How Y Combinator was built: origin stories and founding DNA

Executive overview

In 2005, early-stage funding barely existed — VCs wanted business plans and millions deployed, and angels were scarce. Jessica Livingston and Paul Graham filled that gap by creating a small, standardised seed vehicle designed to help technical founders incorporate, test an idea, and survive long enough to raise real money.

The batch model, weekly dinners, and free events weren't afterthoughts — they were the core product. The same structure that ran in summer 2005 is still running today.

The insight: treating founders as people worth believing in — before they've proven anything — is itself a competitive advantage.

Why YC was founded

• VCs in 2005 required tested ideas, business plans, and million-dollar minimums.
• Angels existed but were rare; no small seed funds existed.
• Graham and Livingston wanted a vehicle that wrote small checks so founders could quit their jobs and test an idea before raising VC.
• The proximate trigger: a Boston VC dragged its feet for 6+ weeks, giving Graham time to propose they just start their own thing.
• Early name was "Cambridge Seed" — dropped quickly because they didn't want to be geographically limited.

Standardising the mechanics

• Graham's pain point from ViaWeb: legal paperwork nearly stopped the company before it started.
• YC created fill-in-the-blank incorporation docs and a standard investment agreement — unprecedented in 2005.
• Standard deal terms eliminated back-and-forth legal negotiation for every company.
• Goal was to make it as easy as possible for technical people to start companies — "not very far from programming to founding."

The batch model and why it worked

• Originally planned to fund companies asynchronously, like traditional angels.
• The batch idea emerged as a summer experiment to learn how to be angel investors.
• Funding companies simultaneously let founders compare progress with peers for the first time.
• Peer pressure from weekly dinners drove output: founders didn't want to show up without new things to demo.
• First batch of eight companies included Sam Altman, Reddit (Huffman and Ohanian), Twitch (Kahn and Shearer), and Aaron Swartz.

Events as core infrastructure

• Livingston had an event-planning background; every YC touchpoint — interviews, dinners, demo days — was treated as an event.
• Graham had run a scrappy anti-spam conference that modelled the format: cheap venue, stick-on name tags, excellent speakers, no sponsors.
• VCs held one or two portfolio events per year; YC held weekly dinners with every company in the batch.
• Events built a community that compounded: alumni returned to advise new batches, creating a flywheel of knowledge and connections.
• The rule: any time you can bring founders together, good things happen.

Startup School: getting more people to start companies

• First Startup School was held at Harvard, partnering with an undergraduate programmer group to avoid venue costs.
• Entry was free — unusual at the time when every other conference charged.
• No sponsors: Graham and Livingston had an "allergic reaction" to opening with sponsor acknowledgements.
• Speakers gave unfiltered, candid talks because there was no press and no Twitter.
• Audience was self-selected serious builders, not networkers or scenesters — that tone matched YC's DNA.

The Yuri Milner moment

• Around 2011, Russian investor Yuri Milner — introduced via Ron Conway — showed up wanting to fund every startup in the current batch, unconditionally, with $150,000 uncapped.
• Graham tried to avoid the meeting; Harj Taggar took it and had to drag Graham in when he realised Milner was serious.
• The announcement was made at a surprise dinner; founders speculated Obama or Steve Jobs was coming.
• Milner appeared on a telepresence robot (Trevor Blackwell's Anybot). The room erupted.
• The move gave every company instant confidence and runway going into demo day.
• Investors in the Start Fund (which followed) caught Coinbase, DoorDash, and Instacart at entry.
• Funding the whole batch, in hindsight, outperforms most VC funds by returns.

How YC went from underdog to institution

• The shift was slow and organic — no single inflection point.
• Portfolio companies becoming household names (Dropbox, Reddit, Airbnb, Justin.tv) drove legitimacy.
• Speakers invited to dinner left impressed, spreading word of mouth.
• The Social Network (2010) correlated with a visible jump in YC application volume.
• With visibility came distraction: haters, inbound coffee-meeting requests, people seeking attention.
• YC's response was to stay focused, say no to almost everything, and refuse to act like a big company.

What successful founders share

• Smart, curious, independent-minded, and determined — traits consistent from early 20s through success.
• Unconventional by default: in 2005, starting a startup was not a normal post-college path.
• Confident in the idea, if not always in themselves as people.
• Success brings more confidence and more people saying yes — but the core doesn't change much.
• Sam Altman at 19: responded to Graham's suggestion to wait a year with "I'm a sophomore and I'm coming."
• Early Reddit founders were normal 21-year-olds doing dumb things — yet they built something.
• The pattern: founders pivot ideas when users demand it (PayPal started as Palm Pilot encryption).

What has and hasn't changed

• The core YC structure — batch, weekly dinners, demo day — is identical from summer 2005 to summer 2024.
• Scale, brand, knowledge base, and deal size ($10–15K then; $500K now) have all grown.
• The ethos hasn't: earnest, non-pretentious, founder-first, no pedigree required.
• YC created a community where people who felt like outsiders found their people.
• The tree of prosperity: the goal was always to cause more startups to exist, not just to generate returns.