Platform Economics
Why Uber Had to Lose Money on Every Ride to Win Everything
The most powerful technology businesses in the world are not really in the business of selling you anything — they are in the business of making two strangers trust each other.
The Idea
A two-sided market is a platform that serves two distinct groups simultaneously, and here is the strange part: the value it creates depends entirely on both sides showing up at once. This is not just a marketing challenge — it is a structural feature with deep economic consequences. Think of it like a telephone. One telephone is useless. Two telephones is one conversation. A million telephones is a civilisation-changing network. The value of joining grows as others join — economists call this a network effect — but in a two-sided market, there are two separate groups whose participation has to be bootstrapped at the same time. Drivers and riders. Buyers and sellers. Advertisers and eyeballs. Each side only shows up if the other is already there. This creates what is called the cold-start problem: before either side has a reason to join, neither will. Platforms solve this in different ways — sometimes by subsidising one side heavily to attract the other, sometimes by pretending to be something else entirely in early days. The side you charge and the side you subsidise is a strategic choice that shapes the entire business. Often, the subsidised side is the scarcer, harder-to-attract group. What makes two-sided markets so competitively brutal is that once a platform achieves critical mass on both sides, it becomes extremely difficult to dislodge. You cannot compete with a platform that already has the drivers by building a better app — you have to simultaneously recruit drivers away and convince riders to switch, which is almost impossible without enormous capital. Winner-takes-most dynamics are not accidental; they are baked into the structure.
In the World
In 2014 and 2015, Uber was losing staggering sums of money in city after city — not through mismanagement, but by design. In markets like China and Southeast Asia, it was offering rides at below cost and paying drivers bonuses that bore no resemblance to what the market would sustain. Observers called it irrational. It was anything but. Uber's leadership understood that the asset they were building was not software or brand recognition — it was liquidity. A rider who opens an app and gets a car in three minutes will never switch to a competitor where wait times are eight minutes. A driver who earns reliably on a platform with dense demand will not risk their income for a new entrant with empty streets. Each successful ride made the next one more likely, tightened the circle, and raised the barrier to entry for anyone who came after. The same logic explains why Google Search has always been free to users. Search is a two-sided market between people with questions and advertisers with answers. Google subsidises the reader completely because the reader's attention is what it sells. The reader is not the customer; the reader is the supply. Once Google had billions of readers, it had built an asset no competitor could replicate without first acquiring those same billions — and those readers had no reason to leave a product that cost them nothing and worked extremely well. In both cases, the economics that look strange in year one are the strategy. The losses are the moat being dug.
Why It Matters
Understanding two-sided markets changes how you read almost every major tech story. When a platform offers something for free, the instinct is to feel lucky. The more useful instinct is to ask: which side of this market am I on, and what is the other side paying for my presence? It also reframes how you think about regulation and competition. When politicians argue that breaking up a dominant platform would increase competition, they are sometimes right — but in a two-sided market, fragmentation can destroy the very value that made the platform worth using. A world with five smaller ride-hailing apps might mean longer wait times and thinner driver earnings across all of them. The monopoly critique and the network-value defence are both partially true, which is what makes this policy territory genuinely hard. Finally, it sharpens your instincts as a professional or entrepreneur. If you are building something that requires two groups to show up simultaneously, your first and most important question is not what to build — it is which side to solve for first, and how to fake critical mass long enough for the real thing to arrive.
A Question to Ponder
In the platforms you use daily, which side of the market are you actually on — and what would have to be true for you to be worth less to that platform than you are today?
Get a new one of these every morning.
Start learning with Thinkable