SEO / GEO / AI Search Glossary 2026
150 essential terms of search engine optimization, Generative Engine Optimization and e-commerce, defined by Hi-Commerce. Each term links to 1 or 2 authoritative sources (Google, Schema.org, Wikipedia, MDN, OpenAI, Anthropic, etc.) to learn more.
Classic SEO
Title tag
The Title tag (or title tag) is the main HTML element for SEO. It appears as a clickable title in SERPs and should contain the main keyword at the beginning, with an optimal length of 50 to 60 characters. For e-commerce, it is crucial to optimize it per product with unique data to improve CTR.
Meta description
The meta description is a summary under the title in search results. Although it does not directly influence ranking, it strongly impacts the click-through rate (CTR). Ideally 150-160 characters, including a call to action and the keyword. In e-commerce, it should describe the product and its benefits to encourage clicks.
H1
The H1 is the main heading of a page, essential for structuring content and signaling the topic to search engines. It should be unique, contain the main keyword, and be consistent with the title tag. In e-commerce, the H1 of the product page should match the product name, possibly with a variation.
Slug
The slug is the part of the URL that identifies a page in a readable way. It should be short, descriptive, contain hyphens and the main keyword, without special characters. For e-commerce, the product page slug should include the product name to improve understanding and SEO.
Sitemap.xml
The sitemap.xml file lists all important pages of a site to facilitate their indexing by search engines. It indicates priority and update frequency. Essential for e-commerce sites with many product pages, it helps signal new pages or updates.
Robots.txt
The robots.txt file tells search engine crawlers which pages not to crawl or index. It is crucial to avoid wasting crawl budget on irrelevant pages like shopping carts or filters. In e-commerce, session or sorting URLs are often blocked.
Canonical
The canonical tag indicates the preferred URL of a page to avoid duplicate content. In e-commerce, it is essential for product variants (sizes, colors) or filters, to concentrate SEO value on the main page and avoid penalties.
Alt text
The alt text describes the content of an image for search engines and accessibility. It should be concise, relevant, and include the keyword if possible. In e-commerce, the alt text of product images should accurately describe the item to improve image SEO.
Internal anchor
The internal anchor is the clickable text of a link pointing to another page on the same site. It should be descriptive and contain relevant keywords. Good internal linking with optimized anchors improves user navigation and distributes link equity.
Breadcrumb
The breadcrumb is a navigation trail that shows the hierarchy of pages. It improves user experience and helps search engines understand the site structure. In e-commerce, it allows users to easily go back to categories.
301 redirect
The 301 redirect is a permanent redirect from one URL to another, passing most of the SEO value. Essential when changing URLs or restructuring an e-commerce site to avoid broken links and preserve rankings.
www vs non-www
The choice between www and non-www concerns the main domain of the site. One must be chosen and the other redirected via 301. This avoids dispersing link equity. In e-commerce, www is recommended for cookies and multiple domains.
http vs https
Switching to HTTPS (encrypted) is a trust signal for Google. It is mandatory for e-commerce since 2018, as it protects sensitive data. An HTTP site may be penalized and lose traffic. Migration requires a 301 redirect.
Mobile-first indexing
Google primarily uses the mobile version of a site for indexing and ranking. An e-commerce site must be responsive, fast, and offer an optimal mobile experience. Without this, rankings decline, affecting sales.
Structured navigation
Structured navigation (or site architecture) organizes pages into logical categories and subcategories. It facilitates crawling and user experience. For e-commerce, a well-thought-out architecture groups products by theme and improves internal linking.
Semantic SEO
Semantic cocoon
The semantic cocoon is an internal linking strategy that connects a pillar page (general) to satellite content (specific) around the same theme. This strengthens topical authority and improves rankings for related queries. In e-commerce, it can be used for a product category.
Topical authority
Topical authority measures a site’s credibility on a given topic. Google favors expert sites. To build it, you need to publish comprehensive and interconnected content. In e-commerce, this means covering each product family in depth.
