Secrets of Sand Hill Road: Venture Capital and How to Get It

By Scott Kupor & Eric Ries

11 min read

Why does it matter? Because the rules of this game were written for insiders — and you're signing a 10-year contract.

The person across that table has negotiated hundreds of term sheets. You will negotiate two or three in your lifetime, if you're lucky. That gap — in pattern recognition, in knowing which clauses actually bite and which your lawyer waved through as "standard" — is where fortunes quietly disappear. Venture capital looks like judgment from the outside. From the inside, it's a set of interlocking financial incentives, each negotiated long before you ever walked in the room, each governing exactly how much pressure your board member is under to push you toward a sale, and why that VC stopped returning calls. None of it is personal. All of it is legible — once someone finally shows you the machinery.

You're Negotiating a Decade-Long Partnership With Someone Who Has Played This Game Hundreds of Times

A couple in a Charles Schwab commercial interrogates their contractor on cedar versus synthetic wood, debates school options with the precision of a litigator, and grills a car salesman on the difference between 467 and 423 horsepower. Then they sit across a mahogany desk from a financial advisor who says, "I think we should move you into our new fund." Blank looks. A pause. They nod. Scott Kupor, managing partner at Andreessen Horowitz, uses this scene to name what founders routinely do with VCs — abdicate every instinct for scrutiny they'd apply to any other significant decision.

The gap is structural. A venture capitalist evaluates hundreds of deals across a career, sits on dozens of boards, and has negotiated the same term sheet provisions so many times each clause is muscle memory. The typical founder does this once, maybe twice. That asymmetry costs founders more than price: the documents signed at the first check determine how power gets distributed and what happens when growth misses and the company needs more money.

The timeline sharpens the problem. When you take a first check from a VC, you're entering a relationship that typically runs a decade or longer — more than many marriages, more than most founders' relationships with their early employees. Yet founders routinely sign without asking what happens if the company stalls, or who controls the board when things get difficult. The information gap was tolerable when founders had no access to this world. Kupor's argument is that they do, once they learn the rules the other side has always known.

VCs Celebrate Losing Money on Half Their Investments — Because Batting Average Is the Wrong Metric

The venture capitalist who tries hardest to protect capital will consistently underperform the one who accepts losing money on most investments. This sounds perverse until you see the math.

Kupor walks through the actual distribution of returns inside a typical VC fund: roughly half of all investments are "impaired," meaning they lose most or all of the money. Another 20 to 30 percent are what he calls singles and doubles, returning 2 or 3 times what went in, which sounds fine until you account for the 50 percent that evaporated. Run the blended math and 70 to 80 percent of a fund's capital has generated something like 75 to 90 cents on the dollar. That's not a fund. That's a problem.

What saves it — what defines the entire business — is the remaining 10 to 20 percent: investments that return 10 to 100 times the original amount. In VC, these are called home runs, and they don't just pad the returns; they are the returns.

Accel Partners illustrates this with arithmetic that's almost too clean to believe. When Accel invested in an early round of Facebook, the company was valued at roughly $100 million. By the time Facebook went public, its market cap was around $100 billion, a return of roughly 1,000 times the original investment. The question that sounds like it should matter — what did Accel's other investments in that fund return? — is a trick question. It doesn't matter. One investment returning 1,000x means you could have been wrong on everything else in the portfolio and still produced a top-performing fund. Everything Accel made beyond Facebook (and they did make other strong investments) was gravy.

That arithmetic reframes the entire logic of how VCs operate. The correct performance metric is not batting average — how often you pick a winner — but what Kupor calls "at bats per home run": how frequently any given investment produces 10x or better. A VC who bats .600 by being selective and cautious, steering clear of the risky swing, is probably underperforming a peer who bats .400 but lands a 1,000x every few funds. Trying to minimize losses by avoiding long shots is, structurally, the losing strategy.

Every behavior that looks strange about the VC world follows from this one mathematical fact. The fixation on market size: a small market can't produce a home run no matter how good the company. The appetite for moonshots over incremental improvements. The relative indifference when a portfolio company fails. These aren't attitudes. They're outputs of a specific financial incentive, as logical as any formula once you know which variable you're solving for.

Your VC's Pressure to Exit Has Nothing to Do With You — It's About the Pension Fund Behind Their Fund

Why would a VC on your board push you toward an acquisition offer you think is too low? The instinct is to read it as betrayal: someone who pledged to back you for the long term now steering you toward a door you didn't want to open. The reality is colder and, in a way, more useful: your investor is doing exactly what the math of her situation demands.

