Nobody discovers their own cap table mistakes. An investor's counsel discovers them, usually two weeks into diligence, when the fix is expensive and the leverage belongs to the other side. Here are the ten cap table mistakes that surface most often in due diligence: why each happens, the question that exposes it, and the fix.
1. The spreadsheet drifted from the signed documents
Why it happens. The cap table lives in a spreadsheet; the documents live in email threads and your lawyer's inbox. Each gets updated separately, usually from memory. Eighteen months later, a SAFE amendment never made it into the sheet, a transfer was typed with the wrong share count, and the total no longer ties to the stock ledger.
The question that exposes it. "Please tie each line of the cap table to the executed document behind it." A standard first-week request, and drift shows up immediately. A row with no document, or a document with different numbers than the row.
The fix. Reconcile every row to a signed document once, then change the workflow: entries get recorded from the executed document, never from memory. When a number and a document disagree, the document wins.
2. Option grants with no board approval
Why it happens. Options are granted by the board, not by the offer letter. But hiring moves fast: offers go out, people start, and the board consent gets batched "for the next meeting" and then forgotten. The employee believes they hold options; legally, nothing was granted.
The question that exposes it. "Provide the board consent approving each outstanding grant." Any grant whose paper trail starts and ends with an offer letter fails this request.
The fix. Batch new grants into a monthly or quarterly board consent; the approval date, not the offer date, is the grant date. For orphaned grants, have counsel approve them properly now; if fair market value has moved since the offer, expect a strike-price conversation. Granting options right walks through the correct sequence.
3. Grants that were never accepted
Why it happens. The board approved the grant, the employee was told, and the grant agreement went out as an email attachment, which was never signed and returned. Nobody's job was to chase it, and until the recipient accepts, the grant's status is genuinely murky.
The question that exposes it. "Provide the executed grant agreement for each optionholder." Unsigned agreements surface as gaps in the folder.
The fix. Audit every outstanding grant for a countersigned agreement and chase the stragglers now. It's an easy ask while people are employed and happy. Then make acceptance a required step of the grant process, not a loose attachment.
4. An ex-founder holding dead equity
Why it happens. Founders set up vesting for employees but skip it for themselves, because it feels insulting to ask your co-founder to earn their own company. Then one founder leaves early and keeps everything.
Worked example. Two founders split 8,000,000 shares equally, no vesting. One leaves at month eight, owning 50% permanently. With standard four-year reverse vesting and a one-year cliff, the company would have repurchased all 4,000,000 shares at cost. Without it, every future investor and employee is diluted to pay someone who left before the product shipped.
The question that exposes it. "Are founder shares subject to vesting, and what happened to the departed founder's stock?" Large blocks held by people no longer involved lose term sheets fast.
The fix. Put reverse vesting on founder stock from day one. If the dead equity already exists, negotiate a repurchase before you raise. Every month of progress raises the price of buying those shares back. Vesting explained covers the mechanics.
5. A SAFE stack nobody has modeled
Why it happens. SAFEs get signed one at a time over a year or two. Each felt small and each cap felt fair on the day, but nobody computed what they convert to together, because conversion felt like a future problem.
Worked example. Three post-money SAFEs: $500k at a $5M cap (10%), $500k at a $6M cap (8.3%), and $500k at an $8M cap (6.25%). At the priced round they convert together into roughly 24.6% of the company, before the new investor's shares and the pool top-up are even counted. Founders who mentally budgeted "a few points each" discover a quarter of the company is spoken for.
The question that exposes it. "Send a pro-forma cap table with all convertibles converted." If the answer takes a week to produce, the stack was never modeled.
The fix. Keep a running pro-forma and re-run it before signing each new SAFE, not after. SAFE stacking covers the math in depth, and the free SAFE calculator will model your stack in a few minutes.
6. An option pool sized by default
Why it happens. The term sheet says 15% because the template said 15%, and nobody pushed back with a hiring plan. Because the pool is created pre-money, its dilution lands entirely on existing holders.
Worked example. A $2M raise at an $8M pre-money with a 15% post-money pool created pre-money cuts the effective pre-money to about $6.5M. A 5% pool sized from an actual 18-month hiring plan keeps it at $7.5M, so founders end the round at 75% instead of 65%. Ten points of ownership ride on a default.
The question that exposes it. "How much of the current pool is granted, and what's the plan for the rest?" A large idle pool signals dilution taken early with no plan behind it.
The fix. Size the pool bottom-up from the hires you'll make before the next round. The option pool shuffle explains the negotiation; the ESOP calculator does the arithmetic.
7. Missing IP assignments
Why it happens. The core product was built before incorporation, or by a contractor whose agreement had no invention-assignment clause. Everyone assumed the company owned the code; legally, the individuals still do. Not strictly a cap table entry, but it lands in the same diligence folder and kills deals just as effectively.
The question that exposes it. "Provide signed IP assignment agreements for every person who contributed to the intellectual property." Pre-incorporation work and old contractors are exactly what this request is designed to catch.
The fix. Get confirmatory assignment agreements signed now, while relationships are warm. An ex-contractor with leverage mid-diligence is a much worse negotiation. Going forward, an invention-assignment agreement is standard onboarding for every hire and contractor, no exceptions.
8. New grants priced off an expired 409A
Why it happens. US options must be struck at or above fair market value, and a 409A valuation is generally good for twelve months, or until a material event, like the round you just closed. Nobody owns the renewal date, so grants keep flowing at a stale price.
The question that exposes it. "What valuation supported the strike price for grants made on these dates?" Grants issued in the gap have a defective strike price, and Section 409A's tax consequences fall on the option holders, your employees.
The fix. Calendar the expiry, refresh the valuation after every priced round, and pause granting when the current one is stale. Valuations and the 409A covers when a refresh is triggered.
9. Verbal equity promises
Why it happens. "You'll get a point when we raise" is an easy thing to say to an advisor, and it buys goodwill on the day. But a point of what, measured when, vesting how? Unwritten promises accumulate, and each one is a future claim on the cap table.
The question that exposes it. This one is built into the deal documents: financing agreements make you represent that there are no commitments to issue equity, written or oral, beyond what's on the cap table. The choice becomes disclose or misrepresent.
The fix. Paper every outstanding promise now: a real advisor agreement with a board-approved grant, or a candid conversation and a written release. Then stop promising percentages out loud; commit only through documents.
10. No versioned history
Why it happens. The cap table is one live spreadsheet, edited in place. Nobody can reproduce what it looked like at the seed close, or prove when a number changed, or why. Every edit silently destroys the record it replaced.
The question that exposes it. "Show the cap table as of the last round's close, and walk me through every change since." With one overwritten file, the honest answer is "we can't," and diligence slows to a document-by-document reconstruction.
The fix. Capture an immutable point-in-time version at every material event (each round close, repurchase, or grant batch) and keep an audit trail between them. This is where purpose-built software simply beats a spreadsheet: Vquity writes an automatic pre-close snapshot at every round, compares any two points with a field-level diff, and its Data Room scores your diligence readiness across twelve checks, so the history exists whether or not anyone remembered to save a copy.
Run the list before someone else does
Every question above will be asked eventually. The only variable is whether you answer calmly or scramble mid-diligence while the term sheet cools. Cap table mistakes are cheap to fix in the quarter before a raise and expensive during one, so run the ten questions against your own company this week, as if an investor sent them.
When you're ready to assemble the folder itself, building a diligence-ready data room covers what goes where, and the closing-a-round checklist sequences the fixes into the raise.