Communicate

Draft a capitalization summary memo

Turns round terms and a cap table into a memo with reproducible math: pre/post tables, the pool-shuffle cost in points and dollars, and a waterfall that names the exit value where preferred converts.

About 15 minintermediateTransactional

Your prompt4,765 characters

Still to fill in: Round terms, Current cap table

RoleYou are a transactional partner who has walked founders, boards, and employees through the same cap table without getting the arithmetic wrong. You show the division behind every number because someone always checks it, you write "the pre-money pool costs the founders 2.4 points" rather than "there is dilution," and you will not publish a waterfall without naming the exit value at which the preferred converts.What I needDraft a capitalization summary memo for Founders / management from the round terms and cap table below. Convert everything in Notes and SAFEs converting before you compute post-round ownership, and run the waterfall at Exit values to model.InputsRound terms: Round terms Current cap table: Current cap table Notes and SAFEs converting: Notes and SAFEs converting Exit values to model: Exit values to model Who reads this: Founders / managementHow to work this1. Open with three sentences: what the round is, what the founders hold post-close on a fully diluted basis, and the one number this reader will repeat to someone else. 2. Build the pre-round and post-round tables: holder, shares, % outstanding, % fully diluted, dollars invested. Derive the price per share and show the division. 3. Convert every instrument in Notes and SAFEs converting before computing post-round percentages. Caps, discounts, and pre- versus post-money SAFE math each change the share count; state which convention you applied. 4. If the pool expansion is created pre-money, break out the shuffle: founder dilution caused by the round versus by the pool, in points and in dollars of pre-money value. 5. Run the waterfall at each value in Exit values to model. Show the preference stack recovering first, then state the exit value at which the preferred is better off converting and show that arithmetic. 6. Caption every table in one plain sentence: "above roughly $160M everyone shares pro rata" beats a row of percentages, and footnote every figure so a reader with a spreadsheet could reproduce it. 7. Flag where the math turns fragile: weighted-average anti-dilution in a down round, pay-to-play conversion, a preference stack that grows with the next priced round.Close with these four sections, every time, without being askedAssumptions I made. Share counts I inferred, the SAFE conversion convention I used, whether the pool is measured pre- or post-money, whether the preferred participates. Mark each [verify] or [safe]. Where this is weakest. The two or three figures most likely to move once someone opens the actual cap table. Name the number and the row it sits on, not "the model generally." What only you can decide. Options with tradeoffs, not flags. At minimum: push the pool expansion post-money: recovers roughly two points of founder ownership but reopens a term the lead treats as settled, or trade it for a higher pre-money and leave the pool alone. Also yours: how much of the waterfall Founders / management should see. What would make this materially better. Ranked: the executed charter, the SAFE side letters, the current 409A report, or the pro forma the lead's associate built.Output formatThree-sentence bottom line, then tables: pre-round, post-round, dilution by holder with point deltas, pool-shuffle math, and the waterfall with a plain caption under each. Risk callouts, a closing paragraph pitched to Founders / management, and numbered footnotes carrying the arithmetic.Never do this- If the memo would read the same for any startup at any stage, it is too generic. Anchor every sentence to these share counts and this preference stack. - No hedging filler. Cut "arguably," "generally speaking," and "it depends." Do not tell me to consult counsel or a valuation firm. I am the counsel. - Never invent a term that is not in my inputs. If the charter is silent on participation or the anti-dilution formula, mark it [UNVERIFIED - confirm in the charter] rather than assuming the market default. - Where you do not know a number, leave the cell blank and say what you need. Do not smooth over a gap with a plausible-looking figure. - Do not pad. If the only real story is the pool shuffle, say that in three sentences and go to the tables. Length is not value.Before you answer- Does every percentage trace to a footnote a reader could reproduce, and do the fully diluted columns sum to 100%? - Did I convert the SAFEs and notes before computing ownership, and name the convention? - Did I state the exit value at which the preferred converts, with the math? - Would this memo be useless to a different company's founders? It should be.

Adds driver's-seat tunes: options instead of answers, questions before work, every citation flagged. Your values come with it.

Pressure-test it

Makes the AI switch hats and attack its own answer.

Somewhere in this memo is the line that will not tie to the lead investor's own pro forma. Their finance associate built it last week and will flag any figure that differs from theirs by more than a tenth of a point. Run the reconciliation as the associate. Where would their model and yours diverge: SAFE conversion convention, whether the pool is measured on the pre- or post-money share count, the treatment of unallocated options in the fully diluted denominator? List each divergence, say which convention is more defensible and why, and rewrite the two footnotes most likely to be challenged so the arithmetic is unarguable on its face.

Go deeper

Pushes the work further once the basics are right.

Employees hear about the round long before they understand what it did to their options. Produce the one-page employee equity FAQ for the same round: what changed for existing option holders, why the strike price on new grants is set by the 409A rather than the preferred price, how the round affects vesting and the exercise window, and what an option is worth at each exit value already modeled. Plain English, no defined terms, no more than one page.

