We don't ask founders "what is your revenue?" at the first meeting. We ask: can you reach $10M ARR within 36 months of launch?
Not "will you" — founders in YC are optimistic by definition, and the answer is always yes. We're asking whether they can see how. Whether they can describe the sales motion, the customer segment, and the growth rate that gets them from where they are to $10M ARR in three years. The difference between a founder who has thought through this and one who hasn't is visible in the first ninety seconds.
Why this number and this timeline
$10M ARR at a 10x revenue multiple is a $100M valuation — the floor where a Series B becomes a real conversation and the unicorn path becomes mathematically plausible. 36 months is what's achievable given the portfolio's median annual revenue growth of approximately 350%. Starting from any traction, that growth rate reaches $10M ARR in roughly 36 months.
We're not asking for a forecast. We're asking for a mechanism. The forecast will be wrong. The mechanism tells you whether the founder understands their business.
What a good answer sounds like
A good answer is specific: "We're selling to mid-market financial services at $120K ACV. We close 2 deals a month at current velocity. We need 8 deals a month by month 24. Here's how: two more AEs, a VP of Sales by month 12, two channel partners sharing our ICP." That's a plan.
A bad answer is general: "We're going after a massive market and think we can get to $10M ARR if we execute." That's a wish. The founders who can't articulate the mechanism almost never hit the number. The plan reveals the understanding.
When the filter doesn't apply directly
Hardware companies have supply chain constraints. Marketplace companies report revenue as a take rate on GMV. Regulated industries have longer sales cycles. The filter adjusts: we look for a plausible path given the structural constraints, not a rigid 36-month ARR number. The question is always the same — can you describe the mechanism?
The $10M ARR in 36 months question isn't a forecast request. It's a mechanism test. Founders who can answer it have thought through the business. Founders who can't have a problem that isn't a metrics problem.