106 terms, explained plainly

A/B testing compares two versions of a page, headline or element to see which one performs better against a defined goal, such as clicks or conversions.

In SEO work it is most useful for testing title tags, meta descriptions and on-page calls to action before rolling a change out site-wide, so a ranking or conversion drop can be caught early rather than after a full rollout.

An AI agent is software that can plan and carry out multi-step tasks on its own, such as researching a topic, comparing options or completing a booking, rather than just answering a single question.

For SEO this matters because agents increasingly browse and act on behalf of users, which means a site's content and structure need to be usable by a program, not only readable by a human.

An AI citation is a direct mention, quote or link to a specific source that an AI system includes when generating an answer, similar to a footnote.

Citations are the AI-era equivalent of a ranking position. Brands cited regularly across a topic build the kind of visibility a first page Google ranking used to guarantee on its own.

AI content is text, images or video produced wholly or partly by a generative AI model rather than written from scratch by a person.

Google has been clear that AI generated content is not penalised for how it was produced, only for whether it is genuinely useful. The risk is publishing generic, unedited AI output at scale with no first hand expertise behind it.

An AI crawler is a bot, such as GPTBot, ClaudeBot or PerplexityBot, that fetches web pages to train or ground an AI model's answers, separate from Google's traditional search crawler.

Blocking these crawlers in robots.txt keeps a site out of the AI systems people increasingly use to research and buy, so the decision to allow or block them should be deliberate, not a default.

An AI hallucination is confident sounding information generated by an AI model that is factually wrong or has no basis in its source material.

For brands, hallucinations are a visibility risk in themselves. If an AI system cannot find a clear, well sourced statement of a fact about your business, it may generate its own version instead of citing you.

An AI Overview is Google's AI generated summary that appears above the traditional blue link results for many searches, synthesising information from multiple sources.

Being cited inside an AI Overview does not require the top organic ranking. It rewards pages with a clear, quotable answer near the top of the content, which is why structuring pages for extraction has become its own discipline.

AI search visibility is how often and how prominently a brand or page is cited, mentioned or linked to across AI answer engines such as Google AI Overviews, ChatGPT, Perplexity and Copilot.

Unlike a traditional ranking position, it is usually measured by tracking mentions and citations across a set of prompts over time, since there is no single results page to check.

An algorithm update is a change Google makes to how it ranks and evaluates web pages, ranging from small daily tweaks to major named releases like Core Updates.

Sites that rely on manipulative tactics tend to see the sharpest swings after a core update, while sites built around genuine expertise and a clean technical foundation are generally more stable.

Anchor text is the clickable, visible text of a hyperlink, which search engines use as a signal for what the linked page is about.

Over-optimised anchor text, the same exact match keyword used for every link, is one of the clearest patterns Google's spam systems look for, so natural variation matters as much as relevance.

An answer engine is a search tool, such as Google AI Overviews, ChatGPT search or Perplexity, that returns a direct synthesised answer instead of a list of links to click through.

Optimising for answer engines means writing the answer itself in clear, self-contained language, since that is the exact text most likely to be lifted and cited.

An attribution model is the rule an analytics platform uses to decide which marketing touchpoint gets credit for a conversion when a user interacted with several channels before converting.

Organic search is frequently under-credited by last-click models because it often plays an early, research-stage role in the journey rather than the final click. Comparing models matters before cutting an SEO budget on last-click numbers alone.

Average position is the mean ranking a page held in Google's results for a given query or set of queries over a chosen date range, as reported in Search Console.

It is a directional metric rather than a precise one, since it blends very different SERP layouts and personalised results into a single number. It is best read alongside impressions and clicks, not in isolation.

A backlink is a link from another website pointing to yours, which Google has used as a core ranking and trust signal since PageRank.

Quality still matters more than quantity. A handful of links from relevant, genuinely authoritative sites moves the needle more than hundreds from low quality or unrelated domains.

A backlink profile is the full picture of every link pointing to a website, including the number of linking domains, their relevance, their authority and the anchor text used.

A healthy profile grows gradually from a mix of sources. A sudden spike of low quality links, often from a past campaign or negative SEO attempt, is one of the first things worth auditing when rankings drop unexpectedly.

