SEO, GEO, web dev, React, Vue, Node.js & AI terms — defined

The practice of optimizing content to be cited by AI assistants like ChatGPT, Perplexity, Claude, Gemini and Google AI Overviews.

Generative Engine Optimization (GEO) is the practice of optimizing your brand, content, and entity signals to be cited by AI assistants — ChatGPT, Perplexity, Claude, Gemini, Microsoft Copilot, and Google AI Overviews. GEO is the natural successor to SEO. Where SEO targets ranked links, GEO targets cited sentences. The two overlap heavily on the technical side (schema, semantic HTML, E-E-A-T) but diverge on content strategy: GEO rewards statistical specificity, expert quotes, citation-worthy data, and authoritative third-party mentions far more than keyword density. The Princeton University GEO study (2023) demonstrated that certain content modifications can improve AI citation rates by 30–40%.

Optimizing content to be selected as the direct answer in AI-generated responses.

Answer Engine Optimization (AEO) is the practice of structuring content so that AI assistants select and surface it as the direct answer to a user query. AEO is closely related to GEO but emphasizes the specific structural patterns that make content 'extraction-ready' — front-loaded answers, question-format H2s, FAQPage schema, concise 40–60 word answer passages, and named entities. AEO and GEO are often used interchangeably; technically AEO is a subset of GEO focused on the answer-extraction layer.

A plain-text manifest at /llms.txt that gives AI crawlers a curated map of your site's most important content.

llms.txt is an emerging standard (proposed by Answer.AI's Jeremy Howard in 2024) for a plain-text Markdown manifest located at the root of a domain (e.g., domain.com/llms.txt). It provides AI assistants with a curated, human-readable index of the site's most important pages, organized by H1/H2 sections and descriptive links. The companion standard llms-full.txt provides the full text of the indexed pages in a single file for direct AI ingestion. Major AI platforms (ChatGPT, Perplexity, Claude) explicitly support llms.txt. Lighthouse now audits for llms.txt as part of agentic-browsing best practices.

A type of AI model trained on massive text corpora to understand and generate natural language.

A Large Language Model (LLM) is a type of deep-learning AI model trained on massive text corpora (often hundreds of billions of tokens) to understand and generate natural language. Modern LLMs like GPT-4, Claude 3, Gemini 1.5, and Llama 3 use the transformer architecture and are typically fine-tuned with reinforcement learning from human feedback (RLHF). For SEO/GEO purposes, the key distinction is between LLMs in 'memory' mode (relying on training data) versus 'retrieval' mode (running a search query and synthesizing from live results). Citation by LLMs is only possible in retrieval mode.

Google's AI-generated answer box that appears at the top of search results for many queries.

An AI Overview is Google's AI-generated answer box that appears at the top of search results for many queries, particularly informational ones. AI Overviews synthesize a response from multiple web sources and display inline citations. They were rolled out to 100+ countries in 2024–2025 and now appear on a significant share of queries. AI Overviews have reduced click-through rates to organic results by 30–60% for affected queries, making it critical for brands to be cited in the AI Overview itself (via GEO) rather than relying solely on traditional blue-link rankings.

An AI search engine that synthesizes answers from live web results with inline citations.

Perplexity is an AI-powered search engine that retrieves live web results and synthesizes natural-language answers with inline citations. It is the fastest-growing AI search product, with 10M+ monthly active users. Perplexity uses its own retrieval index (not Bing or Google) and applies a query fan-out process to gather context. Brands can be cited in Perplexity by being indexed (via PerplexityBot in robots.txt) and producing content that passes Perplexity's authority and specificity filters.

Anthropic's family of large language models, used in claude.ai and Claude API.

Claude is Anthropic's family of large language models, including Claude 3 Opus, Sonnet, and Haiku. Claude powers the claude.ai consumer chat interface and is available via API. Claude's retrieval is similar to ChatGPT's — in retrieval mode, it fetches live web results and cites sources. The Claude-specific crawlers are Claude-SearchBot and Claude-User. To be cited by Claude, brands should allow these bots in robots.txt and produce content with the same statistical specificity and expert attribution that other LLM retrievers prefer.

Structured data vocabulary (schema.org) embedded in HTML to help search engines and AI understand page content.

Schema markup is a structured data vocabulary maintained at schema.org, embedded in HTML (typically as JSON-LD) to help search engines and AI systems understand the meaning and context of page content. Common schema types for B2B and SaaS sites include Organization, WebSite, Service, Product, FAQPage, Article, BreadcrumbList, LocalBusiness, HowTo, SoftwareApplication, Review, and Person. Schema is critical for GEO: it gives AI retrievers a deterministic, machine-readable map of your entities, services, and content relationships.

