The SEO glossary for buyers
Short, sourced definitions of the terms that come up when you buy search work: SEO, GEO and AI search, conversion rate optimization, analytics and the platforms your site runs on. Each entry opens on a plain answer, says why it matters to you, and links to the deeper guide.
30 minutes. No pitch deck. You leave with a prioritised plan.
Search has picked up a second vocabulary in two years. The old one, rankings, backlinks, Core Web Vitals, never went away, and a new one arrived on top of it: AI Overviews, AI Mode, query fan-out, citations, GEO. Proposals mix the two freely, and it is hard to judge a plan when half of its nouns are unfamiliar.
This glossary is the reference we wish buyers had before a sales call. Every entry opens on a definition short enough to read in one breath, then says two things a generic dictionary does not: why the term should matter to someone paying for the work, and how it is actually measured or applied. Where a claim rests on a search engine's documentation or a piece of research, the source is listed at the foot of the entry.
How is the glossary organised?
Alphabetically, with a filter for five areas: SEO, GEO and AI search, conversion rate optimization, analytics, and the platforms sites are built on. Terms link to their neighbours, so an entry on AI Overviews points on to query fan-out and zero-click searches, and each one links down to the longer guide on the blog when that guide is live.
What will you not find here?
Invented numbers. Where a figure appears it comes from the named source. We also leave out jargon that exists only to sell a tool, and we say plainly when a term is overhyped, as the entry on llms.txt does.
A
- A/B test
- An A/B test is an experiment that splits visitors at random between two versions of a page or element, the original and a variation, and compares how each performs on one chosen goal, such as sign-ups or purchases, to find out whether the change caused a real difference or just random variation.
- Above the fold
- Above the fold is the part of a webpage visible on screen before a visitor scrolls, named after the fold on a physical newspaper.
- AI citation
- An AI citation is a link or named source that an AI assistant attaches to its answer to show where a statement came from.
- AI crawler
- An AI crawler is an automated bot run by an AI company to fetch web pages, either to build a search index its assistant retrieves from, to fetch a page a user has asked about, or to collect training data.
- AI hallucination
- An AI hallucination is a confident, fluent statement an AI model produces that is factually wrong, made up, or not actually supported by the sources it cites.
- AI visibility
- AI visibility is how often, how accurately and how favourably a brand appears in the answers AI assistants and AI-powered search features give to relevant buyer questions, whether as a citation, an unlinked mention, or a recommendation.
- Alt text
- Alt text is a written description added to an image's HTML alt attribute.
- Anchor text
- Anchor text is the visible, clickable words in a hyperlink.
- Answer engine
- An answer engine is a search tool, such as ChatGPT, Perplexity, Claude or Google's AI Overviews, that responds to a query with a direct, synthesised answer written in natural language rather than a list of links a person has to click through and read themselves to find the answer.
- Answer engine optimization (AEO)
- Answer engine optimization (AEO) is the practice of structuring content so that search features and AI assistants can lift a direct answer from it, whether that answer appears as a featured snippet, a voice assistant reply, a People Also Ask box or a cited sentence inside an AI-generated summary.
- Attribution model
- An attribution model is the rule a reporting system uses to decide how much credit each marketing touchpoint gets for a conversion when a visitor interacted with several channels before converting, such as an organic search visit followed later by a direct return visit that completes a purchase.
- Average order value (AOV)
- Average order value is the average amount spent each time a customer places an order, calculated by dividing total revenue by the number of orders over a period.
- Average position
- Average position is the mean ranking position a page held in Google's search results for a given query or set of queries, as reported in Search Console, averaged across every time it appeared, at any position, not only when a visitor actually clicked through.
B
- Backlink
- A backlink is a link from another website to a page on your site.
- Branded search
- Branded search is a query that includes a business's own name or a close variant of it, such as "RankPearl reviews" or "RankPearl pricing," as opposed to a non-branded query that describes a need or category without naming any specific business, such as "SEO agency."
