Technical Glossary

A formal written announcement that publicises newsworthy events, product launches, company milestones or important updates. Press releases follow a standard format (headline, dateline, body, boilerplate and contact information). Today they serve three audiences at once: journalists, search engines, and the AI assistants that answer questions by quoting published sources.

A press release written and structured so AI answer engines can understand, trust and quote it: a clear factual lead, specific numbers and quotes, an accurate company boilerplate, and machine-readable Schema.org markup (NewsArticle, Organization, FAQPage). It may be drafted with AI, but the defining feature is that it is built to be cited in AI answers, not only read by journalists.

A press release structured for autonomous AI agents (systems that research, compare and act without a human prompting each step) as well as for people. That means consistent entity data, machine-readable endpoints (structured data, feeds, llms.txt) and facts an agent can verify, so the announcement can be discovered and used inside agent-driven research and commerce.

The standard "About the company" paragraph at the end of every press release, providing consistent company background. Boilerplates typically include founding date, mission, key products or services, notable achievements and location. Because AI systems often lift this paragraph when asked "Who is this company?", an accurate, specific boilerplate is one of the most reused parts of a release.

The process of getting a press release in front of its audiences: newswires and syndication networks, direct journalist outreach, publication on indexed pages, and feeds and notifications that search engines and AI crawlers pick up. Effective distribution maximises visibility for both human readers and machine indexers.

A "do not publish before [date/time]" restriction placed on a press release to give journalists advance notice while controlling publication timing. Embargoes give reporters time to research and prepare coverage while keeping the announcement coordinated across outlets. Breaking an embargo damages journalist relationships and future coverage.

A collection of company assets provided to support press coverage, including high-resolution logos, product images, executive headshots, fact sheets, a company timeline and previous press releases. Media kits streamline journalist workflows by providing ready-to-use, publication-quality materials.

The compelling angle or timely element that makes a press release newsworthy. Strong news hooks tie an announcement to trending topics, industry challenges, customer pain points or current events. Without a clear hook, press releases get ignored by media and readers.

A traditional press release distribution network that syndicates announcements to media outlets, news aggregators and journalists. Major newswires include PR Newswire, Business Wire and GlobeNewswire. Pricing varies widely: as of 2026, PR Newswire's published list pricing is roughly $350 (local) to $805 (national) per 400-word release plus a $195 annual membership, and add-ons can push a single release past $3,000. Newswires are strong on syndication reach; they were not built around tracking whether AI answer engines cite the release.

A personalised email or message sent to a journalist proposing a story angle or offering exclusive access to company news. Pitches are shorter and more conversational than press releases, tailored to the individual reporter's beat. Effective pitches explain why the story matters to that journalist's audience.

Descriptive text added to an image's HTML that describes its content for screen readers (accessibility) and search engines. Alt text lets visually impaired users understand images and helps search engines index them. Well-written alt text improves both user experience and image search visibility.

Links from external websites pointing to your content. Search engines treat editorial backlinks as votes of confidence, and quality matters more than quantity: one link from a respected publication outweighs hundreds from unknown sites. Links inside press releases are normally marked nofollow or sponsored and do not pass ranking value; the backlink value of PR comes when journalists and publishers pick up the story and link to you editorially.

The preferred version of a page URL, declared with a canonical tag, used to consolidate duplicate or near-duplicate pages. When the same content exists at several URLs, the canonical tells search engines which version to index and rank, so your own pages don't compete against each other.

Google's page-experience metrics: Largest Contentful Paint (LCP, loading), Interaction to Next Paint (INP, responsiveness, which replaced First Input Delay in March 2024) and Cumulative Layout Shift (CLS, visual stability). They became part of Google's ranking signals in 2021. Fast, responsive, stable pages give users a better experience and can edge ahead of slower competitors.

The process by which search engine bots (like Googlebot) discover and fetch web pages by following links and sitemaps. Crawling precedes indexing: if a page isn't crawled, it can't be indexed or ranked. robots.txt and XML sitemaps help guide crawling.

Informal name for an ordinary link with no rel qualifier, which lets search engines pass ranking signals to the destination. That is right for editorial links. For paid or sponsored placements, including links in press releases, Google's spam policies ask for rel="sponsored" (or rel="nofollow"); a followed paid link can lead to a manual action. See Sponsored Links.