JavaScript Object Notation for Linked Data — the recommended format for embedding schema markup.

JSON-LD (JavaScript Object Notation for Linked Data) is the format Google officially recommends for schema markup. It is embedded as a <script type='application/ld+json'> tag in the page <head> or body, and is fully machine-readable without affecting the visual presentation of the page. JSON-LD is preferred over Microdata and RDFa because it is cleaner, easier to maintain, and Google's documentation explicitly recommends it. AI retrievers also prefer JSON-LD for the same reason — it is deterministic, parseable, and unambiguous.

Code embedded in a page that explicitly labels content entities (products, reviews, people, organizations) for machines.

Structured data is code embedded in a web page that explicitly labels content entities — products, reviews, people, organizations, events, articles, FAQs — for search engines and AI assistants to understand. The two most common formats are JSON-LD and Microdata; Google recommends JSON-LD. Structured data powers rich results in Google Search (star ratings, FAQ dropdowns, recipe cards, knowledge panels) and is the primary mechanism for GEO: AI retrievers consume structured data to build entity graphs and determine which sources to cite.

A <link rel='canonical'> tag that tells search engines which URL is the authoritative version of a page.

A canonical URL is specified via the <link rel='canonical'> tag in a page's <head> to tell search engines which URL is the authoritative version of that page. This is critical when the same content is reachable via multiple URLs (e.g., with and without trailing slashes, with UTM parameters, with sort/filter parameters on e-commerce sites). Without proper canonicalization, search engines may split ranking signals across duplicates or pick the wrong URL as the canonical, diluting SEO value. For AI retrieval, canonicals help disambiguate which version of a page should be cited.

A file at /sitemap.xml that lists all indexable pages on a site for search engine crawlers.

An XML sitemap is a file located at /sitemap.xml that lists all indexable pages on a site in a machine-readable format. Sitemaps help search engines discover and prioritize pages, especially on large sites or sites with deep navigation. For SEO, the sitemap should be auto-generated and updated whenever new pages are added, then submitted to Google Search Console and Bing Webmaster Tools. For GEO, sitemaps help AI retrievers understand site structure and content hierarchy. A well-formed sitemap is one of the most basic — and most frequently neglected — SEO foundations.

A file at /robots.txt that tells crawlers which paths they are allowed or disallowed from accessing.

The robots.txt file is located at the root of a domain (e.g., domain.com/robots.txt) and provides directives to web crawlers about which paths they are allowed to access. For modern SEO, robots.txt is critical for two reasons: (1) blocking sensitive areas (/api/, /admin/, /thanks/) and (2) explicitly allowing or disallowing AI crawlers. For GEO, brands must allow retrieval bots like OAI-SearchBot, PerplexityBot, Claude-SearchBot, and Claude-User, while often blocking training bots like GPTBot, CCBot, and Google-Extended. A misconfigured robots.txt can completely hide a brand from ChatGPT, Perplexity, or Claude search.

HTML attribute that signals the language and geographic targeting of a page to search engines.

The hreflang attribute (implemented as <link rel='alternate' hreflang='en-US' href='...'> in a page's <head>) signals the language and geographic targeting of a page to search engines. It is critical for sites with multiple language versions, regional variants, or country-specific content. Proper hreflang implementation prevents duplicate content issues across regional versions and ensures users see the right version for their locale. For global brands, hreflang is also a GEO signal — AI retrievers use locale signals to determine which version of a page to cite for a given query.

Experience, Expertise, Authoritativeness, Trustworthiness — Google's framework for assessing content quality.

E-E-A-T stands for Experience, Expertise, Authoritativeness, and Trustworthiness. It is Google's framework for assessing content quality, formalized in the Search Quality Rater Guidelines. While E-E-A-T is not a direct ranking factor, it influences the algorithms that assess authority signals. For GEO, E-E-A-T is even more critical: AI retrievers explicitly weight author credentials, named experts, third-party citations, and brand mentions when deciding which sources to cite. Demonstrating E-E-A-T requires author bios, credentials, primary-source citations, expert quotes, and consistent entity signals across the web.

A structured representation of entities and their relationships, used by Google and AI systems to understand the world.