- Breadcrumbs
- Breadcrumbs are a row of links near the top of a page showing the path back to the homepage through each parent category, such as Home, then Shoes, then Running Shoes.
C
- Call to action (CTA)
- A call to action is the specific instruction, usually a button or a short line of text, that tells a visitor exactly what to do next, such as "Book a free audit" or "Start free trial." It is the point where a page converts interest into a measurable action.
- Canonical URL
- A canonical URL is the one version of a page that a site owner or a search engine treats as the master copy when several URLs show the same or very similar content.
- Cart abandonment rate
- Cart abandonment rate is the share of shopping carts that are started but never completed as a purchase, calculated as one minus completed transactions divided by carts created.
- ChatGPT search
- ChatGPT search is the web search capability built into ChatGPT that lets it look up current information and cite sources when answering a query, rather than relying only on what it learned during training.
- Click-through rate (CTR)
- Click-through rate (CTR) is the share of people who saw a link and clicked it, calculated as clicks divided by impressions.
- Content chunking
- Content chunking is splitting a longer document into smaller, self-contained sections before it is indexed for AI retrieval, so a system can fetch and cite the specific passage that answers a question instead of returning an entire page.
- Content decay
- Content decay is the gradual loss of search traffic a page experiences over time as its information goes out of date, competitors publish more current pages, or the query it targets changes in intent, even though the page itself was never edited or removed.
- Content delivery network (CDN)
- A content delivery network is a distributed set of servers in different geographic locations that cache and serve a site's static content, such as images, scripts and stylesheets, from a location physically closer to each visitor.
- Conversion funnel
- A conversion funnel is the sequence of steps a visitor moves through on the way to a goal, such as landing page, product page, cart, checkout, confirmation, with a smaller number of people completing each successive step.
- Conversion rate
- Conversion rate is the percentage of visitors who complete a defined goal, such as a purchase, a form submission or a booked call, out of total visitors in a given period.
- Conversion rate optimization (CRO)
- Conversion rate optimization (CRO) is the practice of increasing the share of website visitors who take a desired action, such as booking a call, submitting a form or completing a purchase, by finding what stops them and testing changes to pages, forms and checkout flows.
- Core update
- A core update is a broad change Google makes to its main ranking systems, rolled out several times a year and announced publicly, that can move rankings across many sites and topics at once.
- Core Web Vitals (CWV)
- Core Web Vitals (CWV) are three Google metrics that describe real users' experience of a page: Largest Contentful Paint for loading, Interaction to Next Paint for responsiveness, and Cumulative Layout Shift for visual stability.
- Crawling
- Crawling is the process by which a search engine's automated bot, such as Googlebot, follows links and sitemap entries to discover URLs on the web and download their content.
- Cumulative Layout Shift (CLS)
- Cumulative Layout Shift is a Core Web Vitals metric that scores how much a page's visible content unexpectedly moves as it loads, such as text jumping down when an ad or image loads late.
- Customer acquisition cost (CAC)
- Customer acquisition cost is the average amount spent to acquire one new customer through a given channel, calculated by dividing total spend on that channel over a period by the number of new customers it produced in the same period.
D
- Disavow file
- A disavow file is a list of backlinks or linking domains a site owner submits to Google, asking Google to ignore them when assessing the site, typically because the links were bought, come from a spam attack, or are otherwise damaging and cannot be removed directly at the source.
- Domain Authority (DA)
- Domain Authority is a third-party metric from Moz, scored 0 to 100, that predicts how well a whole domain is likely to rank in search results based on its backlink profile.
- Duplicate content
- Duplicate content is substantial text that appears, identically or in near-identical form, on more than one URL, whether within one site or across different domains.
E
- E-E-A-T
- E-E-A-T stands for experience, expertise, authoritativeness and trustworthiness.
- Engagement rate
- Engagement rate is a GA4 metric measuring the percentage of sessions that were "engaged," meaning they lasted at least ten seconds, included a key event, or included at least two page or screen views.
