The simple version

Websites, Explained Simply

No jargon needed. A website works a lot like a restaurant. Once you see the restaurant, you understand the website.

One sentence: a website is a set of files on a computer somewhere. When you type its name, your browser finds that computer, asks for the files and turns them into the page you see.

01

A website is a restaurant

Every part of a website has a match in a restaurant. The small label is the restaurant version, and the bold name is what builders call it.

The restaurant's name

Domain name example.com

What people type to find you. You rent it for about $10–15 a year.

The maps app

DNS

Turns the name into an exact address so your browser knows where to go.

The building

Hosting / server

The computer, always switched on, where your website's files live.

Branches in every city

CDN

Copies of your site around the world, so everyone is served from somewhere close.

The menu's words

HTML

What's on the page: headings, paragraphs, pictures, links. The content itself.

The decor and plating

CSS

How it looks: colors, fonts, spacing, and how it rearranges on a phone.

The waiter

JavaScript

Reacts to you: opens menus, checks forms, loads more without reloading the page.

The kitchen and pantry

Backend & database

Where orders are handled and records are kept: accounts, messages, purchases. Simple sites don't need one.

The sealed delivery bag

HTTPS 🔒

Nobody can read or tamper with what travels between you and the site. Free and automatic on most hosts today.

The guidebook listing

SEO

Getting listed well in Google, with a clear name and a short description, so people pick you.

A friend's recommendation

GEO

When people ask ChatGPT or Claude, the AI recommends places it has read about. It can only repeat what your site says clearly.

The renovation plans

GitHub

A saved history of every change, so you can always undo. Pushing a change there can update the live site automatically.

02

What happens when you visit

All of this takes about a second.

  1. You type the nameexample.com
  2. Look up the addressDNS finds which computer has the site
  3. Ask for the pageOver a locked, private connection
  4. Get the files backHTML, CSS, JavaScript, pictures
  5. Draw the pageThe browser puts it all together on your screen

Want the detailed version? See the animated journey, which shows every step.

03

How Google and AI find you

Before most people find your website, a robot has already read it.

1

Robots visit. Google's robot, and now AI robots, regularly read websites the way you would, only much faster.

2

They read the labels. Every page has hidden labels: a title, a one-line description and facts like "this is a person, this is their job". Visitors never see them, but robots read them first.

3

They decide who to show. Clear pages that load fast, work on phones and that other sites link to get shown first.

4

AI repeats what it read. When someone asks an AI assistant about you, it answers from what it found. If your site is vague, the answer is vague too.

The takeaway: SEO and GEO aren't tricks you add later. They start with a website that is clear, fast and honest about what it is.

04

A short history of the web

From a two-letter message between universities to AI that answers questions for you. Each step added one of the pieces above. Every event links to its sources.

  1. The roads get built

    1969–1988 · Before the web, there was the internet: computers learning to talk to each other.

  2. ARPANET sends its first message

    UCLA tries to send "LOGIN" to Stanford Research Institute. The system crashes after "LO". The internet's first word is an accident.

    Sources: ARPANET

  3. TCP/IP is designed

    Vint Cerf and Bob Kahn describe a common language that lets any network connect to any other. ARPANET switches to it on January 1, 1983, often called the internet's birthday.

    Sources: Cerf & Kahn, 1974 paper · RFC 801: NCP/TCP transition plan

  4. DNS: names instead of numbers

    Paul Mockapetris designs the Domain Name System, so people can remember names instead of numbers. In 1985, symbolics.com becomes the first .com.

    DNS

    Sources: RFC 882 (Mockapetris, 1983) · Symbolics: first .com

  5. The web is born

    1989–1994 · Pages you can link together and anyone can read.

  6. Tim Berners-Lee proposes the World Wide Web

    At CERN, the physics lab in Switzerland, he proposes linked documents researchers can share. His boss writes "vague but exciting" on the proposal.

    Sources: CERN: Tim Berners-Lee's proposal · The 1989 proposal text

  7. The first website

    Berners-Lee builds the first browser, the first server, HTML and HTTP. info.cern.ch, a page explaining the web itself, goes live. You can still visit it.

    HTML HTTP

    Sources: info.cern.ch: home of the first website · History of the World Wide Web

  8. The web becomes free, and gets pictures

    CERN releases the web to the public domain, free for anyone forever. The Mosaic browser shows images inside pages, and the web starts to look like magazines rather than manuals.