Here's the chain. Yale's endowment allocates 16 percent of its assets to venture capital, three times the average at peer institutions, because VC has returned roughly 77 percent annually over two decades. Yale's chief investment officer has no choice but to keep chasing those returns; the endowment funds a third of the university's annual operating budget. But Yale can't live on paper marks. It needs realized cash to write those checks to the university, to reinvest with managers it trusts. So it presses its VC managers for liquidity. Those managers press you.

Clawback makes it personal. When a GP — the fund manager — distributes carried interest, her share of the profits, based on interim valuation marks, and then the rest of the portfolio collapses, she must personally return that money to her LPs. If your company is the lone bright spot in an otherwise struggling fund, an acquisition offer that locks in real cash may be the only thing keeping her out of clawback territory. She isn't thinking about your upside. She's calculating a personal legal obligation that predates your first meeting.

The fund's age sharpens the math further. VC funds run roughly ten years, with new investments typically limited to the first five or six. A VC who backed you in year five is running a different clock than one who came in during year two: less runway, fewer reserves for follow-on rounds, more pressure to manufacture liquidity before the fund closes. Kupor's advice is blunt: ask which numbered fund is writing your check and when it was raised. That single conversation reveals the actual timeline you're both operating against.

VCs Don't Invest in Ideas — They Bet on Whether You Specifically Can Win a Huge Market

When Andreessen Horowitz heard the Series A pitch for Square, the product was obvious: a thumb-sized card reader that plugged into a smartphone and let anyone — a street vendor, a craftsperson at a country fair — accept credit cards. The problem was real. The market was enormous. They passed anyway.

The reason was Jim McKelvey. Jack Dorsey's cofounder had started the company after losing a sale at a glass-blowing fair because he couldn't take a credit card. Classic founder-origin story. But McKelvey was still CEO, and Kupor's team couldn't evaluate whether he could run a payments company at scale. They wondered whether Dorsey would be a better fit. So they passed.

The frame was wrong. Within months, Dorsey took over as CEO. But the deeper miss was what his involvement meant in practice: Dorsey was a celebrity in a market where celebrity had direct commercial value. He landed Square a spot on The Oprah Winfrey Show, free marketing at a scale money couldn't buy. He called Jamie Dimon at JPMorgan directly and struck a deal to bundle Square's card reader with JPMorgan's credit card business. Square went public at a $25 billion valuation.

The question a VC is actually asking isn't whether the product is right — at the early stage, almost certainly it isn't. It's whether this founder has an unfair advantage that can't be replicated. Could a better-suited team walk through my door tomorrow? If yes, pass. If no — if the founder's background or network is so precisely fitted to the problem that it's implausible elsewhere — that's where the check goes.

The Term Sheet Clauses Your Lawyer Called 'Standard' Can Take Half Your Company in a Down Round

Imagine two job offers landing on the same day: same salary, same title, same start date. You choose the one with the nicer office. Six months later, the company hits a rough patch. Under your contract, "underperformance" triggers a forfeiture of 40 percent of your vested equity. Under the other letter, it doesn't. Two documents that looked identical on signing day. One clause neither your recruiter nor your lawyer mentioned.

That's what antidilution structure is, and where founders who optimize on headline valuation while skimming the rest of the term sheet tend to end up.

Two hypothetical term sheets — Haiku Capital and Indigo Capital — both come in at the same $8 million pre-money valuation. The numbers on the cover page look nearly identical. The difference is buried in the economics section. Haiku proposes broad-based weighted average antidilution. Indigo proposes full ratchet.

For eighteen months, it doesn't matter. Both investors hold their stakes. Founders run the company. Then things go sideways. The company burns through its runway without hitting its milestones, and the only new investor willing to write a check wants to do it at a $6 million pre-money valuation. A down round.

Under Haiku's terms, the math is painful but survivable. The weighted average formula blends the original price per share with the new lower one, issuing Haiku a modest number of additional shares to compensate for the markdown. Founders land at 44 percent ownership.

Under Indigo's full ratchet, what follows is a different species of painful. Full ratchet doesn't blend anything — it resets Indigo's original price entirely to the new lower valuation. To make Indigo whole, the company must issue so many new shares that Indigo's position quadruples. Indigo ends up owning 50 percent of the company. Founders collapse to 19.4 percent. The same down round. The same new money. One clause. More than a two-times difference in what the founders own.