Black hat SEO refers to tactics that violate a search engine's guidelines to gain rankings quickly, such as cloaking, link schemes or keyword stuffing.

These tactics can work briefly but carry real risk. A manual action or algorithmic penalty can remove a site from search results entirely, which is why sustainable strategy avoids them regardless of short term gains.

Bounce rate is the percentage of visits where a user left a site after viewing only one page, without any further interaction recorded.

In GA4 it has been largely replaced by engagement rate, which counts a visit as engaged if it lasts a set time, includes a conversion event or has more than one page view, giving a fuller picture than the old single page bounce metric.

A canonical tag is an HTML element that tells search engines which version of a duplicate or near-duplicate page is the one to index and rank.

It is a hint rather than a strict directive. Google can choose a different canonical if it disagrees, so canonicalisation problems are best fixed at the source, parameters, pagination, filters, rather than patched with tags alone.

Chain of thought is a technique where an AI model works through a problem in explicit intermediate steps before giving a final answer, which tends to improve accuracy on complex questions.

Understanding this matters for content strategy because pages that lay out reasoning clearly, not just conclusions, give a model more structured material to draw on when it constructs its own step by step answer.

Click-through rate is the percentage of people who click a result after seeing it, calculated as clicks divided by impressions.

CTR varies enormously by position and by SERP feature. A result sitting under an AI Overview or a featured snippet typically earns a lower CTR than the same position would without one, which is important context when CTR drops without a ranking change.

A content brief is a document that sets out what a piece of content needs to cover before it is written, including target keywords, search intent, required entities, structure and competitor gaps.

A good brief exists to stop content being written from opinion alone. It should be built from what is actually ranking and being cited for the topic, not a generic template applied to every page.

A content cluster is a group of related articles that all link back to one central pillar page, together covering a topic in depth rather than as isolated posts.

Clusters help both users and search engines understand the full scope of a topic, and the internal linking between cluster pages and the pillar concentrates topical relevance where it matters most.

Content decay is the gradual loss of organic traffic or rankings a page experiences over time, even without any change being made to it.

Causes include competitors publishing fresher or more thorough content, facts going out of date, and shifting search intent. Regular content audits are how decaying pages get caught before they fall out of the results entirely.

Content gap analysis is the process of comparing a site's existing content against competitors and search demand to find topics, subtopics or questions it does not yet cover.

It is most useful when done at the entity and subtopic level, not just the keyword level, since a competitor can rank for the same head term while covering several supporting questions a site has missed entirely.

Content pruning is the practice of removing, merging or noindexing low value, outdated or duplicate pages to improve the overall quality signal of a site.

Search engines evaluate quality at the site level as well as the page level, so a large volume of thin or duplicate pages can quietly drag down how the rest of the site is trusted.

A context window is the amount of text, measured in tokens, that a large language model can consider at once when generating a response.

A larger context window lets a model read an entire long page or several documents in one pass, part of why comprehensive, well structured pages are increasingly favoured over content split thinly across many shallow pages.

Conversational search is a search interaction carried out as a back and forth dialogue, where each query can build on the previous one, rather than a single isolated keyword search.

It rewards content that answers the natural follow-up questions a real conversation would raise, one reason FAQ style structuring and clear subheadings have become more valuable, not less, in the AI search era.

Conversion rate is the percentage of visits or users that complete a defined goal, such as a form submission, phone call or purchase.

Organic traffic volume on its own says little about business impact. Conversion rate is what connects rankings and traffic to actual enquiries or revenue, which is why it belongs in every SEO report, not just clicks and impressions.

Core Web Vitals are Google's three measured page experience metrics, Largest Contentful Paint, Interaction to Next Paint and Cumulative Layout Shift, used as part of how a page's user experience is assessed.

They are a real, if modest, ranking factor, but their bigger impact is usually on conversion rate. A slow or unstable page loses visitors before it ever gets the chance to convert them.

Crawl budget is the number of pages on a site Googlebot is willing and able to crawl within a given period of time.

It mainly matters for large sites, thousands of pages or more, where wasted crawling on low value URLs, filtered pages, thin tags, infinite parameter combinations, can delay new or updated pages being discovered and indexed.

Crawlability is how easily a search engine's bots can find and access a website's pages by following links and reading its technical directives.