Links marked rel="sponsored", the attribute Google asks sites to use for paid or sponsored placements (rel="nofollow" is also accepted). Their value is transparent, guideline-compliant attribution, referral traffic and a clear association between the page and the linked company. Every Pressonify release links to the customer's own site this way.

A 0-100 score created by Moz to estimate how well a website might rank, based mainly on its backlink profile. Similar third-party scores include Ahrefs' Domain Rating. Google does not use DA; it is a comparative indicator, useful for sizing up sites, not a ranking factor in itself.

Selected search results shown at the top of Google in a special box (position zero), often answering a question directly from page content. Featured snippets increase visibility, and voice assistants frequently read them aloud.

Content that Google has crawled, analysed and stored in its search index, making it eligible to appear in results. Unindexed content cannot appear in search results regardless of quality. Google Search Console shows indexing status.

The words and phrases people type into search engines. Effective SEO means identifying relevant keywords with real demand and working them naturally into content, headlines and metadata. Keyword research reveals what your audience actually searches for, not what you assume they search for.

The ranking value passed from one page to another through links, also called PageRank, link equity or "link juice". Links from highly trusted sites pass more value, which is why a link from a major news site matters more than one from an unknown blog. Links marked nofollow or sponsored do not pass it.

The short summary (roughly 150-160 characters) that can appear below a page title in search results. It is not a direct ranking factor, but a compelling meta description improves click-through rate by telling searchers what they'll find. Think of it as your page's pitch in the results.

Meta tags that control how content appears when shared on social platforms such as Facebook and LinkedIn. Open Graph tags specify the title, description, image and URL shown in the preview. Without them, shares show generic previews that get less engagement.

A permanent redirect from one URL to another. Use 301s when moving pages, changing URL structure or consolidating duplicate content. Google has said 301 redirects no longer lose PageRank, so a properly implemented redirect preserves a page's SEO value through a migration.

Enhanced search results that show extra information such as star ratings, images, dates or FAQs. Rich results improve click-through rates by adding context and visual appeal. They depend on valid Schema.org markup, and Google decides case by case whether to show them.

A text file at a site's root (yoursite.com/robots.txt) that tells crawlers which paths they may or may not fetch. It helps manage crawling of admin areas, duplicate content or staging sections. Note that robots.txt controls crawling, not indexing: use a noindex directive to keep a page out of search results.

Content structured to rank well in search results through relevant keywords, meta tags, structured data, fast loading, mobile responsiveness and genuine authority signals. SEO combines on-page elements (content, HTML) with off-page factors (links, reputation).

An XML file listing a site's important URLs so search engines can discover and crawl them efficiently. The most useful field is an accurate lastmod date; Google ignores the priority and changefreq fields. Submitting a sitemap in Google Search Console helps new content get found faster.

Meta tags that control how links appear when shared on X (Twitter), similar to Open Graph tags. Cards enable rich previews with images, video and article summaries. Without them, links can show as plain text.

The practice of optimising content to provide direct, concise answers to specific questions, targeting AI assistants, voice assistants and featured snippets. AEO puts a clear answer in the first two or three sentences, uses FAQ sections with Schema.org FAQPage markup, and builds answer-focused content blocks. While SEO aims to "rank on page one", AEO aims to "be the answer".

The practice of optimising content to be cited as a source in AI-generated answers. When ChatGPT, Perplexity or Google AI Overviews generate an answer, they usually cite only a handful of sources: GEO is about becoming one of them. That takes authoritative, fact-dense content with clear attribution, specific statistics, verifiable facts and Schema.org structured data.

The practice of making sure large language models interpret and describe your content accurately. LLMs don't "read" like humans: they tokenise text and draw meaning from context and learned patterns. LLMO uses semantic HTML, clear and unambiguous language, explicit definitions and examples, a logical information hierarchy and consistent Schema.org vocabulary. Poor LLMO leads to misinterpretation; good LLMO means AI describes your company correctly.

A framework for content optimisation in the AI era with five complementary layers: SEO (search engine ranking), AEO (direct answers), GEO (AI citations), LLMO (AI interpretation) and ADP (AI discoverability). Each layer addresses a different part of how content gets discovered, interpreted and cited. Pressonify applies all five to every release it publishes: structured data, FAQ answers, llms.txt and ADP endpoints.

