Our Series A term sheet arrived on a Wednesday, signed and dated. Nine days later, our lead investor's counsel put the round on hold, not over our revenue or our roadmap, but over our cap table. If you're a founder heading into diligence with SAFEs, an old advisor agreement, or an option pool nobody has audited in a year, this is what almost sank a signed deal for us, and the eight days it took to fix it.
We thought we owned 71 percent of the company going into the round. The diligence associate's fully diluted model put us at 61 percent. That 10-point gap between what we believed and what the paperwork actually said is the reason cap table cleanup has to happen before a term sheet shows up, not after.
Where the mess actually came from
Three SAFEs, signed across 14 months of scrappy fundraising, were the first problem. Two were uncapped, written when we were desperate for runway and didn't think hard about what they would look like once they converted. The third had a $6 million cap we'd negotiated with a lead angel. None of the three lived in the same spreadsheet. Each one sat in a folder, and we simply never modeled what happened when all three converted at once against a priced Series A.
The second problem was a departed advisor. She had left the company 18 months earlier on good terms, and we assumed her unvested shares had lapsed. They hadn't. An acceleration clause buried in her original agreement, one none of us remembered negotiating, meant her full 0.75 percent had vested the moment we signed a term sheet with a new investor. Nobody had flagged it because nobody had reread the agreement since the day it was signed.
The third problem was our own option pool. We had topped it up twice as we hired, both times informally, over Slack, without board minutes documenting either increase. On paper, it looked like equity had appeared from nowhere.
Why this almost killed the deal, not just delayed it
Cap table problems rarely surface in the partner meeting. They surface three to five weeks into diligence, when an associate builds a pro forma cap table from your actual signed documents instead of your pitch deck summary. If that model doesn't reconcile with what you told the partner your ownership would be, the investor has two options: re-price the round to reflect the real dilution, or walk.
Our lead investor's first instinct was to re-price. A 10-point swing in founder ownership is not a rounding error to a board member deciding whether the founding team still has enough skin in the game to stay motivated through a seven-year hold. We had eight days to produce a clean, reconciled table before that conversation became final.
How we fixed it in eight days
- We pulled every original signed document for every SAFE and side letter, not the summary spreadsheet we had been maintaining, which turned out to be wrong in three places.
- We built one waterfall model showing every SAFE converting at its actual cap or discount, alongside the priced round terms, so the fully diluted picture matched what an investor's own model would produce.
- We called the departed advisor directly instead of routing it through counsel first. She agreed to sell back half of the accelerated shares at a nominal price once we explained the acceleration was unintentional. That conversation took 20 minutes and saved weeks of legal back and forth.
- We drafted retroactive board consents for both option pool increases, got every director's signature, and attached them to the cap table as documentation.
- We sent the reconciled table to investor counsel before they had to ask a second time. That single move signaled we had gotten our arms around the problem, rather than waiting to be caught again.
What we would tell a founder starting today
Run this audit three to six months before you plan to raise, not after a term sheet lands. Every time you sign a SAFE or a note, add it to a live waterfall model that shows fully diluted ownership at the moment of signing, not a running list in a spreadsheet tab you will forget to update. Put every option pool change, no matter how small, in writing with a board signature the same week it happens. None of that is expensive or complicated. It is just easy to skip when you are focused on closing the next check instead of the one after it.
A messy cap table does not show up in your metrics dashboard, and it will not come up on a first call with an investor. It surfaces exactly when you can least afford it, in the weeks between a signed term sheet and a closed round. The founders who avoid this fire drill are not the ones with simpler cap tables. They are the ones who reconciled theirs before anyone else went looking.