How Do You Evaluate Nothing?

You're sitting across from a founder who has: a deck, a demo that half-works, maybe some letters of intent that aren't binding, and a story about why this problem desperately needs solving. No revenue. No users. Sometimes not even a complete product.

How do you decide if this is the next Stripe or the next cautionary tale?

When you can't underwrite revenue multiples or customer acquisition costs, VCs fall back on proxies. Here's what actually gets scrutinized:

The Founder's Obsession Level: You can teach someone to build a business. You can't teach them to care enough to survive the next five years.

The tell isn't passion; everyone's passionate in the pitch meeting. It's specificity. Do they know their space with unsettling detail? Can they explain the problem's nuances without checking notes? Have they been thinking about this for years, or did they just spot a "market opportunity"?

Best signal: they've already been working on this longer than makes rational sense. They pivoted into this space, or they've been adjacent to the problem their whole career. Either way, this isn't a side project that sounded interesting; it's the thing they can't stop thinking about.

The Learning Velocity: Six months into building, what's different about this founder?

The best ones have visible growth—they've taught themselves technical skills, developed industry relationships, learned to sell, or absorbed domain knowledge at an unnatural pace. The pitch evolved from "we're building X" to a nuanced understanding of why X is harder than it looks and exactly which version of X will work.

Red flag: the pitch from month one sounds identical to month twelve. Nothing learned, no pivots considered, no intellectual honesty about what's not working. That's not conviction, that's inability to process feedback.

The 5-Year Earning Curve of Capital: This is where most founders fumble. They know what the money does in 12 months: hire engineers, build a product, maybe run some ads. They get vague past that.

VCs are writing a check that locks up capital for 5-10 years. The question isn't "what can you build in a year?" It's "what compounding advantage does this capital create that makes you defensible by year five?"

Smart founders explain the progression: Year 1 is product-market fit. Year 2 is the initial scale proving unit economics. Year 3 is when network effects or data moats start kicking in. Year 4 is when you're hard to compete with. Year 5 is when you're the category winner, or you've failed definitively.

If the answer is "we'll be profitable and growing" without explaining the compounding mechanism, you haven't thought about it hard enough.

Here's what makes early-stage venture capital different from every other investment: you're not buying a company. You're buying a decade of someone's life, and betting they'll become the person who can build something valuable.

The capital is locked up. No quarterly earnings. No liquidity events for years. The only thing that changes the outcome is how much this specific founder can grow into the role.

The Pattern Matching Trap

Every breakthrough looks like a bad idea at first. Airbnb was "stranger danger meets hospitality." Uber was "illegal taxi service via app." DoorDash was "Uber for food" in a world where Uber for X was already a punchline.

VCs who rely too heavily on "this looks like that successful thing" miss the genuinely novel. VCs who ignore patterns entirely fund nonsense.

The balance: use patterns to identify red flags (has this exact idea failed three times already?), but stay open to the idea that context has changed enough to flip the outcome.

The real skill in evaluating nothing isn't finding perfect signals; they don't exist. It's identifying founders who will compound learning faster than their competitors, and backing people obsessive enough to still be building when everyone else has moved on.

Most early-stage investing is buying options on people, not companies. The company doesn't exist yet. You're betting on who this founder becomes over the next decade.

That's the only metric that matters when you're evaluating nothing.