A knowledge graph is a structured representation of entities (people, places, organizations, concepts) and the relationships between them. Google's Knowledge Graph powers the information panels that appear on the right side of branded search results, and is the underlying structure that AI retrievers use to understand brand identity. For GEO, aligning your brand entity across your website, Wikidata, Wikipedia (if eligible), Google Business Profile, LinkedIn, GitHub, Crunchbase, and review platforms creates a consistent, machine-readable knowledge graph that AI systems can confidently cite.

HTML that uses meaningful tags (<article>, <section>, <nav>) rather than generic <div> and <span>.

Semantic HTML uses meaningful tags (<article>, <section>, <nav>, <header>, <footer>, <main>, <aside>, <h1>–<h6>, <p>, <ul>, <ol>, <li>) to convey the structure and meaning of content, rather than generic <div> and <span> elements with CSS styling. Semantic HTML is critical for both SEO and GEO: search engines and AI retrievers use the tag structure to understand the role of each piece of content (e.g., is this the main heading? a navigation list? a content article?). A site that uses real <h1>–<h6> tags rather than <p class='heading'> consistently ranks higher and is cited more often by AI.

A React-based framework for production-grade web applications, with SSR, SSG, and ISR built in.

Next.js is an open-source React-based framework developed by Vercel. It is the de-facto standard for production-grade React applications, with built-in support for Server-Side Rendering (SSR), Static Site Generation (SSG), Incremental Static Regeneration (ISR), React Server Components, App Router, image optimization, font optimization, code splitting, and edge functions. Next.js is used by some of the world's highest-traffic sites (Vercel, Stripe, Notion, Linear, The Verge, Hulu, TikTok). For SEO and GEO, Next.js produces clean semantic HTML, fast Core Web Vitals, and full control over schema, llms.txt, and head markup — the three things AI retrievers and search engines need most.

A product with the minimum features needed to validate a market hypothesis with real users.

A Minimum Viable Product (MVP) is a product with the minimum feature set needed to validate a market hypothesis with real users. The MVP is built quickly (typically 8–16 weeks), launched to early adopters, and then iterated based on real feedback. For SaaS, an MVP typically includes: core feature set, basic auth, Stripe billing, one or two integrations, a simple admin dashboard, and basic analytics. The opposite of an MVP is a 'full product' build, which can take 6–18 months and $100K+. We specialize in SaaS MVPs — most of our SaaS engagements ship in 8–16 weeks for $25,000–$100,000.

A JavaScript library for building user interfaces, maintained by Meta. The most widely used frontend framework.

React is a JavaScript library for building user interfaces, maintained by Meta (Facebook). It uses a component-based architecture with a virtual DOM for efficient rendering. React is the most widely used frontend library in the world, powering Instagram, Facebook, Netflix, Airbnb and millions of websites. Key features include JSX syntax, hooks (useState, useEffect, useMemo), React Server Components (RSC), and a massive ecosystem (Next.js, Remix, Gatsby). React is ideal for single-page applications (SPAs), progressive web apps (PWAs), and server-rendered sites via Next.js.

A progressive JavaScript framework for building user interfaces, known for its gentle learning curve and flexibility.

Vue.js is a progressive JavaScript framework created by Evan You. It is known for its gentle learning curve, excellent documentation, and flexibility. Vue uses a template-based syntax with reactive data binding, a composition API (similar to React hooks), and single-file components (.vue files). Vue powers Alibaba, Xiaomi, GitLab and thousands of production applications. The Nuxt.js meta-framework provides SSR, SSG and full-stack capabilities on top of Vue. Vue is ideal for teams that want a simple, approachable framework with strong TypeScript support and excellent performance.

A platform and framework for building single-page client applications, maintained by Google.

Angular is a platform and framework for building single-page client applications, maintained by Google. It uses TypeScript, dependency injection, RxJS for reactive programming, and a powerful CLI for code generation. Angular is used by Google, Microsoft, Samsung, BMW and large enterprises that need a batteries-included framework with strong opinions on architecture. Key features include modules, components, services, decorators, template syntax, two-way data binding, and Angular Universal for SSR. Angular is ideal for large enterprise applications with complex state management and real-time features.

A compiler-based UI framework that shifts work from the browser to the build step.

Svelte is a compiler-based UI framework created by Rich Harris. Unlike React, Vue or Angular, Svelte has no virtual DOM — it compiles components into highly efficient vanilla JavaScript at build time. This results in smaller bundle sizes, faster runtime performance and less boilerplate. SvelteKit is the official meta-framework for Svelte, providing SSR, SSG, routing and full-stack capabilities. Svelte is used by Apple, Spotify, The New York Times and Shopify (for parts of their admin). It is ideal for projects that need minimal JavaScript, excellent performance and a delightful developer experience.