Before you run it

What to gather first

Round size, pre-money, lead, and whether the pool expansion is pre- or post-money
The current cap table with actual share counts, not just percentages
Every SAFE and convertible note outstanding, with cap, discount, and conversion convention
The full preference stack across prior rounds: amount, multiple, participation
Who reads the memo and how much math they will tolerate

Watch for

The model will confidently produce arithmetic that is subtly wrong. Re-run the price per share, the fully diluted denominator, and the conversion threshold in a spreadsheet before the memo leaves your desk.
SAFE conversion is where models break: pre-money and post-money SAFEs on the same cap table produce different share counts, and the memo must say which convention it applied.
Liquidation preferences stack by class and do not average. Model each class in seniority order, and re-check whether participation is capped.
The 409A fair market value is determined separately from the preferred price. Never tell an employee the strike will equal the round price.
Broad-based weighted-average anti-dilution math is easy to state and easy to get wrong. Verify the formula against the charter language, not against a remembered template.

What comes back

A three-sentence bottom line, pre- and post-round cap tables, a dilution-by-holder table with point deltas, pool-shuffle math where the expansion is pre-money, a waterfall across the stated exit values with plain-English captions, risk callouts, a bottom-line paragraph pitched to the audience, and numbered footnotes carrying the arithmetic.

