The Future of Web Startups
There's something interesting happening right now. Startups are undergoing the same transformation that technology does when it becomes cheaper. — There's a pattern that we see over and over in technology. Initially there's some kind of device that's …
Context & Ripple Effects
Paul Graham's essay arrives a year after O'Reilly Radar's look at where startup activity clusters, which framed the question geographically — which cities attract founders. Graham reframes it economically: the act of starting a web company is following the same cost curve technology itself always has, getting cheaper until it transforms.
That framing matters because it attacks the assumption that founding a company is inherently scarce or expensive. If incorporation, hosting, and reaching users keep dropping in cost, the bottleneck moves from building a product to judging which of the many cheaply-built products deserves capital and attention.
First-order effects
- Founders gain leverage: smaller teams can launch real products with little capital, so the population of people who can plausibly start a company expands well past those with VC connections.
- Early-stage investors face a changed intake problem — more candidates at lower cost each, forcing selection to happen earlier and in higher volume than traditional funding was built for.
Second-order effects
- As launching gets cheap, me-too products multiply and compete away each other's returns, pushing the value in web startups toward whatever cannot be replicated for free — audience, brand, and network position.
- Intermediaries that batch-select and train founders, the model behind the startup-center phenomenon, become more valuable precisely because filtering thousands of cheap launches is the new scarce function.
Third-order effects
- If the cost curve keeps falling, startup formation industrializes: standardized pipelines for creating, funding, and graduating companies, with geography mattering less than access to the selection layer.
- The definition of a startup drifts toward a commodity input — and the durable advantages concentrate in whoever owns distribution and judgment rather than whoever writes the code.
The trend: Startup formation is riding technology's own cost curve downward, shifting scarcity from building products to selecting and scaling the right ones.