How early Facebook and Dropbox shaped one founder's approach to risk and growth

Executive overview

Most people wait until they feel qualified before taking on hard problems. Aditya Agarwal learned at Facebook that the gap between "I don't know how" and "I built it" is smaller than it looks.

That shift — from deferring to experts to just building — shaped how he led engineering at Dropbox and how he now runs South Park Commons, a pre-idea venture community.

The only thing that remains constant in hyper growth is change.

Joining Facebook before knowing what Facebook was

• Met Mark Zuckerberg through a mutual friend; the conversation lasted 15–20 minutes
• Had never used the product — Facebook launched at Carnegie Mellon after he graduated
• Chose to join because of the people, not the idea: driven, ambitious, willing to iterate
• Parents called it crazy; he framed it as: if not now at 23, when?
• Key early lesson: you can't evaluate an early-stage company on idea clarity alone — bet on the people

The search engine moment that reframed self-belief

• Mark assigned him to build Facebook's search engine three months in
• First reaction: hire someone from Google or Yahoo who actually knows search
• Mark's reply: "Yeah, we could, or you could just write it"
• Realisation: he hadn't known how to build Facebook either — until he built it
• Founders inevitably do finance, sales, marketing without prior expertise; figuring it out is the job
• Discounting your own experience is a mistake — contextual knowledge often outweighs formal expertise

Facebook vs Dropbox: two cultures, one lesson

• Facebook's mantra: move fast and break things — viable because failures were low-stakes (a missed photo upload)
• Dropbox's standard: ultra reliable, ultra low latency, ultra secure — a hard drive that fails once per thousand reads feels broken
• Same underlying values (great technology, strong product, great design), but product risk tolerance diverged completely
• Lesson: don't copy Google's or Stripe's culture; find the values that fit your mission and product

The CTO role at hyper-growth scale

• North Star: good technical decisions + maximal team productivity
• The job changed every six months as the team grew from 12 to 1,500–2,000 engineers
• Danger signal: doing the same things you did a year ago
• If you think you have the role figured out, you're probably already behind

South Park Commons: the negative-one-to-zero phase

• Most accelerators are built to accelerate — SPC is built to slow down first
• Negative one to zero: the phase before you know which mountain to climb; zero is the base camp
• Previously, founders were expected to figure out their idea alone at home; SPC gives them a place and community to do it together
• Three supports SPC provides:
  1. A physical home (San Francisco and New York) to reduce disorientation after leaving a company
  2. A curated peer community of smart, driven people in the same phase — reduces isolation, improves decision quality
  3. Programming to bootstrap creativity: turning a blank page into ten ideas, nudging toward that first line of code or first customer conversation
• Partners give time equally to funded and unfunded members
• Category creation is the hardest type of startup; SPC actively pushes founders toward bigger risks, not safer ones
• Hardest part of the job: convincing a founder to throw away a good idea in pursuit of a great one