A page cannot rank if it cannot be crawled, which makes crawlability the foundation every other SEO effort sits on top of, before content quality or backlinks come into play.

Cumulative Layout Shift measures how much visible content unexpectedly moves around on a page while it is loading, scored from 0 upward, with lower being better.

It is most often caused by images or ads loading without a reserved size, pushing text and buttons around just as a user is about to interact with them.

Digital PR is the practice of earning media coverage and backlinks by pitching original data, expert commentary or newsworthy stories to journalists and publications.

Done well it produces exactly the kind of editorially earned, contextually relevant links that are hardest to replicate through outreach alone, because the coverage exists on its own merit, not because a link was asked for directly.

A dofollow link is a standard hyperlink with no rel attribute blocking it, which passes ranking value between the linking page and the page it points to.

Most links on the web are dofollow by default. The term only exists in contrast to nofollow, sponsored and UGC links, which explicitly tell search engines not to pass that value in the same way.

Domain Authority and Domain Rating are third party metrics, from Moz and Ahrefs respectively, that estimate how likely a website is to rank based on the size and quality of its backlink profile.

Neither is a Google ranking factor. They are useful shorthand for comparing sites during competitor research and link prospecting, but should never be treated as a target to hit for its own sake.

Duplicate content is identical or substantially similar content that appears on more than one URL, either within the same site or across different sites.

It rarely triggers a penalty on its own, but it does force search engines to choose one version to rank, which can dilute signals like backlinks and engagement that would otherwise consolidate on a single page.

E-E-A-T stands for Experience, Expertise, Authoritativeness and Trustworthiness, the framework Google's Search Quality Rater Guidelines use to describe what separates genuinely helpful content from low quality content.

It is not a direct ranking factor that can be measured in a tool, but the guidelines it comes from shape how Google's automated systems are built and evaluated, which is why author credentials, first hand experience and transparent sourcing matter more than they used to.

An entity is a single, well defined thing, such as a person, place, organisation, product or concept, that a search engine can recognise and connect to other known facts about it, independent of the exact words used to describe it.

Google's understanding of the web has moved from matching strings of text to recognising entities and their relationships, which is why consistent naming, structured data and clear factual context matter as much as keyword usage now.

Evergreen content is written to stay accurate and useful over a long period, rather than tied to a specific event, trend or date.

It is a durable source of organic traffic because it keeps earning clicks and links long after publication, though it still needs periodic review to catch details that quietly go out of date.

A featured snippet is a highlighted answer box shown above the standard organic results, pulling a short passage, list or table directly from a ranking page.

Winning a snippet usually means the page already ranks on page one and answers the query in a clean, extractable format close to the top of the content, a structure that also happens to help with AI Overview citations.

Generative AI describes AI systems that create new content, text, images, audio or code, rather than simply classifying or retrieving existing information.

Tools like ChatGPT, Gemini and Claude are generative AI. Understanding what they can and cannot reliably produce is the starting point for deciding how much AI assistance belongs in a content workflow.

Generative Engine Optimisation is the practice of structuring a website and its content so it is more likely to be retrieved, cited and quoted by AI answer engines such as ChatGPT, Perplexity, Google AI Overviews and Copilot.

It builds on traditional SEO foundations, crawlability, authority, clear writing, but adds a specific focus on passage level clarity, structured data and being a genuinely citable source rather than only a ranking one.

Goal tracking is the configuration of specific user actions, such as a form submission or phone click, as measurable conversion events inside an analytics platform.

Without goal tracking set up correctly, an SEO campaign has no reliable way to show which traffic actually turned into leads or sales, which makes it one of the first things worth auditing on a new client account.

Google Analytics 4 is Google's current web analytics platform, built around an event based data model rather than the older session based model of Universal Analytics.

Its event model makes it more flexible for tracking modern user journeys across devices, but it also changed how familiar metrics like bounce rate are calculated, which trips up anyone comparing GA4 data directly against older reports.

A Google Business Profile is the free listing that controls how a business appears in Google Maps and the local pack, including its address, hours, reviews and photos.

It is the single most important asset for local SEO. An incomplete or unclaimed profile means a business is effectively invisible to the map based searches that drive most local, ready to buy demand.

Google Search Console is Google's free tool for monitoring a site's presence in search, showing indexing status, ranking queries, click and impression data, and technical issues Google has detected.

