The Video Site That Quietly Became Infrastructure

In April 2005, a 19-second clip appeared on a barely-finished website. A young man stood in front of the elephant enclosure at the San Diego Zoo and observed, more or less, that elephants have long trunks. The video was called "Me at the zoo." The man was Jawed Karim, one of three former PayPal employees who had registered a domain that February. Two decades later, that clip still lives on the same platform - now watched by around 2.7 billion people a month, or roughly a third of everyone alive.

YouTube did not set out to replace television. Its founders - Karim, Chad Hurley, and Steve Chen - were solving a smaller, more human problem: it was 2005, and there was no easy way to put a video online and send it to a friend. Hurley, a designer, built the interface. Chen and Karim, both engineers, built the plumbing. What they shipped was less a media company than a piece of infrastructure - a place to upload, store, and press play. The media company grew on top of it, uploaded by everyone else.

01 / What it actually isA two-sided machine dressed as a website

Strip away the interface and YouTube is a marketplace with two sides that never meet. On one side are creators, who supply the content for free. On the other are viewers, who supply attention. YouTube sits in the middle, selling that attention to advertisers and handing back a majority share of the money to the people who made the videos. The Partner Program, introduced in 2007, formalized the deal: creators keep about 55 percent of the advertising revenue their videos earn, and YouTube keeps the rest.

That single split is arguably the most consequential decision in the company's history. It turned uploading into a job. A generation that might once have wanted to be on television instead grew up wanting a channel - and YouTube has said it has paid creators, artists, and media companies more than $100 billion over the past three years alone.

Google saw the shape of this early. In 2006, roughly a year after launch, it bought YouTube for $1.65 billion in stock - a sum that struck plenty of observers as a mistake for a site with no obvious path to profit and a mounting pile of copyright headaches. What Google was really buying was distribution and an ad business waiting to be attached. The site kept its name, its culture, and its San Bruno offices, and Google supplied the one thing a video platform burns through fastest: the money to store and stream video for anyone, anywhere, at no charge to the viewer.

02 / Who uses itAlmost everyone, for almost everything

The customer base is unusually broad because the product is unusually general. There are the viewers - billions of them, watching more than a billion hours of video a day. There are the creators, from teenagers filming in a bedroom to studios uploading polished series. There are musicians, whose catalogs power YouTube Music. There are educators, marketers, newsrooms, and gamers. And there are the advertisers and brands who pay to reach all of the above.

The scale creates its own problems. When 500 hours of video arrive every minute, no human team can screen it, and no rights-holder can police it by hand. YouTube's answer, in both cases, was to build a system rather than hire a department.

03 / The problems it solvesCopyright and chaos, handled by software

Two problems could have killed YouTube early: copyright and moderation. Instead of litigating each one, YouTube built Content ID - an automated system that scans every upload against a library of reference files supplied by rights holders. When it finds a match, the owner can block the video, track it, or, more often, choose to leave it up and collect the ad money. A record label's song in someone's vlog becomes revenue rather than a lawsuit. It is quietly one of the most important pieces of rights infrastructure on the internet.

Moderation is the harder, unfinished problem. Community guidelines, trust-and-safety teams, and a filtered YouTube Kids app all exist to keep a planetary-scale open platform usable - and it is a problem no platform of this size has fully solved. YouTube treats it as an ongoing operation, not a settled one.

04 / How it's differentIt owns the pipes, not the shows

Netflix commissions its catalog. Broadcast networks own theirs. YouTube owns almost none of it - and that is the difference. It supplies hosting, distribution, discovery, and payments, then lets millions of strangers fill the shelves. Against TikTok and Instagram Reels, it competes on economics as much as format; against Netflix and traditional TV, it competes on breadth and the fact that its content costs it nothing to produce.

The real moat is discovery. YouTube's most valuable asset is not any single video but the recommendation engine that decides what 2.7 billion people watch next. That engine is what turns an unnavigable ocean of uploads into a feed that feels, to each person, like it was built for them.

Mohan describes the whole operation as a flywheel, and the metaphor is unusually literal. Creators bring viewers. Viewers bring advertisers. Advertising money pays creators, which attracts better creators, which brings more viewers. Each turn makes the next one easier, and every part of the loop feeds the others. Competitors can copy the format of a short vertical video in a weekend; copying a flywheel that has been spinning for twenty years is a different kind of problem. The subscription business - Premium, Music, and YouTube TV - bolts a second, steadier engine onto the same wheel, so that a slow quarter for advertising no longer stalls the entire machine.

05 / Products and business modelFree on one side, paid on the other

The free, ad-supported platform is the foundation. On top of it sits a growing subscription business: YouTube Premium (no ads, background and offline play), YouTube Music, and YouTube TV, a live-television bundle that includes NFL Sunday Ticket and functions as a cable replacement. Subscriptions now make up roughly a third of YouTube's revenue - a deliberate hedge against depending on advertising alone.

Then there is Shorts, launched in 2020 as an answer to TikTok. The interesting part is not that YouTube copied the format but that it made the format pay: Shorts has reached parity in revenue per watch-hour with core YouTube in the United States, and in some countries it earns more. Newer AI tools - Dream Screen backgrounds, automatic dubbing across languages, generative editing aids - aim to widen what a single creator can produce.

06 / Expertise & where it fitsTwenty years of running video at scale

What YouTube knows better than almost anyone is how to move video reliably to billions of screens, match it against rights libraries in real time, recommend it well enough to hold attention, and pay for all of it through advertising. That is a narrow set of skills, practiced for twenty years, and it is very hard to replicate from scratch.

Where it fits in the market has shifted. YouTube began as a website, became a mobile app, and in early 2025 crossed a symbolic line: the television - the screen it was once expected to disrupt - became the most popular way to watch it. Google bought the company for $1.65 billion in 2006, a price critics called reckless. It now generates something close to $60 billion a year. In 2025, CEO Neal Mohan, the longtime product executive who took over in 2023, was named TIME's CEO of the Year.

• Founded2005, San Bruno, California
• FoundersChad Hurley, Steve Chen, Jawed Karim
• OwnerGoogle (Alphabet), acquired 2006 for $1.65B
• CEONeal Mohan (since 2023)
• Revenue~$60B in 2025 (~$40B advertising)
• CompetitorsTikTok, Reels, Netflix, Spotify, Twitch, cable TV

The first video was one person, one camera, and 19 seconds about elephants. The system underneath it now carries a third of humanity's video habit, pays out more than any studio, and increasingly plays on the same television set it was supposed to make obsolete. Television did not die. It moved.