Platform Economics
Why the Second-Best Search Engine Is Worth Almost Nothing
In most markets, coming second is a viable business; in platform markets, it is basically a different industry.
The Idea
Most markets tolerate a range of competitors. The third-best hotel in a city still fills rooms. The second-best accounting firm still wins clients. Platforms are structurally different, and the reason comes down to a single mechanism: network effects. A platform becomes more valuable to each user as more users join it. This sounds benign — even pleasant — but the compounding logic is brutal. Once a platform crosses a critical mass of users, every rival faces a market that is shrinking not because of price or quality, but because the leading platform's sheer size is itself the product. You join the network your contacts are already on. You search the engine that has indexed the most behaviour. You sell on the marketplace where the buyers already are. The result is that winner-take-all dynamics emerge not from monopolistic scheming but from the basic geometry of networks. What makes this counterintuitive is that the winner rarely wins on quality alone — or even primarily. Timing, luck, and which early adopters happened to join first can tip an otherwise identical competition decisively. This means winner-take-all markets systematically reward being early and being big over being better. Once that feedback loop is established — more users attracting more users — the window for a challenger to compete on product merit alone closes rapidly. Understanding this helps explain why so many tech markets look, from the outside, like they should be more competitive than they are.
In the World
In the late 1990s, AltaVista was the search engine. It was fast, technically sophisticated, and widely used. Google launched in 1998 and was, by most measures, a better product — but "better" alone rarely determines these outcomes. What Google understood, and relentlessly exploited, was that every search query was a data point that could train its ranking algorithms. More searches meant better results, which attracted more users, which generated more searches. Within a few years, this feedback loop had created a moat that had nothing to do with the original algorithm and everything to do with accumulated scale. AltaVista was eventually sold for a fraction of what it had been worth at its peak and quietly shut down in 2013. But here is the part worth sitting with: Yahoo had the chance to buy Google in 2002 for a relatively modest sum. They passed, reportedly because they wanted to remain a portal — a destination — rather than a utility people passed through. In a winner-take-all market, that kind of categorical misreading of what you are actually competing over can be terminal. Yahoo later tried to buy Google again, and couldn't. The window had closed. The network had compounded beyond reach, and the second-best search engine had become, functionally, irrelevant.
Why It Matters
This isn't just a story about tech companies — it reframes how you think about competition, regulation, and even your own choices as a user. When a government regulator asks whether a dominant platform is harming consumers through price, they are often asking the wrong question. Many of these platforms are free. The harm is structural: the winner-take-all dynamic itself narrows the ecosystem, reduces the incentive to innovate after dominance is achieved, and makes it nearly impossible for alternatives to gain traction regardless of quality. As a user, knowing this changes how you interpret your own inertia. Staying on a platform because "everyone is there" isn't just habit — you are actively contributing to the network effect that makes leaving harder for everyone else. And for anyone thinking about where technology markets are headed — whether in AI tools, cloud infrastructure, or social media — the single most useful question to ask isn't who has the best product today, but who is accumulating the feedback loop fastest.
A Question to Ponder
If being early and big can matter more than being good, what does that imply about how we should evaluate the dominant platforms we use every day — and whether the alternatives we've never tried might actually be better?
Get a new one of these every morning.
Start learning with Thinkable