Citation (AI Search) When an AI system references a source by linking to it within a generated response. Getting cited in AI answers is the goal of GEO. AI systems typically cite 2–7 domains per response — far fewer than the 10 blue links on a Google results page.
Click-Through Rate (CTR) The percentage of users who click on a link, ad, or search result after seeing it. Calculated as clicks ÷ impressions × 100. A low CTR in Search Console despite high impressions indicates the page title or meta description isn't compelling enough for the search intent. Tracked in Ooty Octopus.
Claude Anthropic's AI assistant. The primary AI client for MCP — Claude Desktop and Claude's API natively support MCP connections. Ooty's MCP servers are designed to work with Claude.
CMS (Content Management System) Software used to create and manage digital content. Examples: WordPress, Webflow, Contentful. Some CMS platforms have MCP servers that allow AI assistants to read and write content directly.
Consent Mode (Google) A framework for adjusting how Google's tracking tags behave based on users' cookie consent choices. Essential for GA4 compliance in markets with privacy regulations (GDPR in Europe, similar laws elsewhere). Relevant to Ooty Compass setup.
Content Score A numerical measure of how well a piece of content is optimised for a target keyword, as calculated by content optimisation tools like Surfer SEO or Clearscope. Based on comparing your content to the top-ranking pages for that keyword.
Conversion A desired action completed by a user — a purchase, a form submission, a sign-up, a download. Conversion rate is conversions ÷ sessions × 100. Different platforms define and count conversions differently.
Conversion Rate Optimisation (CRO) The practice of improving the percentage of visitors who complete a desired action. AI tools are increasingly used in CRO to identify friction points, suggest test variants, and personalise experiences.
Copilot (Semrush) Semrush's AI assistant that analyses data from six Semrush tools and provides prioritised recommendations. Available on all Semrush plans. Different from Microsoft Copilot (which is built on OpenAI's GPT models).
Core Web Vitals A set of Google-defined metrics measuring page experience: Largest Contentful Paint (LCP, loading speed), Interaction to Next Paint (INP, interactivity), and Cumulative Layout Shift (CLS, visual stability). Used as a ranking factor. Tracked via PageSpeed Insights, accessible through Ooty Octopus.
Cost Per Acquisition (CPA) The total spend divided by the number of conversions. A key metric in paid advertising performance. AI bid strategies in Google Ads and Meta Ads optimise toward a target CPA.
Cost Per Click (CPC) The amount paid each time a user clicks an ad. CPC varies by keyword competition, audience, and bid strategy. Tracked in Ooty Falcon.
Cost Per Mille (CPM) Cost per 1,000 impressions. The standard buying metric for brand awareness campaigns on display and social platforms. Lower CPM doesn't always mean better value — engagement rate matters too.
Crawl Budget The number of URLs Googlebot will crawl on your site within a given timeframe. Large sites with many pages may have important pages left uncrawled if crawl budget is wasted on low-value URLs.
CRM (Customer Relationship Management) Software that tracks customer interactions, sales pipelines, and marketing touchpoints. Examples: HubSpot, Salesforce. HubSpot has an official MCP server; Salesforce connects via Composio.
CTR (Click-Through Rate) See "Click-Through Rate" above.
D
Data Freshness How recently data was fetched or updated. In MCP contexts: critical to know because some MCP servers cache API responses. Old data can lead to incorrect analysis. Good MCP implementations note the data timestamp in responses.
Demand Gen (Google Ads) A Google Ads campaign type designed to drive product discovery across YouTube, Gmail, and Google Discover. Uses AI to target users with high purchase intent.
Domain Authority (DA) / Domain Rating (DR) Third-party metrics (DA from Moz, DR from Ahrefs) that estimate a domain's overall authority based on its backlink profile. Not Google metrics — Google doesn't use DA or DR directly — but useful as a relative indicator when comparing sites.
Dynamic Search Ads (DSA) Google Ads campaign type where Google automatically generates ad headlines and landing page targeting based on your website content. Uses Google's understanding of your site to match queries to relevant pages.
E
E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) Google's framework for evaluating content quality, used in its Search Quality Rater Guidelines. Experience refers to first-hand experience with the topic; Expertise to subject matter knowledge; Authoritativeness to the reputation of the author and site; Trustworthiness to accuracy and transparency. Relevant for both SEO and GEO.
