How SAFE and note conversions can change Series A ownership more than the price per share, and how to model the numbers before the term sheet.
This is the conversion math ambush.
Most founders are trained to negotiate valuation, option pool, board control, liquidation preference, and pro rata. Those matter. But between seed and Series A, a quieter set of mechanics can shift ownership before the new money even lands: whether SAFEs convert into shadow preferred or standard preferred, what “Company Capitalization” includes for broad-based formulas, whether notes convert on principal only or principal plus accrued interest, whether discount or cap produces the lower conversion price, how MFN elections are handled, and whether the pre-money or post-money SAFE language interacts cleanly with the Series A documents.
The result is a common but underappreciated reality: two Series A term sheets with the same headline pre-money and the same investment amount can produce meaningfully different founder dilution solely because the conversion assumptions differ.
That is not a drafting curiosity. It is economics.
Founders who do not lock the math early often discover too late that the “real” round price is less important than the capitalization definition and conversion waterfall embedded in the financing stack. By the time diligence is deep, leverage has shifted to the lead, existing noteholders are protecting their own outcomes, and counsel is left arguing over ambiguous documents under deal-time pressure.
The problem: founders focus on price per share, while ownership moves somewhere else
In a priced round, the intuitive anchor is price per share. If the Series A is at $2.00 per share instead of $1.80, the founder thinks they preserved value. But the founder’s actual post-closing ownership depends on a broader sequence:
If you miss those inputs, the stated Series A price per share becomes an incomplete statistic. A founder can negotiate hard on the headline pre-money and still lose more ownership than expected because the conversion math widened the share base underneath them.
The trap is especially dangerous because each moving piece sounds technical and local. Shadow preferred sounds like lawyer plumbing. Interest accrual sounds minor. Broad-based capitalization definitions sound standardized. Discount-vs-cap elections sound mechanical. In combination, they are not minor.
They determine how many shares exist before the Series A investor’s money is even applied.
Why current approaches fail
1. Founders over-index on the term sheet and under-index on the conversion schedule
Many term sheets do not fully spell out every assumption used to model SAFE and note conversions. The lead may specify pre-money, amount raised, and pool refresh, but leave conversion details to definitive documents “consistent with existing instruments.” That phrase is where ownership surprises are born.
“Consistent with existing instruments” is only comforting if the instruments are clean, consistent with each other, and modeled the same way by everyone involved. Often they are not.
A company may have:
By the time Series A arrives, the financing stack may reflect three law firms, two CFOs, and one half-remembered data room.
2. “Standard” means different things to different counsel
There is no universal market convention on every conversion input. There are patterns, but not uniformity.
For example:
A founder assumes there is one “correct” answer. In practice, there may be multiple plausible answers, and the answer selected can move founder ownership by points, not basis points.
3. Diligence happens too late for clean leverage
The lead investor often delivers the definitive cap table model once the deal is already underway. At that point:
Yet that is precisely when the hidden ownership shifts become visible.
If the founder pushes back late, the lead can say the docs already require this result. If convertibles are ambiguous, each holder will advocate for the interpretation that benefits them. The company is then negotiating not one term sheet, but an entire conversion ecosystem under deadline.
4. Founders rarely compare “same price” term sheets on a like-for-like converted basis
Two investors can both offer a $12 million pre-money valuation and $6 million raise. One may assume all convertibles are in the pre-money capitalization and the option pool is topped up before pricing. Another may handle pool expansion differently, or model note conversion and SAFE capitalization in a way that narrows the denominator. One may insist on a shadow preferred structure that preserves the convertible investors’ economics with more precision. Another may not.
The term sheets look economically equivalent if you only read the headline. They are not equivalent once you run the cap table through closing.
Core insight: the real negotiation is over the denominator before the round, not just the numerator called valuation
Founders often think Series A dilution is driven by one variable: how much the new investor buys relative to the pre-money valuation. In reality, the biggest hidden driver is how many shares are deemed to exist before the financing closes.
That denominator is shaped by conversion assumptions.
If pre-Series A instruments convert into more shares than expected, founder ownership shrinks before the lead’s cash even enters. A higher Series A price per share cannot fully save you if the underlying converted share count is larger.
This is why “same price” term sheets can produce different outcomes. The investor buying Series A preferred at $2.00 per share may still end up with the same number of shares in both deals, but founders can own less in one version because the convertibles were modeled differently.
The clean way to think about it is this:
Founder post-close ownership = founder common shares / total fully diluted shares after all conversions, pool adjustments, and Series A issuance
Everything in that denominator matters.
The mechanics that quietly move ownership
Let’s unpack the specific areas where dilution shifts hide.