Entity
An entity is a concept or identifiable object (person, place, brand). Google builds a knowledge graph from entities. In semantic SEO, you should optimize for entities (e.g., product name) rather than exact keywords. This improves relevance.
Knowledge graph
Google’s Knowledge Graph is a structured database linking entities. To appear in it, you must be a recognized entity. In e-commerce, brands can appear in the Knowledge Panel by optimizing their Google My Business listing and web presence.
NLP
NLP (Natural Language Processing) allows Google to understand the meaning of queries and content. It detects synonyms, context, and intent. In SEO, you should write naturally, with semantic variations, to align with NLP.
Co-occurrence
Co-occurrence refers to the simultaneous presence of keywords in a context. Google uses it to understand semantic relationships. Well-co-occurring content (e.g., ‘shoe’ with ‘sole’, ‘lace’) reinforces topicality. In e-commerce, describe product attributes.
Topic cluster
A topic cluster is a group of linked pages around a central subject (pillar). Each satellite page covers a subtopic and points to the pillar. This improves internal linking and topical authority. In e-commerce, each product category can be a cluster.
Semantic pillar
The pillar page is a long, comprehensive page covering a main topic. It serves as a central hub for a cluster. It should link to satellite articles. In e-commerce, a buying guide page can be a pillar page pointing to product sheets.
Internal linking
Internal linking is the network of links between pages of a site. Good internal linking distributes link equity, facilitates navigation, and helps Google understand the structure. In e-commerce, it links product sheets to categories and blog posts.
Search intent
Search intent is the goal behind a query (informational, transactional, navigational). Google analyzes it to display the right results. In SEO, content must match the intent: product sheets should respond to purchase intent.
Semantic search
Semantic search uses the meaning of words and their context to deliver relevant results, beyond exact matching. It relies on NLP and entities. For SEO, content should be optimized naturally and thematically.
Embeddings
Embeddings are vector representations of words or sentences that capture their meaning. Used by LLMs and Google, they help understand semantic similarities. In SEO, they assist in generating related content.
Vectorization
Vectorization transforms text into numerical vectors for processing by algorithms. It is the basis of modern NLP. In SEO, vectorization is used in search engines for semantic matching. It is not directly optimizable.
Latent Dirichlet Allocation
LDA is a topic modeling technique that extracts themes from a document corpus. In SEO, it can be used to identify topics to cover. However, Google uses more advanced methods like BERT.
BERT
BERT (Bidirectional Encoder Representations from Transformers) is a Google NLP model that understands bidirectional context. It significantly improves query understanding. In SEO, it requires writing natural, contextual content without keyword stuffing.
GEO / AI Search
GEO
GEO (Generative Engine Optimization) optimizes content to be cited by generative search engines (AI Overviews, chatbots). It aims to appear in synthesized answers rather than traditional links. In e-commerce, this means creating content that is easily extractable by LLMs.
AI Overviews
AI Overviews are AI-generated summaries at the top of SERPs. They directly answer the query. To appear in them, content must be clear, structured, and authoritative. In e-commerce, optimizing buying guides and FAQs can help.
llms.txt
The llms.txt file is a text file at the root of the site, intended for LLMs, listing important pages to scan. It functions as a sitemap for AI. In e-commerce, it can point to key categories and product sheets to guide models.
llms-full.txt
The llms-full.txt file contains the full site content for LLMs in plain text. It allows models to understand the entire content without crawling. For e-commerce, it can include product descriptions and guides, but watch out for size.
LLM citation
LLM citation occurs when the model references the site as an information source. It depends on reliability and visibility. In GEO, content must be optimized to be cited in generative answers.
Agentic search
Agentic search refers to AI agents that autonomously perform complex search tasks. They combine multiple queries and sources. For SEO, information must be structured so that these agents can easily exploit it.
Prompt research
Prompt research involves anticipating the queries users will ask LLMs (e.g., ‘What is the best product X?’). Optimizing content to answer these precise prompts improves visibility in AI responses.
