Facebook vs Twitter — The Graph vs The Stream

The most important architectural decision in the history of social media happened in a dorm room in 2004 when Mark Zuckerberg decided that Facebook's fundamental unit would be the connection between two real people. Not an anonymous handle, not a topic feed — a mutual friendship, verified by both parties, anchored to a real identity. That graph, once built to scale, would prove to be one of the most commercially valuable datasets ever assembled: who you know, what events you attend, when you graduate or marry or move, what you buy. It compounded into a $1.4 trillion business.

Twitter, born two years later, made a different call. Its fundamental unit was the public broadcast — anyone could follow anyone, real name optional, without needing the follow returned. That asymmetric model was revolutionary. It made Twitter the world's most influential real-time information network: the place where breaking news broke first, where markets moved on central banker tweets, where political careers were made and ended in 280 characters. In cultural weight, Twitter punched far above its size. In commercial weight, it never translated that influence into revenue.

The gap in the numbers is vertiginous. Meta generated approximately $165 billion in revenue in 2024. Twitter, before Elon Musk's acquisition, was generating around $5 billion and struggling to grow it. Facebook's average revenue per user in the United States exceeded $230 annually — the kind of ARPU that comes from deep demographic data and high-intent targeting. Twitter's ARPU was roughly $8 globally. The social graph generates purchase intent; the interest graph generates opinions. Advertisers pay dramatically more for purchase intent.

Musk's $44 billion acquisition in 2022 is now one of the most studied overpays in corporate history. He bought the most influential public platform in the world at a price that assumed either significant revenue growth or cost-cutting could sustain the valuation. Revenue fell as advertisers pulled back on safety concerns, staff cuts degraded the product, and competing platforms absorbed attention. By 2024-25, analyst estimates put X's value at $12-20 billion — the $44 billion evaporated not because the interest graph has no value, but because Twitter's business model had never figured out how to capture the full commercial value of what it built. The lesson is one of the sharpest in all of tech: cultural power and commercial power are two entirely different things, and the architecture you choose at founding often determines which one you end up with.