It is the only source of truth for what Google actually sees and knows about a site, which makes it the starting point for diagnosing almost any organic search problem.

Guest posting is writing and publishing an article on someone else's website, usually in exchange for a byline, author bio and often a backlink to your own site.

It can be a legitimate way to build relevance and authority when the target site is genuinely aligned with the topic, but mass guest posting on low quality, pay for placement sites is one of the more common link schemes Google's spam systems target.

A heatmap is a visual report showing where users click, move their cursor or scroll on a page, using colour to represent concentration of activity.

It complements standard analytics by showing not just that a page underperformed, but why, revealing ignored calls to action, confusing layouts or content nobody scrolled far enough to read.

Hreflang is an HTML attribute that tells search engines which language and regional version of a page to show to users in different locations.

Incorrect hreflang implementation is one of the most common causes of the wrong country version of a site ranking for a user, or of pages competing against each other instead of each serving its own market cleanly.

HTTPS is the secure, encrypted version of the HTTP protocol that websites use to transfer data between a browser and a server.

It has been a confirmed, if lightweight, Google ranking signal for years, and browsers now actively flag non-HTTPS sites as not secure, which makes it a baseline requirement rather than an optimisation.

Impressions are the number of times a page appeared in search results for a query, whether or not the user clicked on it.

A page can gain impressions without gaining clicks, which usually points to a mismatch between the title and description shown and what the searcher actually wanted, worth checking before assuming the ranking itself is the problem.

Indexability is whether a search engine is allowed to add a page to its index after crawling it, controlled by directives like the noindex tag and canonical tags.

A page can be perfectly crawlable and still never appear in search results if it carries a noindex tag or canonicalises to a different URL, so the two concepts need checking separately, not assumed to go together.

Interaction to Next Paint measures how quickly a page responds to a user's interaction, such as a click or tap, replacing First Input Delay as a Core Web Vital in 2024.

Unlike its predecessor, INP measures responsiveness across the entire visit, not just the first interaction, which makes JavaScript heavy pages with laggy buttons or menus far more exposed than they were under the old metric.

Internal linking is the practice of linking between pages on the same website, helping both users and search engines navigate a site and understand which pages matter most.

It is one of the few ranking levers a site owner controls entirely. The pages that receive the most internal links, in relevant, well written anchor text, tend to be treated as the most important pages on the site.

JSON-LD is the recommended format for adding structured data to a page, written as a self-contained script block that describes the page's content to search engines in a machine readable way.

Google prefers it over the older microdata and RDFa formats because it can be added without touching the visible HTML, which makes it far easier to implement and maintain correctly at scale.

A keyword is a word or phrase a person types or speaks into a search engine, and the term a piece of content is optimised to rank for.

Modern SEO treats keywords as a starting point for research rather than a target to repeat mechanically on the page, since search engines now evaluate topical coverage and intent match far more than exact phrase matching.

Keyword cannibalisation happens when two or more pages on the same site target the same keyword, causing them to compete against each other instead of one page ranking strongly.

The fix is usually consolidation, merging the weaker page into the stronger one with a redirect, or clearly differentiating each page's intent so search engines and users can tell them apart.

The Knowledge Graph is Google's database of entities and the verified facts and relationships connecting them, used to power features like the Knowledge Panel and to understand context behind a search query.

Getting a business, person or product represented accurately in the Knowledge Graph, through consistent structured data, Wikidata and authoritative third party sources, strengthens how confidently AI systems and Google itself can describe it.

A Knowledge Panel is the information box that appears on the right of Google's search results, summarising key facts about a recognised entity such as a business, person or brand.

It is generated automatically from the Knowledge Graph rather than purchased or directly editable, which makes structured data and consistent information across the web the main lever for improving what it shows.

A large language model is an AI system trained on vast amounts of text to predict and generate human-like language, forming the basis of tools like ChatGPT, Gemini and Claude.

LLMs do not browse the live web by default. They generate answers from patterns learned during training, which is why techniques like retrieval augmented generation exist to connect them to current, verifiable information.

Largest Contentful Paint measures how long it takes for the largest visible element on a page, usually a hero image or heading, to fully render.