- Entity (search)
- An entity is a specific, uniquely identifiable thing, a person, organisation, place or product, that a search engine or AI model represents as a distinct concept in its knowledge base, connected to related facts and other entities, rather than as just a string of matching text on a page.
F
- Faceted navigation
- Faceted navigation is the filter and sort system on a catalogue page, such as filtering by size, colour or price, that lets a visitor narrow a list of products.
- Featured snippet
- A featured snippet is a short extract from a web page that Google shows above or near the top of the organic results, answering the query directly in a paragraph, list or table, with a link to the page it came from.
G
- Generative engine optimization (GEO)
- Generative engine optimization (GEO) is the work of making a brand's pages and facts easy for AI answer engines such as ChatGPT, Perplexity, Claude, Gemini and Google's AI Overviews to find, understand and cite, so the brand is named and linked when a buyer asks an assistant for a recommendation or an explanation.
- Google AI Mode
- Google AI Mode is a conversational search experience inside Google Search that answers a question with an AI-generated response, lets the user ask follow-ups, and cites web pages as sources.
- Google AI Overviews
- Google AI Overviews are AI-generated summaries that appear at the top of some Google results pages, answering the query in a few paragraphs and linking to the web pages the summary drew on.
- Google Analytics 4 (GA4)
- Google Analytics 4 (GA4) is Google's current web and app analytics platform.
- Google Business Profile (GBP)
- A Google Business Profile (GBP) is the free listing a business manages in Google that appears in Maps and in the local results on Search.
- Google Search Console (GSC)
- Google Search Console is Google's free tool for monitoring and managing how a site performs in Google Search: which queries and pages get impressions and clicks, which URLs are indexed and why others are not, and technical issues affecting Core Web Vitals, mobile usability and structured data.
- Google Tag Manager (GTM)
- Google Tag Manager is a free tool that lets a site manage tracking and marketing tags, such as GA4, ad conversion pixels and other scripts, through a web interface rather than editing site code directly for every change.
- Google-Extended (robots token)
- Google-Extended is a robots.txt token that lets a site block its content from being used to train future Gemini and Vertex AI generative models, separately from Googlebot.
- GPTBot
- GPTBot is the crawler OpenAI uses to collect publicly available web content to help train and improve its AI models.
- Grounding (AI search)
- Grounding is connecting an AI model's response to real, current external sources, usually live search results, at the moment it answers, rather than relying only on facts it memorised during training.
H
- Headless CMS
- A headless CMS is a content management system that stores and manages content through an API, with no built-in front-end templating of its own, so a separate application, such as a Next.js site, fetches the content and controls entirely how and where it is rendered and displayed.
- Heatmap
- A heatmap is a visual overlay on a page showing where visitors click, move their cursor, or how far down they scroll, aggregated across many sessions and shown as a colour gradient from cool to hot.
- Helpful content system
- The helpful content system is a Google ranking system, now built into its core ranking, that is designed to reduce visibility for content made primarily to rank in search rather than to genuinely help the person reading it, and to reward content that shows real expertise and first-hand knowledge.
- Hreflang
- Hreflang is an HTML or sitemap annotation that tells search engines which language and, optionally, which country a page is written for, and points to the equivalent page in each other language or region a site offers, so the right version is shown to each searcher.
I
- Impressions
- Impressions are the number of times a page appeared in Google's search results for a query, whether or not the searcher scrolled to see it or clicked on it.
- Indexing
- Indexing is the process by which a search engine analyses a crawled page and stores it in its index, the database it actually searches when answering a query.
- IndexNow
- IndexNow is a protocol that lets a website instantly notify participating search engines, currently Bing and Yandex among others, when a page has been added, updated or deleted, instead of waiting for the engine to discover the change on its own next crawl.
- Interaction to Next Paint (INP)
- Interaction to Next Paint is a Core Web Vitals metric that measures how long a page takes to visibly respond after a visitor clicks, taps or types, sampled across every interaction during a visit.
- Internal link
- An internal link is a hyperlink from one page to another page on the same website.
J
- JavaScript rendering
- JavaScript rendering is the process of running a page's JavaScript to build the final HTML a visitor sees, rather than serving that HTML directly from the server.