Generative Engine Optimization
Optimization for generative engines (GEO). Unlike traditional SEO, GEO aims for content to be used as a source by generative AIs to produce synthetic answers. This includes structuring, clarity, and authority.
Citatability
Citatability is the ability of content to be cited by other sources, including LLMs. It depends on reliability, originality, and structuring. In e-commerce, comparative studies and data-driven content are highly citable.
AI Snippet
An AI Snippet is a content excerpt used by AI in a response. It must be concise, factual, and well-formatted. To appear, direct answers to questions (FAQs, definitions) are needed. In e-commerce, product features can be snippets.
Multi-LLM ranking
Multi-LLM ranking evaluates content visibility across multiple models (ChatGPT, Perplexity, etc.). Each model has its criteria. Optimizing for all requires neutral, well-sourced, and structured content (data, lists).
Retrieval Augmented Generation
RAG (Retrieval Augmented Generation) combines a knowledge base with an LLM to generate more accurate answers. Web content must be easily retrievable (well-indexed, structured) to be used in these pipelines.
Source authority IA
Source authority for AI measures a site’s reliability in the eyes of models. Google and LLMs favor recognized sources. In e-commerce, having quality backlinks and positive reviews strengthens this authority.
Bounce clicks
Bounce clicks are clicks on a search result that lead to a page immediately left. Google interprets them as a signal of dissatisfaction. To avoid them, the page must match the user’s intent, especially in e-commerce.
Bounce clicks
Defensive metric used by Google in 2025-2026 to distinguish a productive click (user stays, scrolls, interacts) from a bounce click (immediate return to SERP). LLMs and AI Overviews introduced a new bias: a visitor arriving via AI citation often leaves quickly, which can trick traditional quality algorithms. Optimizing bounce clicks means refining above-the-fold content and consistency with the snippet promise to maximize deep engagement.
Schema.org
JSON-LD
JSON-LD is a format for embedding Schema.org structured data in a page. It is preferred by Google because it is easy to implement and does not degrade HTML. In e-commerce, it is used to mark up products, reviews, and FAQs.
Article schema
The Article schema (Article, NewsArticle, BlogPosting) marks up blog posts or guides. It improves SERP display with rich snippets. For e-commerce, it can be used for content articles.
FAQPage
The FAQPage schema allows displaying an enriched FAQ with expandable questions/answers in SERPs. It attracts attention and improves CTR. In e-commerce, it is ideal for frequently asked questions about products.
SoftwareApplication
This schema marks up software applications. For a software e-commerce, it allows displaying ratings, prices, and direct downloads. It can also be used for SaaS.
DefinedTermSet
DefinedTermSet allows marking up a glossary or dictionary. Each term has a definition. Useful for e-commerce sites that have ‘definitions’ sections related to products.
Product schema
The Product schema is essential for e-commerce. It describes a product with its name, price, availability, and ratings. It allows displaying rich snippets (price, stock) in SERPs, increasing CTR.
Organization schema
The Organization schema represents a company. It can include name, logo, and contact information. For e-commerce, it reinforces credibility and can display a Knowledge Panel.
BreadcrumbList
BreadcrumbList marks up the breadcrumb trail. It improves SERP display with navigation links. Easy to implement, it helps Google understand site structure.
HowTo
The HowTo schema marks up step-by-step guides. It can display with numbered steps in SERPs. Useful for product usage tutorials in e-commerce.
Review
The Review schema marks up a review of a product or service. It can display with a star rating. In e-commerce, it encourages customer reviews and improves visibility.
AggregateRating
AggregateRating summarizes product ratings (average rating, number of reviews). It appears in rich snippets. Essential for product pages in e-commerce.
Person
The Person schema represents an individual. Used for article authors or founders. In e-commerce, it can be used for the ‘About’ page or testimonials.
Author schema
The Author schema indicates the author of content. It can improve E-E-A-T by showing experts. For e-commerce, blog articles can have an identified author.