Google recommends an LCP under 2.5 seconds. Slow LCP is most often caused by unoptimised images, render blocking scripts or a slow server response, all fixable without a full rebuild.

Link building is the deliberate process of earning or acquiring backlinks from other websites to increase a site's authority and rankings.

The most durable approach earns links because content, data or commentary is genuinely worth linking to. Approaches built purely around volume or paid placement tend to age badly as search engines get better at detecting them.

Grounding is the process of connecting an AI model's answer to verifiable, up to date external sources, rather than relying only on patterns learned during training.

Search-connected tools like Google AI Overviews, Perplexity and ChatGPT search ground their answers in live web results, exactly the mechanism that makes a page's clarity and citability a visibility factor, not just its content quality.

llms.txt is a proposed plain text file, placed at a site's root, that gives AI models a curated summary of a site's most important pages, written specifically for machine consumption rather than for browsers.

It is not yet a confirmed standard the major AI platforms officially consume, but adoption is growing, and publishing one costs little while potentially helping an AI system understand a site's structure faster than crawling alone would.

The local pack is the block of typically three local business listings, shown with a map, that appears for searches with local intent, such as 'plumber near me'.

Ranking in it depends heavily on Google Business Profile completeness, review volume and quality, and proximity to the searcher, factors quite different from the ones that drive standard organic rankings.

Local SEO is the practice of optimising a business's online presence to attract more customers from relevant local searches, typically centred on Google Business Profile and localised content.

It matters most for businesses serving a physical area or service radius, dentists, tradespeople, clinics, where 'near me' and location plus service searches are often the highest intent traffic a site can capture.

Log file analysis is the review of a website's raw server logs to see exactly how search engine bots are crawling a site, which pages they visit, how often, and which they ignore.

It is the most accurate crawl data available, since it comes directly from the server rather than a simulated crawl, especially valuable for diagnosing crawl budget waste on large sites.

A long-tail keyword is a longer, more specific search phrase that has lower individual search volume but typically higher intent and less competition than a short, broad keyword.

Targeting a wide net of long-tail terms is often a faster path to traffic and conversions for newer or smaller sites than competing directly for a handful of highly competitive head terms.

Mobile-first indexing means Google predominantly uses the mobile version of a page's content for indexing and ranking, even for searches performed on desktop.

Any content, links or structured data that exists on desktop but is missing or hidden on mobile is effectively invisible to Google, which makes checking parity between the two versions a basic technical audit step.

Multimodal AI refers to models that can understand and generate more than one type of content, such as text, images, audio and video, within a single system.

As multimodal models become the norm, image alt text, video transcripts and clear captions carry more weight than before, since they give a model additional material to interpret alongside the written page.

Natural language processing is the branch of AI focused on enabling computers to understand, interpret and generate human language.

It underpins how Google interprets search queries beyond exact keyword matching, and how content is assessed for topics, entities and sentiment, which is why writing naturally for a human reader remains the safest long term strategy.

A nofollow link carries a rel attribute telling search engines not to pass ranking credit to the linked page, though Google now treats it as a hint rather than a strict rule.

Common sources include comments, forum posts and paid placements. A natural backlink profile always includes a mix of nofollow and dofollow links, so their presence is not itself a problem.

Off-page SEO covers everything done away from a website itself to improve its rankings, primarily backlink building, brand mentions and digital PR.

It works alongside on-page and technical SEO rather than replacing either. Authority earned off-site has little effect if the pages it points to are not crawlable, fast and genuinely useful.

On-page SEO is the optimisation of individual page elements, including title tags, headings, content and internal links, to help a page rank for its target queries and satisfy the searcher.

It sits between technical SEO, which controls whether a page can be found at all, and off-page SEO, which builds authority pointing to it, making all three necessary rather than a choice between them.

Organic search refers to unpaid search engine results, and the traffic that arrives at a website by clicking one of those results rather than a paid advertisement.

It is typically the highest volume, highest trust traffic channel a business has access to, since users generally trust an earned ranking more than a labelled ad.

Organic traffic is the number of visits a website receives from unpaid search engine results, as opposed to paid ads, social or direct visits.

It is the metric most directly tied to SEO effort, but it should always be read alongside conversion rate. A rise in organic traffic that does not translate into leads or sales points to a targeting or landing page problem, not an SEO win.