A JavaScript runtime built on Chrome's V8 engine, enabling server-side JavaScript execution.

Node.js is an open-source, cross-platform JavaScript runtime built on Chrome's V8 engine. It allows developers to run JavaScript on the server side, enabling full-stack JavaScript development. Node.js uses an event-driven, non-blocking I/O model that makes it lightweight and efficient for real-time applications, APIs and microservices. Major frameworks include Express.js (minimal), Fastify (performance), NestJS (enterprise), and Koa.js (modern). Node.js powers LinkedIn, PayPal, Netflix, NASA and millions of production APIs. It is the most popular backend runtime for modern web development.

A high-level, interpreted programming language known for its readability, versatility and rich ecosystem.

Python is a high-level, interpreted programming language known for its simple syntax, readability and vast ecosystem. In web development, Python powers Django (batteries-included), Flask (micro-framework), and FastAPI (async, high-performance). Python is also the dominant language for data science, machine learning (PyTorch, TensorFlow), AI/ML, scripting and automation. Instagram, Spotify, Dropbox, Pinterest and Netflix use Python extensively. Python is ideal for teams that need rapid prototyping, data-heavy applications, ML/AI integration or a simple, readable codebase.

A high-level Python web framework that encourages rapid development and clean, pragmatic design.

Django is a high-level Python web framework that follows the 'batteries-included' philosophy. It provides an ORM, authentication, admin panel, URL routing, template engine, form handling and security features (CSRF, XSS, SQL injection) out of the box. Django is used by Instagram, Pinterest, Mozilla, NASA and The Washington Post. It is ideal for content-heavy sites, data-driven applications, REST APIs (Django REST Framework) and projects that need a robust admin interface. Django emphasizes the DRY (Don't Repeat Yourself) principle and rapid development.

A PHP web framework with expressive syntax, elegant tools and a rich ecosystem.

Laravel is a PHP web framework created by Taylor Otwell, known for its elegant syntax, expressive tools and rich ecosystem. It provides Eloquent ORM, Blade templating, Artisan CLI, migrations, queueing, broadcasting, Laravel Forge (deployment), Laravel Vapor (serverless) and Laravel Nova (admin panel). Laravel is the most popular PHP framework, used by Netflix, Apple, Toyota and millions of production applications. Laravel is ideal for teams that prefer PHP, need rapid development, elegant APIs and a mature ecosystem with extensive documentation.

A statically typed, compiled language by Google, designed for simplicity, performance and concurrency.

Go (Golang) is a statically typed, compiled programming language designed at Google by Robert Griesemer, Rob Pike and Ken Thompson. It is known for its simplicity, fast compilation, built-in concurrency (goroutines and channels) and excellent standard library. Go is widely used for microservices, APIs, CLI tools, cloud infrastructure (Docker, Kubernetes are written in Go) and high-performance backend systems. Web frameworks include Gin, Fiber, Echo and Chi. Go is ideal for projects that need high concurrency, low latency, simple deployment (single binary) and excellent performance.

A systems programming language focused on safety, speed and concurrency, with zero-cost abstractions.

Rust is a systems programming language focused on memory safety, speed and concurrency without garbage collection. It uses a borrow checker to prevent null pointer dereferences, buffer overflows and data races at compile time. Rust is consistently voted the 'most loved programming language' in Stack Overflow surveys. Web frameworks include Actix, Rocket, Axum and Warp. Rust is used by Discord, Figma, Dropbox, Cloudflare and is increasingly adopted for high-performance web backends, WebAssembly and performance-critical services. Rust is ideal for projects that need C++-level performance with memory safety guarantees.

A framework for building native mobile apps using React and JavaScript/TypeScript.

React Native is a framework developed by Meta for building native mobile applications using React. It allows developers to write once and deploy to iOS and Android with near-native performance. React Native uses a bridge to communicate with native UI components, supports hot reloading and has a massive ecosystem of community libraries. Expo is a popular managed workflow that simplifies React Native development. React Native is used by Instagram, Facebook, Shopify, Discord, Microsoft and thousands of production mobile apps. It is ideal for teams with React experience who want to ship to both platforms from a single codebase.

Google's UI toolkit for building natively compiled mobile, web and desktop applications from a single codebase.