Embedding A numerical representation of text (or images, audio, etc.) that captures semantic meaning. AI systems use embeddings to measure similarity between pieces of content. When an AI retrieves relevant context for a query, it often compares embeddings to find the closest matches. Used in RAG (Retrieval-Augmented Generation) systems.
Engagement Rate The percentage of users who interact with content in some meaningful way — likes, comments, shares, clicks, video views. Platform definitions vary significantly. In GA4, engagement rate measures sessions lasting more than 10 seconds, viewing multiple pages, or completing a conversion.
Entity A real-world thing — a person, place, organisation, product, concept — that AI systems and search engines recognise and can reason about. Your brand is an entity. Making sure AI systems have accurate, comprehensive information about your brand entity improves GEO performance. Google's Knowledge Panel is based on entity recognition.
Exact Match (Keywords) A Google Ads keyword match type where your ad shows only for searches that match your keyword exactly (or very close variants). More restrictive than phrase match or broad match; gives more control but limits reach.
F
Feature Gate A mechanism to enable or disable functionality for specific users or environments. In software development: feature flags let you ship code without making it visible to all users. Ooty uses feature gates to control which tools are visible based on licence type.
Fine-Tuning Training a pre-trained AI model on a specific dataset to improve its performance on a particular task. A marketing agency might fine-tune a model on their past copywriting to replicate their voice. Distinct from prompting — fine-tuning modifies the model's weights; prompting guides the existing model.
First-Party Data Data you collect directly from your audience — email subscribers, website behaviour, purchase history, survey responses. Increasingly important as third-party cookies are phased out. First-party data can be connected to AI tools for personalisation without privacy complications.
Frequency (Ads) The average number of times a unique user sees a specific ad. High frequency can cause ad fatigue — the point at which seeing an ad repeatedly reduces effectiveness or creates negative brand associations.
G
GA4 (Google Analytics 4) Google's current analytics platform, which replaced Universal Analytics in 2023. Event-based rather than session-based, meaning it tracks individual user actions rather than page views by default. Requires additional configuration to match UA's reporting. Accessible via Ooty Compass.
Generative AI AI systems that generate new content — text, images, audio, video, code — based on patterns learned during training. GPT-4, Claude, Gemini, and Stable Diffusion are all generative AI systems. Distinct from discriminative AI, which classifies or predicts from existing data.
Generative Engine Optimisation (GEO) The practice of optimising content to appear in AI-generated answers from systems like ChatGPT, Perplexity, and Google AI Overviews. Distinct from SEO, which optimises for traditional search ranking. See our GEO vs SEO guide for a full comparison.
Gemini (Google) Google's family of AI models, used in Google AI Overviews, Workspace AI features, and Google's AI search products. Competes with Claude and GPT-4.
Google AI Overviews Google's AI-generated summaries that appear at the top of some search result pages. Previously called "Search Generative Experience" (SGE) during testing. Powered by Gemini.
Google Indexing API A Google API that lets site owners notify Google when a page is added or removed, prompting faster crawling and indexing. Particularly useful for sites with large numbers of new pages (job boards, news sites). Available through Ooty Octopus.
GPT (Generative Pre-trained Transformer) A class of large language models developed by OpenAI. GPT-4 and its successors power ChatGPT. The architecture (transformer + pre-training + fine-tuning) influenced most modern LLMs.
H
Hallucination When an AI system generates false information presented as fact. A significant problem for marketing — AI tools can make up statistics, misattribute quotes, or invent product features. Always verify AI-generated factual claims before publishing. MCP tools reduce hallucination risk for data-driven tasks by pulling from real sources.
Heatmap A visual representation of where users click, scroll, or focus attention on a page. Tools like Hotjar and Microsoft Clarity generate heatmaps. AI-powered analysis of heatmaps can identify UX problems automatically.
Hyperparameter Configuration settings for an AI model that aren't learned during training but set in advance. For marketers using AI tools: temperature (how random/creative the output is) and max tokens (maximum output length) are common hyperparameters you can adjust in API-based tools.
I
Impression A single instance of content being displayed. In search: your page appeared in a results page. In advertising: your ad was served to a user (whether or not they clicked or noticed it). Tracked in Ooty Octopus via Search Console.