1. Shadow preferred
When SAFEs or notes convert at a lower effective price than the Series A investors pay, companies often issue a separate series of preferred, commonly called shadow preferred, so those converting investors get economics calibrated to their lower price.
Why does this matter? Because the structure can affect:
At a high level, shadow preferred is intended to preserve the bargain the SAFE or note investor earned through the discount or cap. If the Series A investor pays $2.00/share and a note converts at $1.25/share, issuing the noteholder the exact same Series A preferred at the same conversion ratio may not preserve equivalent economics. Shadow preferred adjusts around that.
The founder mistake is assuming shadow preferred is irrelevant because it does not change cash entering the business. It can alter how the conversion is modeled and what rights are attached to those converted shares. If there is ambiguity, counsel choices can materially affect ownership presentation and future governance dynamics.
2. Broad-based capitalization definitions
This is the sleeper issue.
Most founders have heard “broad-based weighted average” in anti-dilution contexts, but broad-based capitalization definitions also sit inside SAFE and note conversion formulas. The question is simple but economically decisive: what share count is used when calculating the conversion price implied by a valuation cap?
A typical cap-based conversion formula is conceptually:
SAFE price = valuation cap / company capitalization
The fight is over “company capitalization.” Does it include:
Small definitional changes here can move conversion price enough to change ownership meaningfully.
If the denominator is broader, the conversion price is lower, which means the SAFE gets more shares, which means founders own less.
3. Interest accrual on notes
Convertible notes are debt. Debt accrues interest unless the instrument clearly says otherwise.
Founders routinely underestimate how much this matters because they think in percentage rates rather than dollars and shares. But over a long fundraising cycle, accrued interest can add a real tranche of conversion shares.
Example: $3.0 million of notes at 8% simple annual interest outstanding for 24 months.
Accrued interest = $3.0 million × 8% × 2 = $480,000.
If the note converts at $1.20/share, that interest alone creates:
$480,000 / $1.20 = 400,000 shares.
If founder common is 8.0 million shares, 400,000 shares is not noise. It is a meaningful ownership shift, especially when paired with multiple SAFEs and a pool expansion.
And this is where process matters: some company models accidentally omit accrued interest, particularly when the note schedule is stale or the maturity date was amended. The noteholder’s counsel will not miss it.
4. Discount-vs-cap election
Most convertible notes and many SAFEs convert at the better of:
Founders often treat this as self-executing and obvious. Usually it is obvious in concept, but the actual election depends on the exact Series A price and the exact capitalization denominator used in the cap formula.
A tiny change to capitalization can flip whether cap or discount wins.
That matters because the difference in conversion price can be large.
Suppose Series A price = $2.00/share.
In both cases the cap beats the discount, but the broader denominator gives the investor even more shares. If a narrower definition pushed cap-based price to $1.68, then discount would win instead, and the investor would get fewer shares. One phrase in the SAFE can change the winner.
5. Pre-money vs post-money SAFE interaction
This article assumes readers know the basics, so the key point is narrower: mixed SAFE stacks create modeling risk.
Post-money SAFEs were designed to make dilution from the SAFE itself more legible at the time of issuance. But many companies have a mix of pre-money SAFEs, post-money SAFEs, and notes. Those instruments do not always interact intuitively in a later priced round.
If counsel models them in different conversion orders or applies different capitalization assumptions across instruments, the founder can get a result that feels inconsistent with how those securities were sold.
The legal answer will depend on the forms and side letters actually signed. The economic point is that mixed stacks amplify ambiguity and make “same price” term sheet comparisons much less reliable unless modeled line by line.
6. MFN side letters and bespoke amendments
The ugliest surprises usually come from documents founders barely remember exist.
An investor with an MFN clause may argue they are entitled to the most favorable economic term granted later to another SAFE holder. A side letter may have changed the cap, the discount, the pro rata right, or the capitalization definition. A note amendment may have extended maturity but left interest running. A waiver may not have been countersigned properly.
These are not rare edge cases. In many seed financings, the company optimized for speed. Fast paper is fine until a priced round forces every shortcut into one cap table.
Example: two Series A term sheets, same headline price, different founder dilution
Let’s build a realistic scenario.
Starting cap table before conversion
Founders and employees:
Total current fully diluted common equivalents before convertibles: 9,500,000
Convertible stack:
- Principal: $3,000,000
- Interest: 8% simple
- Time outstanding: 2 years
- Discount: 20%
- Valuation cap: $10,000,000
- Purchase amount: $2,000,000
- Valuation cap: $12,000,000
- No discount
- Purchase amount: $1,500,000
- Valuation cap: $15,000,000
Series A term sheet headline economics in both cases:
Now we compare two term sheets or, more accurately, two modeling approaches attached to “equivalent” term sheets.