Google's AI-generated summaries at the top of search results, which synthesise information from multiple sources to answer a query directly and link to the sources used. Originally tested as the Search Generative Experience (SGE), AI Overviews mark Google's shift from lists of links towards generated answers.

A search where the user gets an answer directly on the results page without clicking through to any website. Featured snippets, knowledge panels and AI Overviews all contribute. This reduces website traffic, but being the cited source inside the answer still delivers brand visibility and authority. AEO specifically targets zero-click visibility.

Google's framework for evaluating content quality, extended in December 2022 to add "Experience". Experience means first-hand knowledge of the topic; Expertise means demonstrated subject knowledge; Authoritativeness means recognised reputation in the field; Trust is the overall judgement of credibility and matters most. For press releases, E-E-A-T signals include domain verification, named spokespeople, sourced statistics and consistent company information.

Optimising content for spoken queries through voice assistants like Alexa, Siri and Google Assistant. Voice searches tend to be conversational, question-based and often local ("Hey Siri, what's the best plant delivery near me?"). Optimisation means natural language, long-tail question keywords, Speakable Schema where supported, and answers written in two or three sentences that can be read aloud.

The practice of structuring content to win featured snippets (position zero). It involves question-based headings, a direct 40-60 word answer immediately after the question, lists and tables for comparisons, and a clean heading hierarchy. Featured snippets are often what voice assistants read aloud, and the same answer-first structure makes content easier for AI answer engines to quote.

The shift in which being cited as a source inside AI answers (ChatGPT, Perplexity, Claude, Gemini, Google AI Overviews) becomes a primary driver of brand visibility alongside traditional search rankings. As more questions are answered directly by AI, the goal moves from ranking on page one to being the source the answer is built from.

Whether, and how often, your company or your content appears inside AI-generated answers to the questions your customers ask. Answer presence is measured by asking AI engines those questions and recording which sources they cite and which brands they name. It is a different measure from website traffic: an AI answer can name and cite you even when nobody clicks through.

The change in how online visibility is measured. Clicks count visits to your site; citations count how often AI answers reference you as a source. As zero-click and AI answers grow, a brand can gain visibility (and trust) through citations even when clicks fall, so both need tracking.

A press release system that follows the whole citation lifecycle: Publish, Index, Cite, Detect. Traditional PR platforms mostly report distribution (how many outlets received the release); a closed-loop system also checks whether AI engines actually cite it. Pressonify asks AI answer engines (Perplexity for every release, plus ChatGPT and Gemini on Premium and Enterprise) questions about the customer's company and records when a release is cited, giving concrete evidence of answer presence.

Shorthand for a finding in Muck Rack's "Generative Pulse" research, which analysed more than a million links cited by AI engines and found that over 95% came from non-paid media (earned and owned sources such as journalism, company newsrooms and press releases) rather than paid placements. It is why PR and earned media are such a strong route into AI answers. Note the finding covers non-paid media broadly, not press releases alone.

A predictive 0-100 score estimating how likely content is to be cited by AI systems. Pressonify's scorer weights five components: Information Gain 35% (unique facts, statistics, insights), Structure 25% (how easily the body can be extracted and quoted), Freshness 15%, Competitive Context 15% (industry positioning and decision hooks) and Authority 10% (schema richness and verification signals). Any component below 70 comes with specific recommendations.

The unique value content adds beyond what is already available elsewhere. AI systems favour novel facts, original statistics, first-hand perspectives and primary sources. Content that merely summarises existing information has low information gain; content with exclusive data, expert quotes or original research has high information gain and is more worth citing. It carries the largest weight (35%) in Pressonify's Citability Score.

A content strategy that optimises for traditional search (Google, Bing) and AI citation systems (ChatGPT, Perplexity, Claude, Gemini) at the same time. Track 1 covers SEO fundamentals: rankings, organic traffic, links, Core Web Vitals. Track 2 covers AI citation: structured data, freshness, entity consistency and answer-first writing. Much of the work (Schema.org, authority signals, clear facts) benefits both tracks.