Impression Share In Google Ads, the percentage of impressions your ads received compared to the total number they were eligible to receive. Low impression share due to rank means your bids are too low; low impression share due to budget means you're running out of budget before the day ends.
Intent (Search Intent) What a user actually wants to accomplish with a search query. The four main types: informational (learning something), navigational (finding a specific site), commercial (researching before buying), transactional (ready to buy). Matching content to intent is fundamental to SEO and increasingly to GEO.
J
JSON-LD A format for embedding structured data in web pages, used for schema markup. The recommended format by Google for implementing Schema.org vocabulary. Used to communicate to search engines and AI systems what a page is about.
JSON-RPC The communication protocol underlying MCP. When Claude calls an MCP tool, it sends a JSON-RPC request to the MCP server, which processes it and returns a JSON-RPC response. You don't need to understand JSON-RPC to use MCP — it's infrastructure.
K
Keepa A service that tracks Amazon product pricing history, rank history, and availability. Used by Ooty Canopy to provide historical Amazon data to Claude.
Keyword A word or phrase that users type into a search engine. In SEO: you identify keywords your target audience searches for and create content that matches that intent. In paid search: you bid to have your ad shown for specific keywords.
Keyword Difficulty A score (typically 0–100) estimating how hard it would be to rank on page one for a given keyword. Higher difficulty = more competition = more effort required. Available in Ahrefs, Semrush, and other SEO tools. Not a metric available directly from Google.
Knowledge Graph (Google) Google's database of entities and their relationships. The "Knowledge Panel" on the right side of some search results is powered by the Knowledge Graph. Getting into the Knowledge Graph improves brand visibility in both traditional search and AI-generated answers. Accessible via Ooty Octopus.
KPI (Key Performance Indicator) A metric used to evaluate performance against goals. In marketing: could be organic sessions, conversion rate, ROAS, subscriber growth, etc. The choice of KPIs determines what you optimise for.
L
Landing Page A web page a user arrives on after clicking a link. In paid advertising: a dedicated page designed specifically to convert visitors from a particular campaign. Landing page quality affects Google Ads Quality Score.
Latency The time delay between making a request and receiving a response. In MCP contexts: the time it takes for Claude to call an MCP tool and receive data. Most well-implemented MCP servers respond in 1–5 seconds; complex queries to large datasets may take longer.
LCP (Largest Contentful Paint) One of Google's Core Web Vitals. Measures how long it takes for the largest element on a page to become visible. Google considers LCP under 2.5 seconds "good," 2.5–4 seconds "needs improvement," and over 4 seconds "poor."
Lead Scoring Assigning numerical values to leads based on their characteristics and behaviour, to prioritise those most likely to convert. AI-powered lead scoring uses machine learning to weight factors based on historical conversion patterns.
LLM (Large Language Model) A type of AI model trained on large amounts of text to understand and generate language. GPT-4, Claude, Gemini, and Llama are all LLMs. The "large" refers to the number of parameters (weights) in the model — the largest models have hundreds of billions.
Long-tail Keyword A keyword with three or more words, typically lower search volume but higher purchase intent and lower competition. "Running shoes" is a head term; "best running shoes for flat feet women" is a long-tail keyword. Long-tail keywords are typically easier to rank for and convert better.
M
Machine Learning (ML) A subset of AI where systems learn patterns from data rather than being explicitly programmed with rules. The model "learns" from examples. Most modern AI marketing tools use ML to improve their recommendations over time.
MCP (Model Context Protocol) An open standard developed by Anthropic that allows AI assistants to connect to external data sources and tools in a standardised way. MCP servers provide data; MCP clients (like Claude) consume it. The marketing relevance: MCP lets Claude access your real marketing data — Search Console, GA4, YouTube, etc. — rather than working from information you paste in. See our Definitive Guide to MCP for Marketers.
How MCP Fits Together
The three pieces of the MCP architecture and how they connect
MCP Client
Your AI assistant
Claude, ChatGPT, Gemini
MCP Server
The connector
Ooty Octopus, Ahrefs, HubSpot
Tools & Data
Your actual data
Search Console, GA4, YouTube
From your perspective: You ask Claude a question. Claude calls the MCP server. The server fetches your data. You get a real answer. The architecture is invisible.