SameAs
SameAs is a Schema.org property that links an entity to its profiles on other platforms (Facebook, Twitter). This reinforces digital identity consistency and can improve the Knowledge Graph.
ItemList
ItemList marks up a list of items (e.g., product list, top 10). It can display as a carousel. In e-commerce, it is used for category lists or selections.
E-commerce SEO
Product page SEO
An SEO-optimized product page includes a unique title, an attractive meta description, optimized images, structured data, and quality textual content. It must address the purchase intent and contain key information (price, availability).
Faceted navigation
Faceted navigation allows filtering products by attributes (size, color). It can create duplicates if poorly managed. Use canonical tags, robots noindex, or AJAX. For SEO, it is crucial to avoid dilution.
Variants canonical
For products with variants (color, size), use a canonical URL pointing to the main page. Each variant can have its own URL with a parameter, but the canonical points to the base version. This avoids duplicate content.
Google Merchant Center
Google Merchant Center (GMC) manages product data for Google Shopping. It requires a quality product feed with mandatory attributes. An optimized GMC improves visibility in Shopping ads and free listings.
Product feed
The product feed is an XML or CSV file listing products with their attributes (title, price, GTIN). It is used by Google Merchant Center and comparison shopping engines. A well-structured and regularly updated feed is crucial for e-commerce.
Rich results Product
Product rich results display additional information in SERPs: price, availability, reviews. Obtained via Product schema, they increase CTR and visibility. In e-commerce, they can make the difference.
Structured reviews
Structured reviews use the Review or AggregateRating schema. They allow displaying stars in SERPs. Essential for e-commerce, they build trust and can improve click-through rate.
Product availability
Availability (in stock, out of stock) is a key attribute of structured data and the feed. Google displays it in results. In e-commerce, real-time updates are paramount to avoid frustration.
Dynamic pricing
Dynamic pricing changes based on demand, time, or customer. For SEO, use the Offer schema with the priceSpecification property. Google can display the price in real time. Beware of frequent changes that may disrupt.
Schema Offer
The Offer schema describes a specific offer of a product (price, currency, availability). It is used within Product schema. For e-commerce, each variant can have its own Offer.
Internal linking PDP
Internal linking between product pages (PDPs) and categories is crucial. Use relevant links (similar products, accessories). This distributes link equity and aids indexing.
Categories/silos
Categories and silos organize products. Each category should have an optimized page with an H1, a unique description, and links to subcategories or products. This structures the site and improves topicality.
Filter URL
Filter URLs (e.g., /category?color=red) can create duplicate content. Manage them with canonical, noindex, or parameters in Search Console. Ideally, use clean URLs with slugs.
Out of stock SEO
An out-of-stock product should be managed: do not delete it, but mark it as ‘out of stock’ via schema, redirect it, or temporarily noindex it. This avoids losing acquired rankings.
Pagination SEO
Pagination (pages 1, 2, 3) should be managed with rel=next/prev or an indexing strategy. Google may treat subsequent pages as separate entities. In e-commerce, avoid overly long pages.
Netlinking
Backlink
A backlink is a link from an external site to yours. It is a signal of popularity and trust for Google. The more relevant and high-quality the source site, the more valuable the backlink. In e-commerce, backlinks from blogs or articles strengthen authority.
Domain Rating DR
Domain Rating (DR) is an Ahrefs metric that evaluates the strength of a site’s backlink profile (score from 0 to 100). A high DR indicates good authority. For link building, we target sites with high DR to gain popularity.
Domain Authority DA
Moz’s Domain Authority (DA) is a predictive score of a site’s ranking. It ranges from 1 to 100, based on several factors including backlinks. Although not official, it is used to assess SEO difficulty.
Trust Flow
Majestic’s Trust Flow (TF) measures a site’s trust based on the quality of backlinks. It is related to proximity to trusted sites. A good balance between Trust Flow and Citation Flow is important.
Citation Flow
Majestic’s Citation Flow (CF) measures the quantity of backlinks, regardless of their quality. A high CF with a low TF may indicate low-quality links. Ideally, TF should be close to CF.