In practice, the new investor would likely demand Indigo waive its full ratchet as a condition of closing (which Indigo has every right to refuse). The clause that looked like routine boilerplate eighteen months ago is now a veto over whether the company survives.

The specific flavor of every economics term determines who holds what in the scenarios founders tend not to model when things are going well. Valuation is the number that goes in the press release. Antidilution is the number that matters when you most need it not to.

Your Board Member Has Two Sets of Financial Obligations — and Yours Ranks Second

Picture a board meeting you're not in. The company is struggling: it missed its milestones, burned through its runway, and the only offer on the table is from an acquirer willing to pay $60 million. One of the VC board members has stopped attending regular meetings. He shows up for the acquisition calls, tells the CEO to close quickly, and says out loud that he'd rather be spending his time on companies with real upside. He isn't hiding this. He isn't embarrassed by it. He is, from where he sits, being perfectly rational.

That's the Trados case, a real Delaware lawsuit that became the defining legal precedent for VC-backed company governance. Trados had raised $57.9 million across several rounds, with liquidation preferences to match. When the company sold for $60 million, the VCs took their preference back, a management incentive plan paid out $7.8 million to senior executives, and the founders and early employees who held common stock received exactly zero. A common shareholder sued.

Delaware's court looked at the board and counted conflicts. The VC directors were conflicted for a reason that had nothing to do with bad faith: their liquidation preferences had structurally decoupled their incentives from common shareholders'. If the acquirer had offered $45 million instead of $60 million, the VCs would have felt every dollar of the decrease. Their preference wouldn't be fully covered. But at $75 million, they would have collected the same amount either way. Once their preference was satisfied, every additional dollar flowed to common, not to them. Pushing hard for a higher number cost the VCs time and energy with no economic return.

The court gave this a name: the "VC opportunity cost model." A venture portfolio is a competition for a GP's attention, and losing companies lose. The rational move is to extract the liquidation preference efficiently and redeploy the time. That one VC's candor — stop attending, close it fast, better things to do — wasn't a smoking gun. It was the financial structure speaking out loud.

The VC who stopped showing up wasn't betraying the company. He was doing what his financial position predicted. That's the thing about dual fiduciaries: the obligation to common shareholders is real, but it operates in direct competition with contractual rights explicitly designed to pay preferred holders first. The math was never personal.

By the Time You're Exiting, the Outcome Was Largely Determined by the Deal You Signed Three Years Earlier

The outcome at exit is largely a reading of what was signed three years before. By the time an acquisition offer arrives or the IPO window opens, the meaningful negotiations are over: done at term sheet, accumulated across rounds of preference, built or neglected through years of relationships with potential acquirers. The exit reveals the architecture; it doesn't redesign it.

The clearest demonstration is the acquisition path. Founders who exit at strong valuations almost never got there by putting out feelers when they decided it was time to sell. They built working partnerships with natural buyers years earlier — integrations, joint customers, product relationships. By the time someone picks up the phone for an acquisition conversation, the other side has already seen the team operate and has a commercial reason to want the deal closed. The negotiation is still hard, but it starts somewhere different: motivated buyer, established trust, a history that makes the ask plausible.

The same logic runs backward through every chapter in this book. Antidilution structure chosen at Series A determines how much a founder owns after a down round years later. Liquidation preference accumulated across rounds determines whether a $60 million exit feels like a win or a wipeout, and whether the board is aligned when that offer arrives or quietly hoping to close fast and redeploy elsewhere. Board composition locked in at the term sheet stage determines who controls the room when things get difficult.

You'll use the map most on the day you sign, because by the time you receive an offer, the architecture is already drawn.

The Room You Were Never Supposed to Enter

The system was never designed to disadvantage you. It was designed around obligations that existed before you walked in — LP agreements, fund timelines, carried interest calculations, liquidation preferences stacked across rounds. The behavior that felt like pressure, or betrayal, or misaligned priorities was just those obligations doing exactly what they were written to do. None of it was personal. All of it was structural.

That's actually the hopeful part. Structures can be learned. The gap was real. The map is now yours. Now that you can trace any board member's incentives back to the clause that created them, the asymmetry shifts. You'll read the term sheet the way someone reads it after seeing a thousand of them.