Source: Model Context Protocol, modelcontextprotocol.io | ooty.io
MCP Client An AI application that can connect to MCP servers — Claude Desktop, Claude API, ChatGPT Agents, and others. The client sends requests to MCP servers and receives data back.
MCP Server A connector that exposes data or tools to MCP clients. An MCP server for Google Analytics, for example, knows how to authenticate with Google, query the GA4 API, and return data in a format Claude can understand. Ooty builds MCP servers for marketing data sources.
MCP Tool
A specific action or query that an MCP server offers. A Search Console MCP server might offer tools like get_top_queries, get_page_impressions, compare_periods. Claude calls these tools during conversations when it needs specific data.
Meta Description The snippet of text that appears below a page title in search results. Not a direct ranking factor, but affects CTR — a well-written meta description increases the percentage of people who click your result.
Meta (Platform) The company owning Facebook, Instagram, and WhatsApp. In advertising contexts: "Meta Ads" refers to the advertising system covering Facebook, Instagram, and the Meta audience network. Accessible via Ooty Falcon and Ooty Echo.
Multi-Touch Attribution An attribution model that gives credit to multiple touchpoints in the customer journey, rather than just the first or last. AI-powered multi-touch models can weight touchpoints based on statistical analysis of what actually influences conversions.
N
Natural Language Processing (NLP) A field of AI focused on enabling computers to understand, interpret, and generate human language. Relevant to marketing: NLP powers chatbots, email personalisation systems, content analysis, and sentiment analysis.
Natural Language Query A question or request phrased in ordinary human language, rather than in keywords or code. MCP enables natural language queries against marketing data — instead of building a GA4 custom report, you ask Claude "which landing pages have the worst bounce rate?"
Neural Network A computational system loosely inspired by the human brain, used as the underlying architecture for most modern AI systems. LLMs are built on a type of neural network called a transformer.
O
OAuth An authorisation framework that lets users grant an application access to their data on another service, without sharing passwords. When you connect Ooty to your Google account, you're completing an OAuth flow. Standard for secure API access.
On-Page SEO Optimisation done within individual web pages — title tags, meta descriptions, heading structure, internal links, content quality. Contrasted with off-page SEO (backlinks, brand mentions) and technical SEO (site architecture, page speed).
Organic Search Unpaid search results, as opposed to paid search ads. Growing organic traffic requires SEO. Organic search data is available in Search Console and GA4; AI-powered analysis via Ooty Octopus and Ooty Compass.
Output Token In LLM contexts, a token is the basic unit of text (roughly 0.75 words). AI APIs often charge per token generated. Relevant for teams using AI at scale: a 2,000-word blog post takes more output tokens than a 200-word summary.
P
PageRank Google's original algorithm for ranking web pages based on the quantity and quality of inbound links. Still a core component of how Google evaluates authority, though the algorithm has become vastly more complex. The concept that "a link is a vote" derives from PageRank.
PageSpeed Insights Google's free tool for measuring web page performance against Core Web Vitals metrics. Returns a score (0–100) and specific recommendations for improvement. Data accessible via Ooty Octopus.
Perplexity An AI search engine that generates answers with cited sources. Drives 15% of AI referral traffic globally and has a notably high conversion rate from referral clicks.
Performance Max (PMax) A Google Ads campaign type that uses AI to serve ads across all of Google's channels — Search, Display, YouTube, Gmail, Maps — automatically. Google's systems choose the best channels and creatives for each auction.
Personalisation Tailoring content, offers, or experiences to individual users based on their characteristics or behaviour. AI makes personalisation viable at scale — personalised emails deliver 6x higher transaction rates than non-personalised ones.
Phrase Match (Keywords) A Google Ads keyword match type where your ad shows for searches containing your keyword in the same order, possibly with additional words before or after.
Prompt An instruction or question given to an AI system. In marketing: a prompt might be "Write a subject line for an email promoting our summer sale" or "Analyse this campaign data and identify the three biggest opportunities."
Prompt Engineering The practice of crafting inputs to AI systems to get better, more reliable outputs. Good prompt engineering for marketing tasks involves: being specific about the audience and goal, providing relevant context, specifying the format you want, and giving the AI a role to play.
Q
Quality Score (Google Ads) A 1–10 rating Google assigns to keywords based on expected click-through rate, ad relevance, and landing page experience. Higher Quality Scores mean lower CPCs and better ad positions.