Anchor text
Anchor text is the clickable text of a backlink. It should be relevant and varied (brand, keyword, URL). An excess of optimized anchors can be penalized. In e-commerce, brand names or natural anchors are often used.
Dofollow
A dofollow link passes popularity (PageRank) to the target site. It is the default link type. Quality dofollow links are the goal of link building. In e-commerce, they help improve authority.
Nofollow
The rel=’nofollow’ attribute tells Google not to follow the link or pass popularity. It is used for paid links, comments, or untrustworthy content. It does not directly affect ranking.
UGC link
The rel=’ugc’ (User Generated Content) attribute is used for links in comments or forums. It indicates the link is user-generated. Google treats it as nofollow.
Sponsored link
The rel=’sponsored’ attribute is intended for advertising or sponsored links. It avoids a penalty for manipulation. In e-commerce, it is used for affiliate links or partnerships.
PBN
A PBN (Private Blog Network) is a set of sites created to generate backlinks to a main site. This practice violates Google’s guidelines and can lead to severe penalties. To be avoided.
Tier 1/2 links
Tier 1 links point directly to your site. Tier 2 links point to Tier 1 sites. This stratification diversifies sources. In link building, we build quality Tier 1 links and can reinforce with Tier 2.
Reciprocal link
A reciprocal link is an exchange of links between two sites. Google considers this manipulative if excessive. In moderate quantity with relevant sites, it can be natural.
Broken link building
Broken link building involves finding dead links on other sites and offering to replace them with your content. It is an ethical link building technique that adds value. Useful for e-commerce with guides.
Skyscraper
The Skyscraper technique involves creating more comprehensive and higher-quality content than the best existing articles, then asking sites that link to them to link to yours. Effective for link building.
Analytics & Metrics
CTR
CTR (Click Through Rate) is the ratio of clicks to impressions. It measures the attractiveness of a result in the SERPs. A good CTR depends on the title, meta description, and structured data. In e-commerce, aim for 5-10%.
Impressions
Impressions indicate the number of times a page appears in search results. They depend on ranking and query. In SEO, increasing impressions requires good positioning on high-volume keywords.
Average position
Average position is the average ranking of a page for a query over a given period. The lower it is, the better. Google Search Console often shows the position. The goal is to reach the first page.
Conversion rate CR
Conversion rate (CR) is the percentage of visitors who complete a desired action (purchase, sign-up). It is crucial in e-commerce. A CR of 1-3% is typical, optimizable through design and content.
ROAS
ROAS (Return On Ad Spend) measures revenue generated relative to ad spend. A ROAS of 4 means €4 earned for €1 spent. It is often used to evaluate Shopping campaigns.
RPM
RPM (Revenue Per Mille) is revenue per 1000 impressions. Used in display advertising. In e-commerce, it can measure traffic monetization, especially if banners are present.
LTV
LTV (Lifetime Value) is the total revenue generated by a customer over the entire relationship with the brand. In e-commerce, it helps determine the acceptable marketing budget to acquire a customer.
CAC
CAC (Customer Acquisition Cost) is the amount spent to acquire a new customer. It includes marketing and sales costs. Ideally, LTV should be greater than CAC. In e-commerce, it is crucial for profitability.
GA4
GA4 is the latest version of Google Analytics, based on events. It enables cross-device and predictive tracking. For e-commerce, it offers detailed reports on purchase journeys and attribution.
GSC
Google Search Console (GSC) is a free tool to monitor a site’s presence in the SERPs. It provides data on queries, clicks, impressions, and technical issues. Essential for SEO.
Bounce rate
Bounce rate is the percentage of sessions with a single page viewed. A high rate may indicate a lack of relevance or poor UX. In e-commerce, a high bounce rate on a product page is often normal.
Pages per session
Pages per session measures engagement. The higher it is, the more the user explores. In e-commerce, it may indicate that the visitor is viewing multiple products.