A four-layer framework for what drives AI citations. Layer 1 (Authority): are you crawled often? Harmonic Centrality in Common Crawl's web graph is one measure, and well-connected sites appear more in AI training data. Layer 2 (Distribution): is your content on sites AI already trusts? Coverage on established news sites borrows their authority. Layer 3 (Discovery): can AI understand you? Schema.org, ADP endpoints and llms.txt live here. Layer 4 (Citation): did AI actually cite you? This is the outcome that matters. Many companies over-invest in Layer 3 and neglect Layers 1-2, which is why schema alone doesn't guarantee citations.

The gap between how much companies use AI internally and how visible they are to AI externally. McKinsey's State of AI 2025 survey found 78% of organisations use AI in at least one business function, yet most had not yet seen enterprise-level financial impact. Many of those same companies have done little to make their own content discoverable to, and citable by, AI systems: using AI (chatbots, automation) is a different job from being cited by it.

How discoverable, retrievable and citable your content is for AI-powered search systems. Unlike traditional rankings, AI search visibility depends on structured data quality, freshness, entity recognition and citation likelihood. High AI search visibility means your content appears as a cited source when people ask AI systems questions in your field. It is measured through AI crawler activity, citation tracking and AI referral analytics.

The observation that non-paid, news-style content (journalism, company newsrooms and press releases) makes up the bulk of what AI engines cite: Muck Rack's Generative Pulse research put non-paid media at over 95% of cited links. Press releases suit AI citation because they are structured, timestamped, fact-dense and published on crawlable pages, which makes AI-optimised releases one of the most efficient routes to AI visibility.

"Alternative metrics": measures of attention and impact beyond formal citations, such as social mentions, shares, downloads, bookmarks and news coverage. Borrowed from academic publishing, altmetrics help show how widely an announcement travels, and these broader signals feed the reputation that AI systems weigh when choosing sources.

An open specification for making a website structurally discoverable by AI systems such as ChatGPT, Claude, Perplexity and Gemini. Where traditional SEO is built around keyword search, ADP publishes structured, machine-readable metadata for AI crawlers. Its core files are /ai-discovery.json (the entry-point index), /knowledge-graph.json (entity catalogue) and /llms.txt (AI-readable context), plus feeds and change logs so crawlers can pick up new content quickly.

The entry-point file for the AI Discovery Protocol, served at /ai-discovery.json. This index lists a site's AI-oriented resources (knowledge graph, llms.txt, feeds and other discovery endpoints), so an AI system can read one file and find everything else instead of reconstructing entities from hundreds of HTML pages.

A JSON-LD file containing a website's entity catalogue in Schema.org vocabulary. Served at /knowledge-graph.json, it gives AI systems the site's entities (organisations, articles, products, people, defined terms) and their relationships in one structured file, rather than scattered across individual pages.

A markdown file at /llms.txt that gives AI language models a concise, readable overview of a website: its purpose, key pages and most useful content. Proposed by Jeremy Howard in 2024, it works like an "about page" written for machines rather than humans. Pressonify adds every published release to its llms.txt.

The state of a website implementing the AI Discovery Protocol specification. ADP compliance has three levels: Level 1 (Minimal) needs only ai-discovery.json (about 15 minutes to implement); Level 2 (Standard) adds knowledge-graph.json and llms.txt (2-4 hours); Level 3 (Advanced) adds versioning and change logs for incremental crawling (1-2 days). Pressonify implements Level 3 for all published content.

An HTTP response header indicating which version of the AI Discovery Protocol a server implements. For example, X-ADP-Version: 3.0 signals ADP v3.0 to crawlers and tools, so they know which endpoints and response formats to expect.

An HTTP response header (X-LLM-Optimized: true) that a site uses to declare that a response has been prepared for large language model consumption, with structured data, semantic markup and machine-readable metadata. It is a self-declared signal: no major AI crawler documents acting on it, so treat it as a label, not a ranking lever.

A protocol that lets websites notify participating search engines (Bing, Yandex, Seznam, Naver and others) the moment content is created, updated or deleted, instead of waiting for crawlers to find the change. Google does not participate. Pressonify pings IndexNow automatically when a press release is published.

A W3C protocol (formerly PubSubHubbub) that pushes feed updates to subscribers through a hub the moment new content is published, instead of subscribers polling the feed on a schedule. Pressonify pings a WebSub hub for its RSS and JSON feeds whenever a release goes live.