Query A search term entered by a user. In Search Console: the queries report shows what searches led users to your site. Different from a keyword — keywords are what you target; queries are what users actually typed.
R
RAG (Retrieval-Augmented Generation) An AI architecture where the model retrieves relevant information from an external knowledge base before generating a response. This gives AI access to current or proprietary information beyond its training data. MCP is one implementation approach for enabling retrieval.
ROAS (Return on Ad Spend) Revenue generated divided by advertising spend. A ROAS of 3 means you earned £3 for every £1 spent on ads. The target ROAS varies significantly by industry and business model.
Rate Limit A restriction on how many API requests you can make in a given time period. Marketing APIs all have rate limits — Google, Meta, YouTube, etc. MCP servers must handle rate limits gracefully. Large data pulls may take longer than simple queries because the MCP server has to paginate requests.
Remote MCP An MCP server hosted on a remote server (accessible via URL) rather than running locally on your computer. Ooty's MCP servers are available as remote MCP — you don't need to install anything locally, just configure Claude with your Ooty endpoint. Contrast with local MCP.
Retargeting Showing ads to users who have previously visited your website or interacted with your brand. Uses cookies or pixel tracking to identify past visitors across the web. Also called remarketing.
S
Schema Markup / Schema.org Structured data vocabulary used to annotate web pages so that search engines and AI systems can understand their content more precisely. Article, Product, FAQ, HowTo, and Review are common schema types for marketing content. Implementing schema markup is a GEO-relevant tactic.
Search Console Google's free tool showing how your site appears in Google Search — impressions, clicks, average position, and queries. The most accurate source of organic search performance data for your own site. Core data source for Ooty Octopus and Ooty Compass.
Search Intent See "Intent."
Semantic SEO Creating content that covers a topic thoroughly and accurately, rather than repeating a single keyword many times. Google's understanding of language has improved to the point where semantic relevance matters more than exact keyword frequency.
Sentiment Analysis AI-powered analysis of whether text expresses positive, negative, or neutral sentiment. Used in marketing to analyse customer reviews, social comments, survey responses, and brand mentions at scale.
Session In web analytics: a period of user activity on your website. GA4 defines a session as starting when a user arrives and ending after 30 minutes of inactivity.
Smart Bidding Google Ads' set of AI-powered bid strategies that automatically adjust bids to optimise for your defined goal: target CPA, target ROAS, maximise conversions, or maximise conversion value.
Social Proof Evidence that other people have used and approved of a product or service — reviews, testimonials, user numbers, press mentions. AI tools are used to generate, personalise, and test social proof elements in marketing.
Structured Data Data organised in a defined format that machines can easily parse. In SEO/GEO contexts: schema markup is structured data embedded in HTML. AI systems prefer structured, well-organised data over prose when extracting specific facts.
Surfer SEO A content optimisation tool that analyses top-ranking pages for a keyword and recommends terms, structure, and word count. See the AI SEO Tools Comparison for more.
T
Temperature A parameter controlling the randomness of AI output. Low temperature (0–0.3): more predictable, consistent, factual. High temperature (0.7–1.0): more creative, varied, sometimes less accurate. For data analysis tasks: lower temperature. For creative copywriting: higher temperature.
Third-Party Cookie A cookie set by a domain other than the one the user is visiting — used for cross-site tracking and ad targeting. Being phased out by browsers. Relevant to marketing because audience targeting and attribution methods that relied on third-party cookies need to be replaced with privacy-compliant alternatives.
Token (AI) The basic unit of text processing in LLMs. A token is roughly 0.75 words (or 4 characters). GPT-4 processes requests in units of tokens; pricing for AI APIs is typically per token. A 1,000-word article contains approximately 1,333 tokens.
Tool (MCP) A specific capability exposed by an MCP server. When Claude connects to an MCP server, it discovers what tools are available. Tools have names, descriptions, and input parameters. Claude calls tools as needed during a conversation to retrieve data or perform actions.
Topical Authority A search ranking signal based on whether your site demonstrates deep expertise on a topic, not just individual pages ranking for individual keywords. Building topical authority requires creating comprehensive, interlinked content covering a topic from multiple angles.