Time on page
Time on page indicates interest in the content. For articles, a long time is good; for product pages, a short time may be normal if the user quickly finds the information.
Engaged sessions
In GA4, an engaged session lasts more than 10 seconds, has a conversion, or at least 2 page views. It is a better metric than bounce rate. It measures real interaction.
Attribution
Attribution determines how marketing channels are credited for a conversion. There are several models (first click, last click, linear). In e-commerce, choosing the right model helps optimize budget.
Technical & Crawl
Crawl budget
Crawl budget is the number of pages a search engine will crawl on a site within a given time. Optimizing crawl budget involves eliminating useless pages, improving speed, and using the sitemap. Essential for large e-commerce sites.
Robots.txt
The robots.txt file tells crawlers which parts of the site not to explore. It must be placed at the root. In e-commerce, cart, session, or filter URLs are often blocked to save crawl budget.
Status code 200/301/404/410
HTTP codes indicate the state of a page. 200 (OK), 301 (permanent redirect), 404 (not found), 410 (deleted). In SEO, avoid 404s and use 301 redirects when changing URLs.
Hreflang
The hreflang tag indicates the language and region of a page for multilingual sites. It helps Google display the correct version in each country. In international e-commerce, it is essential.
Canonical URL
The canonical URL is the preferred version of a page. It resolves duplicate content issues. In e-commerce, it is used for product variants or filters.
Render budget
Render budget is the time Google spends rendering a page (JavaScript). A site that is too heavy may not be properly indexed. In e-commerce, minimize JS to ensure good indexing.
JavaScript SEO
JavaScript SEO concerns the optimization of sites that rely on JS for content display. Google can execute JS, but not always well. Use SSR or pre-rendering for critical content.
Indexability
Indexability is the ability of a page to be indexed by search engines. It can be blocked by robots.txt, meta robots noindex, or technical issues. Checking indexability is essential.
Crawl depth
Crawl depth is the number of clicks needed from the homepage to reach a page. Ideally, all pages should be within 3 clicks. In e-commerce, this is managed with a good site structure.
Sitemap XML
The XML sitemap is a file that lists all important pages of a site. It facilitates indexing. In e-commerce, it should include categories, product pages, and content pages.
Server-side rendering SSR
SSR generates HTML on the server before sending it to the client. It is beneficial for SEO because Google sees the full content. In e-commerce, it is used for dynamic web applications.
Static site generation SSG
SSG generates static HTML pages at build time. They are very fast and well indexed. Used for e-commerce sites with infrequently changing content.
Edge SEO
Edge SEO consists of applying optimizations at the CDN level (Cloudflare Workers). This allows quick modifications without touching the server. Useful for A/B testing or geolocation.
Log analysis
Server log analysis allows you to see how Googlebot crawls the site. You can identify ignored pages, errors, and optimize crawl budget. Essential for large sites.
Crawler trap
A crawler trap is an infinite loop of pages that traps Google’s robot, wasting crawl budget. This can happen with poorly configured sessions or filters. To be avoided.
Core Web Vitals
LCP
LCP (Largest Contentful Paint) measures the loading time of the largest visible element (image, text block). A good LCP is under 2.5 seconds. In e-commerce, it affects user experience and ranking.
INP
INP (Interaction to Next Paint) measures a page’s responsiveness to user interactions (clicks, keystrokes). A good INP is under 200 ms. It replaces FID as a responsiveness metric.
CLS
CLS (Cumulative Layout Shift) measures a page’s visual stability. A good score is under 0.1. In e-commerce, high CLS can cause purchase buttons to shift, frustrating users.
TTFB
TTFB (Time To First Byte) measures the time between the request and the first byte received from the server. Ideally under 600 ms. It depends on hosting and back-end optimization.
FCP
FCP (First Contentful Paint) measures the moment when the first text or image element appears. A good FCP is under 1.8 seconds. It gives a first impression of speed.
Speed Index
Speed Index measures how quickly a page’s content becomes visible. The lower, the better. It is calculated by Lighthouse and integrated into PageSpeed Insights.