- JSON-LD
- JSON-LD is the data format Google recommends for adding schema markup to a page: a self-contained block of JavaScript-like code, usually placed in the page head, that states structured facts about the page without needing to be woven into the visible HTML the way older microdata formats did.
K
- Key event (GA4)
- A key event is a GA4 event marked as representing an action particularly important to the business, such as a form submission, a purchase or a booked call.
- Keyword difficulty
- Keyword difficulty is a score, usually 0 to 100, that a keyword research tool assigns to estimate how hard it would be to rank on the first page for a given search term, typically based on the authority and number of backlinks held by the pages currently ranking for it.
- Knowledge Graph
- The Knowledge Graph is Google's database of entities, meaning people, organisations, places and things, together with the facts about them and how they relate.
- Knowledge panel
- A knowledge panel is the information box, usually on the right side of Google's desktop search results, that summarises facts about a recognised entity, such as a company, person or place, pulled from the Knowledge Graph, Google Business Profile and other verified or widely corroborated sources.
L
- Landing page
- A landing page is a standalone page a visitor arrives at from a specific source, such as an ad, an email or a search result, built around a single, focused goal rather than general site navigation.
- Large language model (LLM)
- A large language model is an AI system trained on huge amounts of text to predict and generate human-like language.
- Largest Contentful Paint (LCP)
- Largest Contentful Paint is a Core Web Vitals metric that measures how long it takes for the biggest visible element, usually a hero image, headline or block of text, to render on screen after a visitor opens the page.
- Link equity
- Link equity is the value or authority passed from one page to another through a hyperlink, influenced by the relevance and authority of the linking page and site.
- LLM optimization (LLMO)
- LLM optimization (LLMO) is the effort to influence how large language models describe and recommend a brand, both in answers generated from live web retrieval and in what a model has absorbed from its training data.
- llms.txt file
- An llms.txt file is a proposed plain-text Markdown file placed at a site's root that gives language models a short summary of the site and links to its most useful pages.
- Local pack
- The local pack is the block of usually three local business listings, shown with a map, that Google displays for searches with local intent, such as "plumber near me." It sits above ordinary organic results and is driven mainly by a Google Business Profile rather than the website itself.
- Long-tail keyword
- A long-tail keyword is a longer, more specific search phrase, such as "best waterproof running shoes for flat feet," that gets less search volume individually than a broad term like "running shoes" but is usually easier to rank for and converts better because it matches a clearer intent.
M
- Manual action
- A manual action is a human reviewer at Google's decision to reduce or remove a specific site's visibility because it violates Google's spam policies, such as unnatural links, thin content or cloaking.
- Meta description
- A meta description is an HTML tag that summarises a page's content in a sentence or two.
- Minimum detectable effect (MDE)
- The minimum detectable effect is the smallest improvement in a metric, such as conversion rate, that an A/B test is designed to reliably detect given its sample size and duration.
- Mobile-first indexing
- Mobile-first indexing means Google predominantly uses the mobile version of a page's content, crawled by its smartphone crawler, to index and rank that page, rather than the desktop version.
- Multivariate testing
- Multivariate testing is an experiment that tests multiple page elements, such as a headline, an image and a button colour, in combination at once, measuring not just each element's individual effect but how they interact.
N
- NAP consistency
- NAP consistency is having the exact same business name, address and phone number listed identically across a website, its Google Business Profile, and every directory or citation source it appears on.
- Nofollow
- Nofollow is a link attribute, rel="nofollow", that tells search engines not to pass ranking credit through that link.
- Noindex
- Noindex is a directive, set through a meta robots tag or an X-Robots-Tag HTTP header, that tells a search engine not to include a specific page in its index at all.
O
- OAI-SearchBot
- OAI-SearchBot is the crawler OpenAI uses specifically to build the search index that powers ChatGPT search results and citations.
- Off-page SEO
- Off-page SEO covers the work done outside a site's own pages to improve its search visibility, chiefly earning backlinks, but also digital PR, brand mentions, reviews, and citations.