Flutter is Google's open-source UI toolkit for building natively compiled applications for mobile (iOS, Android), web and desktop from a single codebase. It uses the Dart programming language and provides a rich set of customizable widgets, hot reload and excellent performance. Flutter compiles to native ARM code for mobile and JavaScript for web. It is used by Google, BMW, Toyota, Alibaba and eBay. Flutter is ideal for projects that need pixel-perfect UI, complex animations, custom designs and a single codebase across all platforms.

The world's most popular CMS, powering 43%+ of all websites on the internet.

WordPress is an open-source content management system (CMS) that powers 43%+ of all websites on the internet. It started as a blogging platform and evolved into a full CMS with a massive plugin ecosystem (60,000+ plugins), theme marketplace and developer community. WordPress supports both traditional (monolithic) and headless (decoupled via REST API or WPGraphQL) architectures. It is used by The New York Times, BBC, TechCrunch, Sony and millions of businesses. WordPress is ideal for content-heavy sites, blogs, e-commerce (WooCommerce) and projects where the editorial team needs a familiar, user-friendly admin interface.

An open-source headless CMS with a customizable admin panel and REST/GraphQL API.

Strapi is an open-source headless CMS that provides a customizable admin panel, content modeling tools and auto-generated REST and GraphQL APIs. It is built on Node.js and supports multiple databases (PostgreSQL, MySQL, MongoDB). Strapi is used by Walmart, IBM, Toyota, Shopify and thousands of companies. It is ideal for developers who want full control over their content model, API and frontend while giving editors a user-friendly admin interface. Strapi can be self-hosted or used with Strapi Cloud for managed hosting.

A query language for APIs that lets clients request exactly the data they need in a single request.

GraphQL is a query language for APIs and a runtime for fulfilling those queries, created by Facebook (Meta) in 2012 and open-sourced in 2015. Unlike REST, where the server defines the response structure, GraphQL lets the client request exactly the data it needs in a single request. This eliminates over-fetching and under-fetching. Key features include a strong type system, introspection, subscriptions (real-time), and a single endpoint. GraphQL is used by GitHub, Shopify, Twitter, Airbnb and Netflix. It is ideal for complex data requirements, mobile apps (bandwidth optimization) and microservices that need flexible data aggregation.

Google's three official user-experience metrics: LCP, INP, and CLS — a confirmed ranking factor.

Core Web Vitals (CWV) are Google's three official user-experience metrics: (1) Largest Contentful Paint (LCP) — how fast the main content loads, target < 2.5s; (2) Interaction to Next Paint (INP) — how fast the page responds to user interaction, target < 200ms; (3) Cumulative Layout Shift (CLS) — how visually stable the page is, target < 0.1. CWV is measured at the 75th percentile of real-user sessions (CrUX data) and is a confirmed Google ranking factor. Sites that pass CWV get a small but measurable ranking boost. Sites that fail get suppressed. Every site we ship at Sufyaan Studio passes CWV.

Core Web Vital that measures how fast the largest visible content element loads. Target: < 2.5s.

Largest Contentful Paint (LCP) is a Core Web Vital that measures how quickly the largest visible content element (typically the hero image or main heading) loads and renders. The target is < 2.5s at the 75th percentile of real-user sessions. Common LCP fixes: preload the LCP image in <head>, serve the LCP image in WebP/AVIF with responsive srcset, reduce server response time (TTFB) to < 600ms via edge functions or static generation, and eliminate render-blocking resources. Across our 40+ Next.js projects, average LCP is 1.2s; across WordPress projects, average LCP is 3.4s.

Core Web Vital that measures page responsiveness to user interactions. Target: < 200ms.

Interaction to Next Paint (INP) is a Core Web Vital that measures the latency between a user interaction (click, tap, key press) and the resulting visual update. INP replaced First Input Delay (FID) as a Core Web Vital in 2024. The target is < 200ms at the 75th percentile of real-user sessions. Common INP fixes: break up long tasks (> 50ms) using scheduler.yield(), defer non-critical JavaScript, debounce or throttle expensive event handlers (especially scroll and resize), use web workers for CPU-intensive tasks, and reduce JavaScript bundle size through code splitting and tree-shaking.

Core Web Vital that measures visual stability. Target: < 0.1.

Cumulative Layout Shift (CLS) is a Core Web Vital that measures how much the visible content shifts around during page load. The target is < 0.1 at the 75th percentile of real-user sessions. Common CLS causes: images without explicit width/height attributes, web fonts causing FOIT/FOUT (fix with font-display: optional or size-adjusted fallback fonts), dynamically injected content (banners, ads, cookie consent) that pushes existing content down, and animations that change layout (use transform and opacity, not width/height/margin). High-CLS sites have 2–3x higher bounce rates than low-CLS sites.