Configuring robots.txt to control access by AI crawlers and AI-related user agents: GPTBot (OpenAI training), OAI-SearchBot (ChatGPT search), ChatGPT-User (pages fetched for a user), ClaudeBot (Anthropic), PerplexityBot, and the Google-Extended and Applebot-Extended tokens that govern use of content for Gemini and Apple Intelligence. Sites that want AI visibility explicitly allow these agents, for example User-agent: GPTBot followed by Allow: /.

Automated bots run by AI companies to fetch web content for model training, search indexes and real-time answers. Major ones include GPTBot, OAI-SearchBot and ChatGPT-User (OpenAI), ClaudeBot (Anthropic), PerplexityBot (Perplexity) and Meta-ExternalAgent (Meta). Google and Apple crawl with Googlebot and Applebot and use the Google-Extended and Applebot-Extended robots.txt tokens to let sites opt in or out of AI use.

A check of whether AI crawlers can actually reach and understand a site: robots.txt rules for each AI user agent, firewall or CDN bot blocking, llms.txt and sitemap availability, server response codes, and whether key pages render without JavaScript. It often reveals sites that block the very AI systems they want to be cited by.

A lightweight format for publishing a site's knowledge as linked markdown documents with YAML frontmatter, one per concept, so AI agents can read it as plain files. Pressonify publishes its press releases, companies, topics and blog articles in this format, starting from /okf/index.md.

A network measure published in Common Crawl's web graph statistics that captures how "close" a domain is to all other domains in the link graph. Common Crawl uses it to help prioritise crawling, so well-connected domains (Wikipedia, Reddit, major news sites) are crawled more and appear more often in the web corpora many AI models are trained on.

The number of URLs a crawler (Googlebot, Common Crawl or an AI crawler) is willing and able to fetch from your site in a given period. It is shaped by how fast and reliable your server is and how valuable the crawler judges your pages to be. Slow responses, errors, redirect chains and large volumes of low-value URLs waste it, leaving important pages unfetched: a missed opportunity for both search and AI visibility.

A system for tracking and measuring AI crawler engagement with website content. It identifies when AI bots (GPTBot, ClaudeBot, PerplexityBot and others) visit pages, logs their activity and calculates visibility scores, showing how discoverable your content is to AI systems: the AI counterpart to web analytics for human visitors.

A 0-100 metric in Pressonify that summarises how much attention AI crawlers are paying to a specific press release. It combines the number of distinct AI bots that visited (30%), total bot visits (25%), recency of the last visit (20%), how quickly bots discovered the page after publication (15%) and visits from priority AI crawlers such as GPTBot, PerplexityBot, ClaudeBot and Google-Extended (10%). It measures crawler attention, the first step on the path to an AI citation.

A visit that arrives at your website from an AI assistant's answer. When ChatGPT, Claude, Perplexity or Gemini cite your content and a user clicks the link, the referrer (for example chatgpt.com or perplexity.ai) identifies the AI source. LLM referral tracking shows how much traffic AI citations actually drive.

Server-side middleware that inspects incoming requests and identifies AI crawler visits from their User-Agent strings. It distinguishes GPTBot (OpenAI), ClaudeBot (Anthropic), PerplexityBot, Googlebot, Bingbot and other crawlers, logging each visit with timestamp, path and crawler type for analytics.

A single automated bot operated by an AI company, identified by its User-Agent string (for example GPTBot or ClaudeBot). Each can be allowed or blocked individually in robots.txt. See AI Crawlers for the major ones and what each is used for.

An open protocol introduced by Anthropic that lets AI assistants connect securely to external data sources and tools through a standard interface. MCP servers expose tools (actions) and resources (data) that assistants such as Claude can use. Pressonify runs an MCP server, so you can draft press releases and check citation reports directly from an AI assistant.

Combining outbound discovery (ADP) with inbound measurement (crawler tracking and citation checks) to create a feedback loop: publish content, make it discoverable, see which AI systems visit, check whether they cite it, improve the content based on the data, and publish again. The loop replaces guesswork with evidence.

