You're four days from a wire transfer and the lawyer has sent over a cap table spreadsheet with fourteen tabs. The valuation is what you negotiated. The check size is what you asked for. Somewhere in tab nine is a line for a new option pool, sized before the money comes in, that will cost you more equity than the investor's own check does. Most founders don't find that line until after they've signed.
Cap tables are not paperwork you review once and file away. The NVCA 2025 Yearbook lists the cap table as one of the three documents that show up in due diligence for every single funding round, alongside financial statements and legal incorporation paperwork. Investors read it closely because it tells them exactly what they're buying into and what happens to their stake later. Founders should read it with the same attention, because it tells them what they're giving up now and what they'll have left later.
What the table actually tracks
A cap table is a ledger of who owns what: founders, employees, advisors, and every investor across every round, expressed in shares and in percentages. The percentages move every time the company issues new shares, whether that's a funding round, a new employee grant, or an advisor's exercised options. Every issuance dilutes everyone who isn't part of that issuance. This is arithmetic, not politics, but the mechanics of it determine whether a founder owns a meaningful stake at exit or a rounding error.
The dilution is predictable enough that it's been measured. According to Carta's 2025 data, median dilution runs about 19.5% at seed, 18% at Series A, and 14% at Series B. Run that forward and a founder who started at 100% ownership is looking at roughly 36% after Series A and about 15% by IPO, per the same Carta figures reported by Startup.age. That's not a founder who negotiated poorly. That's the median outcome for a company that raised capital successfully all the way to a public listing.
The line item that moves before you notice
Option pools are where a lot of founders lose ground they didn't know was on the table. CRV's research shows early-stage companies typically set aside 10% to 15% of fully diluted shares for an employee option pool, vesting over four years with a one-year cliff. The standard practice is for investors to ask that this pool be created or topped up before the new money comes in, which means the dilution from the pool comes entirely out of the founders' and existing shareholders' side of the table, not the new investor's.
Run the numbers on a $10 million pre-money round with a 15% pre-money option pool top-up and the effective price per share drops in a way that isn't visible in the headline valuation. The investor's percentage ownership is protected. The founder's is not. This is standard, it's not a trick, but it is a negotiation point, and the size of the pool relative to actual hiring plans for the next 18 months is worth a real conversation before signing, not after.
The regulatory limits that cap what you can do with equity
Two statutory thresholds sit underneath every equity compensation decision a company makes, and both show up on a well-built cap table.
The first is SEC Rule 701, which caps the value of securities a private company can issue to employees and consultants at $10 million in any 12-month period without triggering additional public disclosure requirements. Companies granting equity aggressively to move fast on hiring need to track this number or risk an unplanned compliance event.
The second is the ISO $100,000 limit, an IRS rule that restricts how much in incentive stock options can vest for a single employee in a calendar year and still qualify for favorable ISO tax treatment. Options that vest above that threshold convert to non-qualified stock options, which carry a different, generally worse, tax outcome for the employee. This matters most for early, high-equity hires, whose grants are large enough relative to the company's size that they can cross the threshold without anyone flagging it. A cap table that models vesting schedules against this limit protects the employee from a tax surprise; a cap table that doesn't, doesn't.
Reading the founder split before the first outside dollar
The cap table conversation usually starts before a company raises anything, at the point where co-founders decide how to split ownership. Allied Venture Partners' analysis found that around 73% of first-time founders choose an equal split. That's a reasonable default when co-founders are contributing comparably in time, capital, and risk from day one. It becomes a liability when contributions diverge sharply, one founder works full-time for a year before the other joins, or one founder puts in the initial capital, and the cap table doesn't reflect any of that. Equal splits made under time pressure, without vesting schedules attached to the founder shares themselves, are a common source of later disputes precisely because the cap table becomes the referee and the document was never built to reflect what actually happened.
Reading valuation trend lines against your own round
Cap tables are also a way to sanity-check whether the round in front of you is priced the way the market is currently pricing similar rounds. Zeni.ai's analysis put the median Series B pre-money valuation at $118.9 million in Q3 2025, up from $102.8 million in Q3 2024. A founder raising a Series B substantially below that figure isn't necessarily being lowballed, sector and traction matter enormously, but it's a number worth having in hand before a negotiation, not after.
It's also worth knowing what a down round actually looks like in the current market. Carta data cited by Qubit Capital shows down rounds made up less than 14% of all new funding rounds in Q4 2025, the lowest rate in three years. Down rounds are painful on a cap table specifically because of anti-dilution provisions: many preferred share classes carry weighted-average or full-ratchet protection that reprices earlier investors' shares downward in value terms when a later round prices lower, which further dilutes founders and common shareholders beyond the raw new-money dilution. Reading the anti-dilution language attached to each existing preferred class, before agreeing to a new round's price, tells you exactly how a future down round would redistribute the table if one happened.
The case for not obsessing over the percentage
There's a real argument, and venture investors make it often, that founders who spend weeks negotiating an extra point or two of ownership at seed are optimizing the wrong variable. A smaller percentage of a company that raised enough capital to actually win its market can be worth far more in absolute dollars than a larger percentage of a company that ran out of runway trying to preserve ownership. Industry data puts the Series A to Series B conversion rate at roughly 30% to 40%, which means most companies that raise a Series A never raise a Series B at all. Against that base rate, the operator's counterargument is straightforward: the dominant risk is not surviving to the next round, making timely capital with acceptable terms more valuable than delayed capital with better ones.
There's a second version of this argument aimed specifically at the fully diluted share count, which is the denominator used to calculate ownership percentages on most cap tables. Critics point out that fully diluted figures include every outstanding option, warrant, and convertible note as if all of them were exercised or converted today, a scenario that rarely happens all at once. This can make dilution look worse on paper than it will be in practice, and a founder who reads the fully diluted number as a certainty rather than a ceiling can talk themselves out of capital that would have helped the company.
Both arguments are correct as far as they go. They fail under specific conditions. The market-window argument fails when the terms being rushed through include provisions, like an outsized pre-money option pool, a full-ratchet anti-dilution clause, or a large liquidation preference stack, that compound at every subsequent round regardless of how the company performs. Those terms don't just cost percentage points once, they reset the baseline that every future round negotiates from. The fully diluted argument fails when the unlikely scenario it's discounting is actually the likely one, a company that's already burned through its option pool and needs a top-up, for instance, has a fully diluted number that isn't hypothetical at all. Reading the cap table isn't the same as re-litigating every point of ownership. It's distinguishing the terms that cost you once from the terms that cost you every round after this one.
Who benefits from a founder skipping this step
Every party at the table except the founder benefits when the founder doesn't read the cap table closely. The investor's return math is protected either way, that's what the preference stack and anti-dilution language are for. The lawyer bills the same hours whether the founder asks questions or not. The recruiting hire promised a percentage of "the company" benefits from vagueness about which denominator that percentage applies to. The founder is the only party in the negotiation whose outcome actually changes based on whether they understood the document before they signed it. That's the whole case for reading it slowly, with a lawyer who works for the founder and not the round, before the wire lands.