    Sources: CERN: 30 years of a free and open Web · NCSA Mosaic

  9. Netscape, cookies and robots.txt

    Netscape Navigator brings the web to millions. Cookies let sites remember you. robots.txt lets sites tell crawlers where not to go. The W3C forms to set web standards.

    robots.txt

    Sources: Netscape Navigator · HTTP cookie · The 1994 robots.txt standard · W3C history

  10. Style, scripts and search

    1995–2003 · The web learns to look good, react, and be searched.

  11. JavaScript, built in 10 days

    Brendan Eich writes the first version of JavaScript at Netscape in about ten days. Encryption (SSL, later TLS) arrives the same year, making online shopping possible, and Amazon and eBay launch.

    JavaScript HTTPS

    Sources: Brendan Eich · TLS and SSL history · Amazon · eBay

  12. CSS separates looks from content

    CSS becomes a standard, so design can live in its own file instead of being tangled into every page.

    CSS

    Sources: W3C CSS1 Recommendation, Dec 1996

  13. Google, and links as votes

    Google ranks pages by how many other pages link to them (PageRank). Search engine optimization, SEO, becomes a profession.

    Backlinks

    Sources: Brin & Page: The Anatomy of a Large-Scale Hypertextual Web Search Engine

  14. The first big CDN

    Akamai starts copying websites to servers around the world so they load fast everywhere.

    CDN

    Sources: Akamai Technologies

  15. Apps and the cloud

    2004–2009 · Websites start behaving like software.

  16. "Web 2.0" and Gmail

    Gmail and, in 2005, Google Maps update without reloading the whole page. People start creating content, not only reading it.

    fetch()

    Sources: O'Reilly: What Is Web 2.0 · Gmail · Google Maps

  17. The cloud and sitemaps

    Amazon Web Services rents out servers by the hour, so you no longer need to own one. Google, Yahoo and Microsoft agree on one sitemap format.

    Sitemap

    Sources: AWS: Announcing Amazon S3, 2006 · sitemaps.org

  18. The iPhone

    The full web fits in a pocket, and sites now have to work on small touch screens.

    Sources: Apple: Apple Reinvents the Phone with iPhone

  19. The canonical tag

    Search engines introduce rel=canonical, so a page reachable at several addresses can name one official URL.

    Canonical

    Sources: Google: Specify your canonical, 2009

  20. Mobile, social and secure

    2010–2019 · Phones take over, and machines learn to read pages better.

  21. Responsive design and Open Graph

    Ethan Marcotte coins "responsive web design": one site that reshapes itself for every screen. Facebook's Open Graph tags control how shared links look.

    Responsive Open Graph

    Sources: Ethan Marcotte: Responsive Web Design · The Open Graph protocol

  22. schema.org

    Google, Bing and Yahoo agree on one vocabulary for facts in a page ("this is a person", "this is a recipe"). Today it's the structured data that AI also reads.

    Structured data

    Sources: Google: Introducing schema.org, 2011

  23. React

    Facebook releases React, and building big interactive apps in the browser becomes mainstream.

    Framework

    Sources: React (software) · React blog: Why did we build React?

  24. Free padlocks for everyone

    Let's Encrypt gives out free HTTPS certificates, and HTTP/2 speeds up loading. Google starts favoring mobile-friendly pages.

    HTTPS

    Sources: Let's Encrypt: Entering Public Beta · RFC 7540: HTTP/2 · Google: mobile-friendly search results

  25. "Not secure"

    Chrome labels every site without HTTPS "Not secure". Google begins indexing the mobile version of sites first.

    Sources: Chromium: A secure web is here to stay · Google: Rolling out mobile-first indexing

  26. Speed and AI

    2020–today · Search stops being a list of links and starts answering.

  27. Core Web Vitals

    Google defines what "feels fast" means, in three measurements, and starts using them for ranking in 2021.

    Core Web Vitals

    Sources: Chromium: Introducing Web Vitals · Google: Timing for page experience

  28. ChatGPT

    People start asking AI instead of searching. HTTP/3 also becomes a standard the same year.

    Sources: OpenAI: Introducing ChatGPT · RFC 9114: HTTP/3

  29. "GEO" gets a name

    Researchers publish a paper on Generative Engine Optimization: how to get content cited in AI-written answers.

    GEO

    Sources: Aggarwal et al.: GEO: Generative Engine Optimization

  30. AI Overviews and llms.txt

    Google puts AI-written summaries above its search results. llms.txt is proposed as a plain summary of a site written for language models.

    llms.txt

    Sources: Google: generative AI in Search, May 2024 · llmstxt.org

  31. Back to the foundations

    Robots now read your site on behalf of people. What they reward is what the web has always rewarded: clear content, honest labels, fast pages and links from others.

Ready for more detail? The full Anatomy of a Website has the animated request journey, the layer stack and 64 encyclopedia entries.