Google's automated audit tool for performance, SEO, accessibility, and best practices.

Lighthouse is Google's open-source automated audit tool that runs against any web page to measure Performance, SEO, Accessibility, and Best Practices on a 0–100 scale. Lighthouse is integrated into Chrome DevTools, available as a CLI, and powers PageSpeed Insights. The Performance score is synthetic (lab data) — it does not represent real-user experience as accurately as CrUX. For marketing sites we target 95+ Performance, 100 SEO, 95+ Accessibility, 100 Best Practices. For web apps we target 80+ Performance (due to JS overhead), 100 SEO, 90+ Accessibility, 100 Best Practices.

Rendering HTML on the server on each request, then sending it to the browser.

Server-Side Rendering (SSR) is a rendering strategy where the HTML for a page is generated on the server in response to each request, then sent to the browser. SSR is in contrast to Static Site Generation (SSG) and Client-Side Rendering (CSR). SSR is ideal for pages that need fresh data on every request (e.g., a dashboard, a personalized product page, a search results page) but still want fast initial page load and good SEO. Next.js supports SSR via getServerSideProps or React Server Components. SSR is essential for GEO: AI retrievers do not execute client-side JavaScript, so the content must be in the initial HTML response.

Pre-rendering HTML at build time for sub-second page loads and maximum crawl efficiency.

Static Site Generation (SSG) is a rendering strategy where HTML pages are pre-rendered at build time and served from a CDN. SSG produces the fastest possible page loads (often < 500ms LCP), the best Core Web Vitals, and the most efficient crawl budget for search engines. Next.js, Astro, Gatsby, and Hugo all support SSG. SSG is ideal for content that does not change on every request: marketing pages, blog posts, documentation, product listings. For dynamic content, SSG can be combined with Incremental Static Regeneration (ISR) to rebuild pages on a schedule or on-demand.

A backend-only content management system that exposes content via API for any frontend.

A headless CMS is a backend-only content management system that stores and manages content but does not dictate how it is presented. Content is exposed via a REST or GraphQL API and consumed by any frontend — a website, a mobile app, a digital sign, an AI assistant. Popular headless CMSs include Sanity, Contentful, Strapi, Hygraph, Storyblok, and Payload. The opposite of a headless CMS is a traditional CMS like WordPress, which couples content management with presentation. Headless CMSs are preferred for modern stacks (Next.js, Astro, SvelteKit) because they decouple content from code and let developers choose the best frontend for SEO and performance.

Using WordPress as a content backend with a custom frontend (Next.js, etc.) for speed and SEO.

Headless WordPress uses WordPress as a content management backend (the familiar WordPress admin UI for editors) while serving the content via REST or GraphQL API to a custom frontend (Next.js, Astro, Gatsby). The result: editors get the WordPress UX they know, developers get the speed and SEO of a modern framework. We use this architecture for content-heavy sites — news, blogs, large marketing sites — where the editorial team is non-technical and needs the WordPress admin. The frontend hits Lighthouse 95+ while the backend stays familiar.

Decoupling the storefront frontend from the e-commerce backend for full design and speed control.

Headless commerce is an architecture pattern where the storefront frontend is decoupled from the e-commerce backend. The backend (Shopify, BigCommerce, Magento, Medusa, Saleor) handles products, inventory, checkout, payments, and order management. The frontend is a custom web application (typically Next.js, Astro, or Remix) that consumes the backend's APIs (Storefront API for Shopify, GraphQL Storefront API for content). Headless commerce enables: sub-second page loads, full design control, omnichannel flexibility, better SEO, and AI citability. We build headless commerce with Shopify Hydrogen for brands doing $1M+ GMV.

Shopify's React-based framework for building custom headless storefronts.

Shopify Hydrogen is Shopify's official React-based framework for building custom, headless storefronts on top of the Shopify backend. It provides pre-built components, commerce logic, and Oxygen (Shopify's edge hosting for Hydrogen) out of the box. Hydrogen is built on Remix and supports SSR, SSG, and React Server Components. It is the recommended path for brands doing $1M+ GMV that need sub-second loads, full design control, omnichannel flexibility, and the SEO/CWV advantages of headless. We build headless Shopify with Hydrogen and deploy on Oxygen for our enterprise clients.