Case 1: founder-friendlier conversion assumptions
Assumptions:
Step 1: determine Series A price per share
Using the cleanest pre-money capitalization baseline first, assume the negotiated Series A price is set off 9.5 million existing fully diluted shares:
$12,000,000 / 9,500,000 = $1.2632/share
Series A investor shares:
$6,000,000 / $1.2632 = 4,750,000 shares
That is only the starting point because convertibles must convert.
Step 2: note conversion
Accrued interest = $480,000, so total conversion amount = $3,480,000.
Discount price = 80% of Series A price = 0.8 × $1.2632 = $1.0106/share
Cap price depends on capitalization definition. Use 9.5 million shares for a founder-friendlier note cap denominator:
$10,000,000 / 9,500,000 = $1.0526/share
Lower price wins, so discount wins at $1.0106/share.
Note shares = $3,480,000 / $1.0106 = 3,443,699 shares
Step 3: SAFE A conversion
Cap price = $12,000,000 / 9,500,000 = $1.2632/share
SAFE A amount = $2,000,000
SAFE A shares = $2,000,000 / $1.2632 = 1,583,333 shares
Step 4: SAFE B conversion
For simplicity, assume its post-money mechanics here imply an effective conversion price of $1.1500/share under its form and side letter set.
SAFE B shares = $1,500,000 / $1.1500 = 1,304,348 shares
Step 5: pool top-up to 15% post-close
Shares before new pool top-up:
Subtotal = 20,581,380 shares
Existing option reserve included in the 9.5 million was 500,000 unissued. Suppose to hit 15% post-close on an unissued pool basis, the company needs 800,000 more shares added to reserve.
Total post-close shares = 21,381,380
Founder ownership = 8,000,000 / 21,381,380 = 37.42%
Case 2: investor-friendlier conversion assumptions
Same headline pre-money. Same $6 million raise. Same nominal price framing. Different assumptions:
These assumptions lower conversion prices for the convertibles, which means more shares issued before the Series A cash lands.
Step 1: revised capitalization denominator for cap-based instruments
Assume the broad model includes:
This pushes the effective denominator used in cap-based conversion math to 11,500,000 shares for certain instruments.
Series A headline pre-money is still $12 million, but the conversion formulas for SAFEs/notes now reference a broader denominator.
Cap-based prices become:
The note’s discount price remains tied to Series A price. If the Series A price itself is also reset by the broadened pre-money base in the legal model, say to $1.1450/share, then the note discount price is:
0.8 × $1.1450 = $0.9160/share
Cap wins now at $0.8696/share.
Step 2: note conversion under broader denominator
Total note amount with interest = $3,480,000
Note shares = $3,480,000 / $0.8696 = 4,001,840 shares
That is 558,141 more shares than in Case 1.
Step 3: SAFE A conversion under broader denominator
SAFE A shares = $2,000,000 / $1.0435 = 1,916,148 shares
That is 332,815 more shares than in Case 1.
Step 4: SAFE B conversion under broader modeling assumptions
Assume effective conversion price falls from $1.1500 to $0.9800/share because the broader capitalization interpretation applies to its mechanics too.
SAFE B shares = $1,500,000 / $0.9800 = 1,530,612 shares
That is 226,264 more shares than in Case 1.
Step 5: Series A shares and pool top-up
If the broadened model yields Series A price = $1.1450/share, then Series A shares for $6 million are:
$6,000,000 / $1.1450 = 5,240,175 shares
Now tally shares before final reserve:
Subtotal = 22,188,775 shares
Assume final pool true-up requires another 300,000 shares on top of the pre-closing expansion assumptions.
Total post-close shares = 22,488,775
Founder ownership = 8,000,000 / 22,488,775 = 35.57%
What changed?
Case 1 founder ownership: 37.42%
Case 2 founder ownership: 35.57%
Difference: 1.85 percentage points
That may sound small to outsiders. It is not.
On a company later worth $300 million, 1.85 points is $5.55 million of value before tax. On governance, every point also affects future voting power and dilution stacking into B and C rounds. And this example is not extreme. Larger note balances, older notes with more accrued interest, and bigger SAFE stacks can widen the gap materially.
Most importantly, the founder did not lose that ownership because the investor changed the headline valuation. The founder lost it because the conversion math changed the pre-closing share count.
Term sheet comparison: why “same pre-money” is often a false comparison
Consider two simplified Series A term sheets.
Term Sheet Alpha
Term Sheet Beta
Alpha and Beta look the same to an inexperienced founder. They are not.
Alpha turns the conversion math into a negotiated item while leverage still exists.