Page speed is how quickly a page's content loads and becomes usable for a visitor, measured through metrics like Largest Contentful Paint and Time to First Byte.

Beyond its role in Core Web Vitals, speed has a direct, well documented effect on conversion rate. Even a one second delay measurably increases the chance a visitor leaves before the page finishes loading.

Passage-level indexing is Google's ability to identify and rank a specific, self-contained section of a page for a query, even if that passage is not reflected in the page's title or main topic.

It rewards clear, well structured subsections with their own heading and complete answer, since a passage can now earn visibility on its own merit rather than needing the whole page to be about that exact query.

A pillar page is a comprehensive page covering a broad topic at a high level, linking out to more detailed cluster pages that each cover a specific subtopic in depth.

Together with its cluster, a pillar page is how a site demonstrates topical authority to both users and search engines, rather than trying to cram every detail of a subject onto a single page.

Prompt engineering is the practice of writing and refining the instructions given to an AI model to reliably get a more accurate, relevant or well formatted response.

For SEO teams it matters both for using AI tools effectively in research and drafting, and for understanding how real users phrase prompts to answer engines, since those phrasings reveal the search intent behind a topic.

Prompt-to-page match is how closely a page's content aligns with the actual language and structure of prompts users type into AI answer engines, as opposed to the shorter, keyword style queries typed into traditional search.

AI prompts tend to be longer, more conversational and more specific than search keywords, so pages written to answer full questions directly tend to match better than pages built around a single short keyword phrase.

Quotable content is writing structured so a complete, accurate answer can be lifted as a standalone passage, without needing the rest of the page for context.

It generally means putting the direct answer in the first sentence of a section, then supporting it with detail afterward, the opposite structure of building up to a conclusion, which AI systems and featured snippets both struggle to extract from.

Readability is how easily a piece of writing can be read and understood, influenced by sentence length, word choice, structure and formatting.

Search engines do not directly score readability, but users do. Content that is hard to scan gets abandoned faster, which shows up indirectly through weaker engagement and conversion metrics even when rankings look fine.

Retrieval-augmented generation is a technique where an AI model retrieves relevant information from an external source, such as live search results or a document database, before generating its answer.

It is what allows tools like Google AI Overviews and Perplexity to cite current, specific web pages rather than relying only on what the underlying model learned during training, exactly the mechanism GEO work is trying to influence.

Robots.txt is a plain text file at a site's root that tells search engine and AI crawlers which parts of the site they are allowed or disallowed to crawl.

It controls crawling, not indexing. A disallowed page can sometimes still appear in results if it is linked to elsewhere, a common point of confusion when trying to fully remove a page from search.

Salience is how central or prominent a given entity or topic is within a piece of content, as distinct from simply being mentioned in passing.

Google's NLP systems assess salience, not just presence, when judging what a page is really about, which is why burying the main subject under generic filler weakens topical relevance even if the right keyword appears on the page.

Search Engine Optimisation is the practice of improving a website's visibility in unpaid search results by aligning its technical foundation, content and authority with how search engines evaluate and rank pages.

It spans three connected disciplines, technical SEO, on-page SEO and off-page SEO, and increasingly extends into GEO, since search itself now includes AI-generated answers alongside the traditional results page.

Search intent is the underlying goal behind a search query, generally categorised as informational, navigational, commercial or transactional.

Matching a page's format and depth to the correct intent matters more than exact keyword matching. A commercial intent query needs a page that helps someone decide, not an academic explainer, no matter how well the keyword is used.

Search volume is the estimated number of times a keyword is searched for over a given period, usually shown as a monthly average.

It is a useful prioritisation signal but a poor standalone target, since a lower volume keyword with strong commercial intent can drive more actual business value than a high volume, purely informational one.

Semantic search is a search engine's ability to understand the meaning and context behind a query, rather than matching only the literal words typed.

It is why a page can rank for a related phrase it never explicitly contains, and why writing to genuinely cover a topic outperforms writing to hit a list of exact match keyword variations.

Semantic SEO is an approach to content strategy built around covering a topic's full range of related entities, attributes and subtopics, rather than optimising individual pages for isolated keywords.

It treats a site's content as a connected knowledge structure that demonstrates depth on a subject, aligning closely with how search engines now assess topical authority and relevance.