Traffic Source The origin of website visits — organic search, paid search, direct, referral, social, email. GA4 tracks traffic sources and attributes conversions to them. Available via Ooty Compass.
U
UA (Universal Analytics) The previous version of Google Analytics, replaced by GA4 in July 2023. Data from UA cannot be directly imported into GA4. Historical UA data remained accessible until July 2024.
User Journey / Customer Journey The sequence of touchpoints a customer has with a brand before making a purchase. AI tools map user journeys by analysing cross-channel data, identifying where users drop off and which touchpoints correlate with conversion.
V
Value Proposition A clear statement of what benefit a product delivers to a customer and why they should choose it over alternatives. Core to effective marketing messaging. AI tools are used to test value proposition variants in ads and landing pages.
Velocity (Content) How quickly you're publishing content. AI tools have increased content velocity for many teams — AI-powered content creation can increase production speed by 400% according to some studies. Higher velocity is only valuable if quality is maintained.
Velite The build tool used by Ooty's blog system to process MDX content files into the blog. Handles frontmatter parsing, content transformation, and static generation.
W
Web Vitals See "Core Web Vitals."
Webhook A mechanism for one application to notify another when an event occurs — a form submission, a payment, a product purchase. In marketing automation: webhooks trigger actions in connected systems.
Word Count The number of words in a piece of content. Long-form content (2,000+ words) tends to rank better for competitive informational queries, but correlation is not causation — the content needs to be genuinely comprehensive, not padded. AI content optimisation tools often recommend word counts based on what's ranking.
Y
YouTube Analytics Performance data for YouTube channels — views, watch time, average view duration, subscriber changes, audience retention, click-through rate on thumbnails. Accessible through YouTube Studio and via Ooty Iris for conversational analysis.
YouTube Studio YouTube's creator dashboard — the web interface where you manage videos, view analytics, respond to comments, and access monetisation settings. The data behind this dashboard is accessible via Ooty Iris.
Z
Zero-Click Search A search where the user gets their answer directly from the search results page without clicking through to any website. Featured snippets, knowledge panels, and AI Overviews all drive zero-click behaviour. Growing relevance of zero-click searches is one reason why GEO (getting your brand mentioned in AI answers) matters alongside traditional ranking.
Zero-Shot Prompting Asking an AI to perform a task without providing examples. Contrasted with few-shot prompting (providing 2–5 examples) and one-shot prompting (providing one example). For complex or highly specific marketing tasks, few-shot prompting typically produces better results.
Where These Terms Connect to Real Tools
Each Ooty product touches multiple concepts defined in this glossary
Related terms: Average Position, CTR, Impressions, PageSpeed, Knowledge Graph, Search Console
Related terms: GA4, Bounce Rate, Conversion, Sessions, Traffic Source, Consent Mode
Related terms: Engagement Rate, Meta, Sentiment Analysis
Related terms: YouTube Analytics, YouTube Studio
Related terms: CPC, CPM, CPA, ROAS, Quality Score, Smart Bidding
Related terms: Keepa, Amazon product data
Source: ooty.io | ooty.io
This glossary is updated periodically. Terms in rapidly evolving areas (AI search, MCP ecosystem) may change faster than we can update them. If you spot an error or a missing term, we'd welcome the feedback.
Sources:
From Ooty
AI native marketing tools for SEO, Amazon, YouTube, and social — replace your expensive dashboards.
Start free
Written by
Lola Reeves
Head of Content at Ooty. Covers AI marketing research and data strategy.
Related posts
Best MCP Servers for Marketers 2026: The Complete Directory
The MCP ecosystem has grown fast. Over 5,800 MCP servers are now available, which sounds impressive until you try to find the five that are actually useful for your job. This directory cuts through the noise. It covers the MCP servers that matter most for mark
The Definitive Guide to MCP for Marketers (2026)
There's a good chance you've heard "MCP" thrown around in AI conversations lately and nodded along without being entirely sure what it means. This guide is for you. MCP — the Model Context Protocol — is the reason your AI assistant can suddenly look at your li
AI Marketing Statistics 2026: The Data You Can Actually Use
Every number on this page comes from a named, linked source. No invented percentages. No "studies show" without the study. Use this as your reference when you need real data for a strategy doc, a pitch, or a board presentation. We update it regularly — last re