Beta delegates economics to later document interpretation.
If you sign Beta, you are effectively saying, “We will discover later how much of the company existed before this round.” That is not a valuation negotiation. That is surrender.
Where counsel models materially shift outcomes
Founders should understand where legal models diverge, even when everyone is acting in good faith.
Modeling issue 1: circularity treatment
Some cap-based SAFE formulas are circular because the capitalization used to set conversion price may include securities that themselves are converting based on that price. Different models solve this circularity with iterative calculations or simplifying assumptions. If the docs do not dictate the method, there can be room for economically different outputs.
Modeling issue 2: what counts as outstanding in the option pool
Granted, ungranted, reserved, and board-approved-but-unissued options can be treated differently depending on document language. A pre-closing pool refresh is especially contentious. Including it in the conversion denominator benefits cap-based convertibles and hurts founders.
Modeling issue 3: treatment of warrants and exotic rights
Old bridge lenders may have warrants. Advisors may have unusual equity-linked compensation. If the document says “all outstanding convertible securities, options, and warrants,” these can creep into the denominator.
Modeling issue 4: interest conventions
Simple versus compounding, exact day count, maturity default rates, extension amendments, and whether unpaid interest capitalized into principal under an amendment, all of this can change note share count.
Modeling issue 5: side letters not integrated into the main schedule
The cap table may reflect the main instrument but miss a side letter that changed the cap from $15 million to $12 million, or granted MFN rights later exercised by email. If that is surfaced during diligence, the company’s original dilution estimate can break.
Negotiation framework: how founders lock the math early
Here is the practical framework.
1. Treat the conversion schedule as a term sheet item, not a closing item
Before signing the Series A term sheet, ask for a fully converted cap table showing:
Do not accept “we’ll finalize with counsel later.” The model is the economics.
2. Mark up capitalization definitions explicitly
For each major convertible form, identify what “Company Capitalization” includes. Then negotiate term sheet language saying the financing will be structured consistently with the attached model and listing any exclusions that matter, especially:
If the documents are ambiguous, economic clarity now beats legal argument later.
3. Force a note interest reconciliation early
Get a note schedule with:
Then have both company counsel and lead counsel confirm the same numbers. If there is any mismatch, solve it before exclusivity dynamics lock you in.
4. Run the discount-vs-cap election line by line
Do not assume one method wins across all instruments. Run both for each note and SAFE. Small pricing or denominator shifts may flip the winner. Ask counsel to show the election result for every instrument in the model.
5. Clean up MFN and side letters before the Series A process begins
As part of pre-financing diligence, collect every:
Founders should ask one blunt question: “Could any investor credibly claim better conversion economics than our current model assumes?” If yes, resolve it early, ideally with a written acknowledgment.
6. Compare term sheets using founder post-close ownership, not headline pre-money
When deciding between competing leads, compare:
This is the right scoreboard.
7. Put ambiguity-resolution language in the term sheet
This can be as simple as:
“Investor and Company agree that the capitalization and conversion assumptions set forth in the attached cap table model shall govern the economic treatment of outstanding SAFEs and notes for purposes of the financing documentation, subject only to changes expressly required by the underlying instruments and agreed in writing by the parties.”
Lawyers will refine that. The point is to prevent model drift.
Diligence checklist: hidden cap table outcomes to surface before you are boxed in
Founders should be able to answer these questions before circulating a Series A draft:
If you cannot answer those questions, you do not yet know your Series A economics.
A direct view on leverage
Investors are not wrong to care about this math. Convertible investors paid for caps, discounts, and accrued returns. Series A leads want certainty that the cap table they are buying is real. Company counsel wants a defensible interpretation. Everyone is acting rationally.
The founder problem is different: if the founder is the only party not modeling every assumption precisely, the founder becomes the residual claimant of ambiguity. Every unclear term is usually resolved by someone else with stronger incentives and better spreadsheets.
That is why this issue feels like an ambush. The founder thinks the valuation negotiation was the main event. In reality, the main event may have happened years earlier when fast bridge documents were signed, or weeks earlier when the Series A term sheet omitted the conversion exhibit.
The practical takeaway
The right founder instinct in a Series A is not “What is the price per share?”
It is:
“Show me exactly how many shares exist immediately before and immediately after closing, instrument by instrument, using the documents that actually govern.”
Then ask the harder question:
“If I had another lead at the same valuation, would this model still be the same?”
If the answer is no, your real negotiation is not about valuation. It is about conversion math.
That math decides who owns what before the new money even hits the bank.
And once you understand that, you stop treating SAFE and note conversion mechanics like legal cleanup. You treat them as first-order financing terms, because that is exactly what they are.
RaiseSimpli Team
Author