A SERP is the page a search engine returns after a query, made up of organic results alongside features like ads, the local pack, featured snippets and AI Overviews.

Modern SERPs increasingly push standard blue-link organic results further down the page, a large part of why ranking position alone no longer guarantees the visibility or clicks it once did.

A SERP feature is any non-standard result shown on a search results page, such as a featured snippet, local pack, image carousel, People Also Ask box or AI Overview.

Each feature has its own separate opportunity to appear in, often independent of the standard organic ranking, which is why a full SEO strategy considers each relevant SERP feature for a query, not just the ten blue links.

Server-side rendering generates a page's full HTML on the server before sending it to the browser, rather than relying on JavaScript in the browser to build the page after it loads.

It matters for SEO because it guarantees crawlers see complete content immediately, without needing to execute JavaScript first, reducing the risk of important content being missed or indexed late.

Share of voice in AI search is the proportion of citations or mentions a brand receives across a defined set of AI-generated answers, relative to its competitors, for a given topic or set of prompts.

It is becoming the AI-era equivalent of tracking keyword rankings against competitors, typically measured by running the same prompts repeatedly across platforms and logging which sources get cited.

Source attribution is whether and how clearly an AI system credits the original source it drew information from when generating an answer.

Clear authorship, publish dates, and well structured claims all make attribution more likely, since an AI system is more confident citing a source it can clearly identify and verify as the origin of a fact.

Structured data is standardised markup, most commonly written as JSON-LD, that describes a page's content in a format search engines and AI systems can read and interpret directly.

It does not guarantee better rankings on its own, but it does make it far easier for a search engine to accurately understand and represent a page, whether through a rich result, a Knowledge Panel or an AI-generated answer.

Thin content is a page with little substantive information, offering minimal unique value to a reader relative to what else already exists on the topic.

It is one of the clearest patterns Google's quality systems target, whether it comes from unedited AI output, doorway pages, or a large batch of near-identical location pages published with only the city name swapped.

A token is the basic unit of text an AI language model processes, roughly equivalent to a word or part of a word, used to measure both a model's input limits and its cost.

Understanding tokens helps explain practical AI limitations, such as why a very long page might get truncated or summarised rather than read in full within a given context window.

Topical authority is the degree to which a website is regarded, by search engines and users alike, as a comprehensive and trustworthy source on a given subject.

It is built through consistent, interconnected coverage of a topic over time, a pillar and cluster structure, internal linking and genuine expertise, rather than through any single standout page.

Trust signals are the various cues, such as reviews, secure connections, author credentials, verifiable contact details and third party mentions, that indicate to both users and search engines that a site is credible.

No single trust signal carries much weight alone, but together they form a large part of how E-E-A-T is assessed in practice, which is why they need to be built deliberately rather than assumed to exist by default.

A vector embedding is a numerical representation of a word, sentence or piece of content that captures its meaning, allowing an AI system to measure how semantically similar two pieces of text are.

Embeddings are what let a search or AI system match a query to relevant content even when no exact words overlap, the mathematical foundation behind modern semantic search.

White hat SEO refers to optimisation practices that fully comply with search engine guidelines, focused on genuinely improving a site's quality, relevance and user experience.

It is a slower path to results than manipulative tactics, but it is the only approach that holds up over time, since it does not carry the risk of a penalty erasing years of progress overnight.

An XML sitemap is a file listing a website's important URLs, submitted to search engines to help them discover and understand a site's structure.

It is a discovery aid, not a guarantee of indexing. A URL still needs to meet a search engine's quality bar to be indexed regardless of whether it appears in the sitemap.

A zero-click search is a query that gets fully answered directly on the search results page, through a featured snippet, AI Overview or knowledge panel, without the user clicking through to any website.

Zero-click searches now make up a large share of all queries, shifting SEO value away from pure click volume and toward being the cited, trusted source behind the answer itself, whether or not that earns a visit.

A 301 redirect is a permanent redirect that sends both users and search engines from an old URL to a new one, passing the vast majority of the original page's ranking signals along with it.

It is the correct way to handle a URL change, whether from a site migration, a rebrand or content consolidation. Skipping it, or using a temporary 302 redirect instead, is one of the most common causes of lost rankings after a site change.