Pressonify's AI discovery architecture, made up of seven layers: (1) Schema.org JSON-LD structured data, (2) knowledge graph entity relationships, (3) llms.txt context, (4) the ai-discovery.json index, (5) instant notification via IndexNow and WebSub, (6) AI crawler permissions in robots.txt, and (7) AI crawler and citation tracking. Together they make each release easy for AI systems to find and understand, and make the result measurable.

AI systems that can plan, make decisions and take actions towards a goal with limited human intervention. Unlike passive tools that wait for each command, agentic AI breaks a task into steps, uses tools, checks results and adapts: an assistant that doesn't just answer questions but works through problems.

Search powered by AI that interprets context, intent and natural language to give direct answers rather than just lists of links. AI search products such as Perplexity, ChatGPT search and Google's AI Overviews synthesise information from multiple sources and cite them, which means content has to be structured to be cited, not only to be read.

Checks built around a language model to stop it presenting false or invented information: validating claims against the source material, detecting inconsistencies and flagging low-confidence output for review before publication.

When an AI system like ChatGPT, Claude or Perplexity references and links to your content as a source for its answer. AI citations are becoming as important as backlinks for visibility, because a cited source is shown to everyone who asks that question. Structured data helps AI systems parse and attribute content, but on its own it doesn't guarantee a citation: clear, specific, verifiable facts matter most.

A distinct, well-defined "thing" in a knowledge graph, such as a person, place, company, product or concept. Entities have identifiers and relationships to other entities: "Apple Inc." connects to "Tim Cook" (CEO), "iPhone" (product) and "Cupertino" (location). Search engines and AI use entity recognition to understand what content is about.

The level of trust and recognition search engines and AI systems give a specific entity (person, company, brand). Entities with strong authority, such as established companies with consistent structured data, reference entries and reputable coverage, are more readily surfaced and cited. Building it takes a consistent, structured presence across the web.

Automated checks of claims, statistics and statements against reliable sources to keep content accurate. Fact verification systems cross-reference data points, detect contradictions and assign confidence scores, which protects content credibility and limits the spread of misinformation.

AI-assisted systems that analyse patterns and signals to catch fraudulent or spam content before publication, examining language patterns, domain reputation, verification status and behavioural signals. Pressonify runs a four-layer fraud check with a 0-100 risk score on every release, which protects the platform against scams, fake companies and misleading announcements.

An AI model trained on very large amounts of text that can understand and generate human-like language, such as OpenAI's GPT models (ChatGPT), Anthropic's Claude and Google's Gemini. LLMs power AI assistants, content tools and, increasingly, search itself. Because they learn from and retrieve published web content, well-structured press releases make good "LLM brain food".

Content that AI systems can easily parse, trust and quote: clear facts in plain language, a logical structure, semantic HTML, accurate metadata and JSON-LD that confirms who said what and when. Organised, unambiguous content is easier for AI to retrieve and attribute correctly, which increases the chance it is cited.

An architecture where several specialised AI agents each handle part of a task and coordinate to complete it. Pressonify uses 16+ specialised PydanticAI agents for press release generation, SEO and schema enhancement, fraud detection and citation tracking. Multi-agent systems handle problems that are too broad for a single model prompt.

Designing effective instructions (prompts) for AI language models to produce the output you want. Good prompts provide context, specify format, include examples and set constraints. Vague prompts produce vague results; precise prompts produce precise results.

A technique where a language model first retrieves relevant information from external sources (databases or indexed web content) and then generates its answer from it. RAG reduces hallucinations by grounding answers in real data. It is how AI search engines like Perplexity work: they retrieve relevant pages, such as a well-structured press release, then answer while citing them.

The body of text, images and other information used to teach AI models during training. Most LLMs are trained largely on web content, including news articles, press releases and structured data. Well-formatted, accurate press releases can become part of that corpus, helping models learn about your company and industry, which is why press releases are now AI training data as well as news.

JavaScript Object Notation for Linked Data: a way of encoding structured data with Schema.org vocabulary that search engines and AI systems can parse easily. JSON-LD is Google's recommended format and sits in a script tag, usually in the page head. It tells machines "this is a NewsArticle with headline X, published on date Y, by organisation Z."