Lighthouse
Lighthouse is an automated Google tool that audits performance, accessibility, SEO, etc. It provides scores and recommendations. Essential for optimizing an e-commerce site.
PageSpeed Insights
PageSpeed Insights (PSI) analyzes a page’s speed on mobile and desktop, based on lab and field data (CrUX). It provides improvement suggestions. Essential for SEO.
CrUX
CrUX (Chrome User Experience Report) is a dataset of real user experiences from Chrome users. It measures real Core Web Vitals. Google uses it for ranking.
Real User Monitoring RUM
RUM (Real User Monitoring) collects real user performance data. Unlike lab tests, it reflects real conditions. Useful for monitoring Core Web Vitals in production.
Largest Contentful Paint
Largest Contentful Paint (LCP) is a Core Web Vital that measures the loading time of the largest content element visible in the viewport. A fast LCP improves perceived speed.
Interaction to Next Paint
Interaction to Next Paint (INP) evaluates the latency of all interactions on a page. A low INP means a responsive interface. Google introduced it as a key metric in 2024.
Cumulative Layout Shift
Cumulative Layout Shift (CLS) quantifies unexpected layout shifts. High CLS can lead to erroneous clicks. In e-commerce, avoid elements that load late.
Total Blocking Time TBT
TBT (Total Blocking Time) measures the time during which the main thread is blocked by long tasks, preventing interactions. A low TBT is essential for good responsiveness.
First Input Delay FID
FID (First Input Delay) measures the delay before a page responds to the first interaction. It is replaced by INP. A good FID is under 100 ms. It affects user experience.
LLM Tools
ChatGPT
ChatGPT is a conversational AI assistant developed by OpenAI, based on GPT. It can generate content, answer questions, and assist with research. In SEO, it is used to write articles, descriptions, and brainstorm.
Perplexity
Perplexity is an AI search engine that generates answers with citations. It combines LLMs with web sources. For SEO, appearing in Perplexity requires high-quality and well-referenced content.
Claude
Claude is an AI assistant developed by Anthropic, focused on safety and reliability. It can be used for data analysis, writing, and research. For SEO, it can help create structured content.
Gemini
Gemini is Google’s multimodal model, capable of processing text, images, and more. It integrates Google data. For SEO, it influences AI Overviews and answers in Search.
Grok
Grok is an LLM developed by xAI, known for its direct tone. It can access X (Twitter) in real time. For SEO, it can be used to generate content related to current events.
Copilot
Copilot is Microsoft’s AI assistant, integrated into Bing. It generates answers with citations. For SEO, optimizing for Bing can improve visibility in Copilot.
RAG
RAG combines information retrieval and text generation. It allows LLMs to use external sources. In SEO, it is crucial to structure content so that it is easily retrievable.
Embeddings
Embeddings are vector representations of data (text, images). They measure semantic similarity. In SEO, they are used for semantic search and content clustering.
Vector database
A vector database stores and queries embeddings. It is used in RAG systems to retrieve relevant information. Optimizing content for these databases can improve visibility.
Fine-tuning
Fine-tuning involves specializing an LLM on a domain by training it on specific data. In SEO, one can fine-tune a model to generate content consistent with the brand.
Prompt engineering
Prompt engineering is the art of formulating instructions to LLMs to obtain precise answers. In SEO, it is used to generate optimized content or analyze data.
System prompt
The system prompt is an initial instruction given to an LLM to define its behavior. In SEO, it can be used to constrain tone, format, or topics to address.
Function calling
Function calling allows LLMs to execute functions (API, calculations) to respond. In SEO, it can be used to automate tasks like generating meta descriptions.
Tool use
Tool use allows LLMs to interact with external tools (web search, calculations). In SEO, this can include plugins to scrape or analyze pages.
MCP Model Context Protocol
MCP (Model Context Protocol) is a standard for providing context to LLMs securely. It allows integrating dynamic sources. For SEO, it can improve answer accuracy.