- On-page SEO
- On-page SEO is everything a site controls directly on its own pages to help them rank: content quality, title tags, headings, internal links, structured data, page speed and mobile usability.
- Organic traffic
- Organic traffic is visits to a website that arrive from unpaid search engine results, as distinct from paid search ads, social media, direct visits or referral links from other sites.
- Orphan page
- An orphan page is a page on a website that has no internal links pointing to it from anywhere else on the site, so a visitor or crawler cannot reach it by browsing normally, only by a direct URL, an external link, or a sitemap entry.
P
- Parametric knowledge
- Parametric knowledge is the factual and linguistic knowledge an AI model has encoded internally in its parameters during training, as opposed to knowledge it retrieves live from an external source at the moment it answers.
- People Also Ask (PAA)
- People Also Ask is a Google search results feature showing an expandable list of related questions, each revealing a short answer pulled from a specific web page when clicked.
- Pillar page
- A pillar page is a broad, comprehensive page covering a core topic at a high level, linked to and from a set of narrower supporting pages that each go deeper into one subtopic.
- Prompt tracking
- Prompt tracking is running a consistent, defined set of realistic buyer questions against AI assistants on a recurring schedule and recording how a brand appears in each answer, whether it is cited, merely mentioned, or absent entirely, and how that compares to competitors named in the same answers.
Q
- Query fan-out (AI Mode)
- Query fan-out is the technique AI search systems such as Google AI Mode use to answer a question: the system breaks the question into related sub-questions, runs many searches for them at the same time, and combines what it retrieves into one answer with citations from several pages.
R
- Rank tracking
- Rank tracking is the practice of regularly recording a site's search ranking position for a defined list of target keywords, using a dedicated tool that performs automated searches, to monitor movement over time.
- Redirect (301 and 302)
- A redirect automatically sends a visitor and a search engine from one URL to another.
- Retrieval-augmented generation (RAG)
- Retrieval-augmented generation (RAG) is a method in which a language model first retrieves relevant documents from a search index or database and then writes its answer using those documents, rather than relying only on what it learned in training.
- Rich results
- Rich results are search results shown with extra visual detail beyond a plain blue link, such as star ratings, product prices, recipe images or event dates, generated when a page's structured data qualifies for one of Google's supported result types.
- robots.txt
- robots.txt is a plain text file at a site's root that tells crawlers which parts of the site they may or may not fetch, by user agent.
S
- Sample ratio mismatch (SRM)
- Sample ratio mismatch is a data-quality problem in an A/B test where the actual split of visitors between variants differs from the intended split, such as a planned fifty-fifty split landing at fifty-six to forty-four, by more than chance would explain.
- Scaled content abuse
- Scaled content abuse is Google's spam policy term for publishing large volumes of pages, whether written by people, automated tools or AI, primarily to manipulate search rankings rather than to genuinely help the reader.
- Schema markup
- Schema markup is structured data added to a web page, usually as JSON-LD, that uses the Schema.org vocabulary to state facts about the page in a form machines can read without guessing: who published it, what product or service it describes, its price, its author and how it relates to other things.
- Search engine optimization (SEO)
- Search engine optimization (SEO) is the work of improving a website so that search engines can crawl, understand and rank its pages for the searches its buyers make.
- Search engine results page (SERP)
- A search engine results page (SERP) is the page a search engine shows in response to a query.
- Search volume
- Search volume is an estimate of how many times a given keyword is searched, usually reported as a monthly average, by a keyword research tool.
- SERP feature
- A SERP feature is any element on a search results page beyond the standard list of blue links, such as a featured snippet, People Also Ask box, local pack, image carousel or AI Overview.
- Server-side rendering (SSR)
- Server-side rendering is generating a page's full HTML on the server for each request, before sending it to the browser, rather than sending a mostly empty page and building the content with JavaScript in the visitor's browser.
- Session recording
- A session recording is a replay of an individual visitor's actual interactions on a page, mouse movement, clicks, scrolling and typing, reconstructed from captured browser events rather than a video recording.