A database of interconnected entities and relationships that search engines and AI systems use to understand and answer complex queries. Knowledge graphs power the information panels beside Google results and let AI answer questions like "Who is the CEO of Apple?" by following entity relationships.

The Schema.org type for news content, with properties such as headline, datePublished, author, publisher and image. NewsArticle markup helps search engines and AI systems understand exactly what a page is, who published it and when, and makes it eligible for article rich results. Every Pressonify release carries it.

The Schema.org type for marking up FAQ content as structured Question and Answer pairs. Since 2023 Google shows FAQ rich results only for a small set of authoritative government and health sites, but the markup still makes question-and-answer content explicit and easy for AI systems to quote accurately.

The Schema.org type for product information: name, price, availability, reviews and images. Product markup enables rich results in Google Search and Merchant Center listings, and gives AI shopping features in ChatGPT, Perplexity and Google clean facts to compare. Essential for Shopify stores that want to appear in AI shopping answers.

A Schema.org property that marks the sections of an article best suited to text-to-speech playback. Google supports it in beta for news content read aloud by Google Assistant. Speakable sections should be two or three concise sentences that stand alone as a complete answer.

A shared vocabulary for structured data, founded by Google, Microsoft, Yahoo and Yandex. Schema.org provides hundreds of types (NewsArticle, Organization, Product, Person) and properties to describe content precisely. It is the common language for telling search engines and AI what your content means.

HTML that conveys meaning about content structure with the right elements, like <article>, <section>, <header> and <nav>, instead of generic <div> tags. It helps search engines, AI systems and accessibility tools understand page structure, and is part of making content machine-readable.

Machine-readable information in a standard format (usually Schema.org in JSON-LD) that tells search engines and AI exactly what a page contains. Structured data turns unstructured HTML into explicit facts that power rich results and knowledge panels and help AI attribute content correctly. It's the difference between "some text" and "this is a company name".

Web pages set up with tracking to measure visitor behaviour, traffic sources and conversions. Analytics shows what content works, where visitors come from and what they do next, which is essential for measuring the return on a press release.

A set of rules that lets software applications communicate. APIs enable integrations: for example, a press release platform's API might let you submit releases from your CMS or pull results into your own dashboard.

The share of visitors who leave after viewing a single page without further interaction. High bounce rates can point to poor relevance, slow loading or mismatched intent. (Google Analytics 4 reports the related "engagement rate" instead.)

A distributed network of servers that caches and delivers content from locations close to each visitor, improving load times. CDNs reduce server load, help Core Web Vitals and provide DDoS protection.

When a visitor completes a desired action, such as subscribing to a newsletter, downloading a guide, requesting a demo or making a purchase. Conversion rate is the percentage of visitors who convert. Press releases drive conversions by attracting qualified visitors and giving them a clear next step.

The percentage of people who click a link after seeing it: clicks divided by impressions. A high CTR points to compelling headlines, meta descriptions and previews. Typical CTRs vary widely by position and query type, and rich results can lift them.

The internet's phone book, translating domain names (pressonify.ai) into the IP addresses computers use to reach servers. DNS records also route email (MX records) and can prove domain ownership (TXT records).

HyperText Transfer Protocol, the foundation of data exchange on the web. HTTPS adds TLS encryption for secure transmission. Modern sites should use HTTPS for security and user trust; browsers flag plain-HTTP pages as "Not secure", and Google uses HTTPS as a (lightweight) ranking signal.

The number of times your content is shown in search results, social feeds or other platforms, whether or not anyone clicks. A result with 10,000 impressions was displayed 10,000 times. Compare impressions with clicks (CTR) to judge how well headlines and previews perform.

Web design that adapts layout and content to any screen size, from phones to desktops. Most searches now happen on mobile and Google indexes the mobile version of pages first, so non-responsive sites give a poor experience and lose visibility.

A site served over HTTPS with a valid TLS (formerly SSL) certificate, shown by the padlock in the browser. It is expected of every modern website: it protects visitors' data, avoids browser warnings and is a lightweight Google ranking signal.

Confirmation that a person controls a specific domain, through email verification, DNS records or similar methods, to prevent spam and fraud. Pressonify sends a 6-digit code to a business email address on the company's domain before a release can be published, so only people who can speak for a company can publish on its behalf. It stops bad actors publishing false announcements under your name.