See an example of what you’ll get

Bottom line. The round is $20M at $80M pre-money, led by Andreessen at $14M. After close and after the $2.5M of SAFEs convert, the founders hold 37.5% fully diluted, down from 61.9%. About 5.0 points of that drop is the pre-money pool expansion rather than the money, roughly $5.0M at the $100M post-money. Post-round cap table | Holder | Shares | % FD | Invested | |---|---|---|---| | Founders | 30,000,000 | 37.5% | n/a | | Seed preferred | 8,000,000 | 10.0% | $4.0M (prior) | | SAFE conversion shares | 9,000,000 | 11.3% | $2.5M (prior) | | Series A - Andreessen | 11,200,000 | 14.0% | $14.0M | | Series A - syndicate | 4,800,000 | 6.0% | $6.0M | | Other common | 2,500,000 | 3.1% | n/a | | Options outstanding | 2,500,000 | 3.1% | n/a | | Pool unallocated | 12,000,000 | 15.0% | n/a | | Total | 80,000,000 | 100% | $20.0M | *Series A price: $1.25 per share, $80,000,000 of pre-money value ÷ 64,000,000 pre-money fully diluted shares.* [^1] *Caption: at the $300M case every point of founder ownership is worth $3.0M.* Pool shuffle. The lead asked for a 15% post-close pool, 12,000,000 shares. The 5,500,000 unallocated shares already outstanding carry into it, so 6,500,000 new shares are created inside the pre-money and the existing holders absorb all of them; the new investors absorb none. Founder dilution attributable to the expansion alone: 5.0 points, or about $5.0M at the $100M post-money. [^2] Waterfall | Exit | Preference recovers | Founders receive | What is happening | |---|---|---|---| | $60M | $20.0M to the Series A; seed and SAFE shares convert | $18.8M | The Series A takes its money off the top; the remaining $40M spreads over the other 64,000,000 shares. | | $100M | $20.0M, or nothing; this is the conversion point | $37.5M | Exactly where the Series A is indifferent between taking its preference and converting. | | $300M | Every class converts | $112.5M | Above the threshold everyone shares pro rata. | | $1B | Every class converts | $375.0M | Same math, larger number. | *Conversion threshold: the Series A converts above $100M of exit value, where its 20.0% pro rata share first equals the $20M preference.* [^3] Fragility. A $60M down round would trigger broad-based weighted-average anti-dilution on the Series A and cost founders an estimated 1.5 to 2 additional points. Confirm the formula against the charter, not against this estimate. Pay-to-play would push non-participating Series A holders into common and work in the founders' favor. [^1]: $80,000,000 ÷ 64,000,000 pre-money fully diluted shares = $1.25. $14,000,000 ÷ $1.25 = 11,200,000 shares; $6,000,000 ÷ $1.25 = 4,800,000; together 16,000,000, which is 20.0% of the 80,000,000 post-close count. [^2]: With the expansion: 30,000,000 ÷ 80,000,000 = 37.50%. Without it (pool left at the 5,500,000 shares already unallocated), the post-close count falls to roughly 70,545,000 and the founders hold 30,000,000 ÷ 70,545,000 = 42.53%. The 5.03-point difference, on a $100,000,000 post-money, is $5.03M. [^3]: Series A pro rata = 16,000,000 ÷ 80,000,000 = 20.0%. Conversion beats the $20,000,000 preference when 20.0% × exit > $20,000,000, that is, above $100M. Below it the Series A takes its preference and the balance spreads over the other 64,000,000 shares, on which basis the seed converts above roughly $52M and the SAFE shares above roughly $38M. --- Assumptions I made. - SAFE conversion: I read the $1.5M instrument as a post-money SAFE and converted it on the $12M post-money cap, giving it 12.5% ($1.5M ÷ $12M) of the 64,000,000 pre-money fully diluted shares, or 8,000,000 shares; I priced the $1M uncapped SAFE at a 20% discount to $1.25, or $1.00, for 1,000,000 shares [verify against the executed documents; a pre-money form on these same facts produces a different share count and moves every percentage in the post-round table]. - The 15% pool is measured against the post-close fully diluted count, and the 3-point expansion is created pre-money [safe - that is what the term sheet says]. - Seed preferred is 1x non-participating with no accruing dividends, and the converted SAFE shares carry a preference equal to the $2.5M invested rather than 1x the $1.25 issue price [verify - the charter controls. On the issue-price convention the SAFE preference is $11.25M rather than $2.5M and the $60M row changes materially, and accrued dividends would push every preference figure up]. - Options outstanding and the unallocated pool both sit in the fully diluted denominator [safe - the lead's model will do the same]. - No prior-round anti-dilution adjustment is triggered at the $1.25 price, which is well above the seed's $0.50 ($4,000,000 ÷ 8,000,000) [verify against the seed charter]. Where this is weakest. - The unallocated pool row: 12,000,000 shares. I sized it as 15% of the post-close count on its own, with the 2,500,000 outstanding options on top, so total plan overhang is 18.1%. If the lead means 15% inclusive of outstanding grants, the unallocated pool is 9,500,000 shares, the post-close count is smaller, and every percentage in the table moves. One sentence in the term sheet decides it. Re-run that row before this table goes anywhere. - The class interaction in footnote 3. The $52M and $38M figures assume the Series A decides first and the seed and the SAFE shares then decide against what is left. That is the conventional order and it is not the only one a model can take; solving all three classes simultaneously moves both numbers. Model them together before a board sees it. - The 9,000,000 SAFE conversion shares. That figure rests entirely on the convention assumed above: the $12M post-money cap alone accounts for 8,000,000 of them. If the $1.5M instrument is a pre-money form, the founders' 37.5% moves and so does every row beneath it. What only you can decide. - Push the pool expansion post-money, or leave it. Creating the 6,500,000 new pool shares after the round rather than inside the pre-money makes the new investors bear their share of them and recovers about 1.7 points for the founders (roughly $1.7M at the $100M post-money) but it reopens a term the lead's paper treats as settled, and the standard counter is a lower pre-money that gives back part of what the ask recovers. Leaving it alone costs those points and keeps the one ask you have available for the board seat or the protective provisions. - Size the pool from a hiring plan, or accept the lead's number. If the plan through the next round needs two points, ask for 14% and keep one. If it needs four, taking 15% now is cheaper than a second expansion priced into the Series B. A pool expansion always costs the people who are already here. - How much of the waterfall the founders should see. The $60M row is the one that changes behavior: a modest exit returns $18.8M to the founders after $20.0M comes off the top for the Series A preference. Showing the full table is the honest version and it is a hard meeting. Stopping at $100M is an easier conversation and leaves them without the single fact that should drive an exit decision. What would make this materially better. Ranked by impact. 1. The executed SAFE documents. The pre- versus post-money convention on the $1.5M instrument moves every percentage in the post-round table and is the largest single source of divergence with the lead's pro forma. 2. The current charter: seed participation, dividend accrual, and the exact weighted-average formula. The waterfall and the fragility paragraph both rest on terms I inferred rather than read. 3. The lead's own pro forma. Reconciling to it now costs an hour; reconciling to it at the board meeting costs the memo its credibility. 4. The current 409A report. Not needed for these tables, but needed before anyone tells an employee what the strike price on a new grant will be.

Why this prompt is built the way it is

## Framework 1. **Bottom line up front.** Three sentences: what the round is, what the founders hold post-close fully diluted, and the one number this reader repeats to someone else. 2. **Pre and post tables.** Holder / shares / % outstanding / % fully diluted / dollars in. Derive the price per share and show the division. 3. **Convert first.** SAFEs and notes convert at the round. Caps, discounts, and pre- versus post-money SAFE conventions each change the share count. State which convention you applied before computing ownership. 4. **Break out the pool shuffle.** If the expansion is created pre-money, separate founder dilution caused by the round from dilution caused by the pool, in points and in dollars of pre-money value. 5. **Waterfall at the stated exit values.** Preference stack recovers first; then name the exit value at which conversion beats the preference, and show that arithmetic. 6. **Plain-English caption under every table.** "Above roughly $160M everyone shares pro rata" beats a row of percentages. 7. **Footnote the math.** A reader with a spreadsheet should reproduce each figure. 8. **Call out fragility.** Weighted-average anti-dilution in a down round, pay-to-play conversion, a preference stack that grows with the next priced round. 9. **Match depth to audience.** Board memo carries the tables; founder memo leads with outcomes; employee FAQ is one page with no jargon.