- Shopify
- Shopify is a hosted ecommerce platform used to build and run online stores, providing built-in checkout, payments and hosting.
- Sitemap
- A sitemap is an XML file listing the URLs on a site that a business wants search engines to know about, optionally with metadata like when each was last modified.
- Social proof
- Social proof is evidence that other real people or businesses have used and trust a product or service, shown through testimonials, client logos, review scores, case studies or usage numbers.
- Statistical significance
- Statistical significance is a measure of how likely it is that a difference observed in a test, such as one page version converting better than another, is a real effect rather than random chance.
T
- Test hypothesis
- A test hypothesis is a written, specific statement of what will be changed, why it is expected to help based on evidence, and what metric will show whether it worked, written before an A/B test begins.
- Thin content
- Thin content is a page that adds little or no original value for a reader: very short, auto-generated, scraped from elsewhere, or a near-duplicate of other pages on the site with only small variables swapped.
- Title tag
- A title tag is the HTML element that names a page, shown as the clickable headline in search results and as the browser tab label.
U
- Unlinked brand mention
- An unlinked brand mention is an instance of a business's name appearing on another website, or in an AI assistant's answer, without a hyperlink back to the business's own site.
- UTM parameters
- UTM parameters are tags added to the end of a URL, naming its source, medium and campaign, that tell an analytics platform exactly where a click came from.
V
- Value proposition
- A value proposition is a clear statement of the specific outcome a product or service delivers, for whom, and why it is a better choice than the alternatives.
- Vector embedding
- A vector embedding is a list of numbers that represents the meaning of a piece of text, image or other content, positioned in a mathematical space where content with similar meaning sits close together.
W
- Wikidata
- Wikidata is a free, structured, machine-readable knowledge base run by the Wikimedia Foundation that stores facts about entities, people, organisations, places and products, as linked data.
- WooCommerce
- WooCommerce is a free, open-source ecommerce plugin that turns a WordPress site into an online store.
Y
- YMYL (Your Money or Your Life)
- YMYL is Google's term for topics that could significantly affect a reader's health, financial stability, safety or major life decisions, such as medical, legal or financial content.
Z
- Zero-click search rate
- Zero-click search rate is the share of searches that end without the searcher clicking through to any website, because the answer was shown on the results page itself, in an AI summary, a featured snippet, a knowledge panel or a map listing, or because the searcher refined the query instead.
Questions we get asked
Who writes these definitions?
The two founders, Ibrahem Helal and Jeremy Zava, write and review every entry. Each one names at least one source you can check, usually the search engine's own documentation or the original research, so you are not taking a definition on our word alone.
Why does every entry open with a short paragraph?
Because that paragraph is the answer. It is written to make sense on its own in forty to sixty words, which is also the shape search features and AI assistants quote. The rest of the entry says why the term matters to a buyer and how it is measured or applied.
What is the difference between SEO, GEO and AEO?
SEO gets pages ranked in search results. GEO gets a brand named and cited in answers from assistants such as ChatGPT, Perplexity, Claude and Google's AI features. AEO is the older label for structuring answers so search features can lift them. The work overlaps heavily, and each has its own entry here.
How often are new terms added?
Two a day from 24 September 2026 until the planned list of 120 terms is complete, then as new terms enter common use. Entries are updated when the underlying documentation changes, and the date on each entry says when that last happened.
Can I quote or link to a definition?
Yes. Link to the entry's own page so readers get the source list with it. If a definition is wrong or out of date, email us and we will correct it and change the date on the entry.
Do you sell every service described here?
No. The glossary covers terms buyers meet whoever they hire. Where we do the work, the entry links to the service page; where we do not, it says what the term means and stops there.
Thirty minutes to see where your site is invisible.
Book a free consultation with Ibrahem and Jeremy. We look at your site on search, AI answers and conversion while we talk, and you leave with a prioritised plan.
30 minutes. No pitch deck. You leave with a prioritised plan.

