How one harmless-looking MFN concession can spread across your pre-seed SAFE stack, and how to handle side letter requests from angels.
An angel wants a side letter. Nothing dramatic, just a “most favored nation” clarification plus one or two sweeteners because they are moving quickly, wiring this week, and helping anchor the round. Counsel says the language is manageable. The company is trying to close a pre-seed on momentum, so nobody wants to turn a $150,000 check into a three-week legal debate.
Six months later, the company is raising a Series Seed or Series A. Lead counsel asks for every SAFE, every side letter, every amendment, and every investor rights document. Then the questions start:
What looked like a harmless investor accommodation starts behaving like a latent system-wide repricing event.
That is the MFN cascade problem.
It is not just a legal drafting issue. It is a financing design issue. A cap table issue. A diligence issue. And if handled badly, it becomes a negotiation issue precisely when the company can least afford distraction: right before a priced round.
The core mistake founders make is treating MFN as if it were a narrow courtesy clause. In practice, across a stacked pre-seed made up of multiple SAFEs issued over time, MFN can function like a distribution mechanism for concessions. If one later investor gets a better cap, a discount, shadow preferred rights, expanded definitions, side-letter governance rights, or procedural protections, prior investors with MFN may claim the right to adopt some or all of those improved terms. If enough prior money can opt in, the “exception” is no longer an exception. It becomes the de facto baseline for the round.
That is why sophisticated Series A counsel diligence MFN exposure with unusual intensity. They are not being pedantic. They are trying to answer a simple economic question: what exactly converts, on what terms, and who can still change that answer?
The problem founders underestimate
Most founders understand obvious dilution. They know that a lower valuation cap means more shares on conversion. They know that stacked SAFEs can produce more dilution than expected. What they often miss is that MFN can change the terms of existing SAFEs after the fact, without a new financing announcement and sometimes without any immediate papering beyond notice and election.
That creates three distinct risks.
1. Economic contagion
A later concession can reprice earlier money.
Suppose the company has raised:
To close the last $300,000 quickly, the company grants:
If the earlier $1.2 million tranche includes MFN rights that allow investors to adopt more favorable terms from later SAFEs, those investors may be able to swap into the $6.5 million cap or pick off selected favorable provisions, depending on the drafting.
That means the company did not just improve terms for $300,000. It may have improved terms for $1.2 million or more.
2. Rights contagion
The concession may not only be about price.
Many founders hear “MFN” and think only about valuation cap or discount. But depending on the SAFE version and side-letter drafting, the fight often centers on rights, not price. Investors may claim access to:
Those rights can be individually manageable. In aggregate, they can become a governance mess.
Giving one investor quarterly reporting is trivial. Giving it to 28 scattered SAFE holders is an administrative burden and a confidentiality risk. Giving one investor pro rata is fine. Extending it through MFN to every earlier check can materially complicate allocation in the next round and reduce room for the new lead and key insiders.
3. Diligence contagion
Even if nobody has exercised MFN rights yet, Series A counsel will ask whether they can.
That is the hidden problem. The issue is not only what has already cascaded. It is what remains electable. If old investors can still claim better terms after the lead has priced the round, the lead is underwriting a moving target.
No serious lead likes moving targets.
At best, this creates a pre-closing cleanup. At worst, the lead conditions signing on obtaining investor waivers, amendments, or a comprehensive SAFE exchange. That turns a financing into a consent solicitation exercise.
Why current founder approaches fail
Founders usually respond to MFN risk in one of four bad ways.
Bad approach #1: “MFN is standard, so it must be harmless”
This is lazy thinking.
Standard does not mean harmless. It means common. SAFEs are standardized forms dropped into non-standard fundraising behavior: rolling closes, mixed cohorts, strategic investors, accelerators, friends and family, angels with side letters, and ad hoc concessions made under time pressure. Once you add bespoke side agreements to standardized instruments, you are no longer in a clean standard regime.
MFN is especially dangerous in rolling SAFE rounds because there is no single negotiated package. Terms drift over time. Each drift point creates optionality for prior holders.
Bad approach #2: “The side letter only applies to that investor”
Maybe. Maybe not.
Whether the concession can spread depends on the exact language of:
Calling something “personal” or “non-transferable” does not necessarily keep it from becoming a more favorable term for MFN purposes. If the earlier SAFE entitles investors to adopt any term more favorable than those in their own SAFE, then relabeling the concession does not solve the substantive issue.
Bad approach #3: “Nobody complained at the time, so it’s fine”
This is not how diligence works.
MFN rights are often dormant until someone has a reason to care. A priced round is a reason to care. Conversion economics are becoming real. Pro rata rights now matter. Governance rights become operational. Investors who ignored paperwork when writing a $50,000 angel check suddenly forward documents to counsel.
Dormant does not mean waived.
Bad approach #4: “We’ll sort it out during the Series A”
That is exactly when your leverage is worst.
Once a lead investor has surfaced the issue, they control timing pressure. Founders are now balancing:
n- investor optics
The right time to solve MFN exposure is when you are granting the concession, not when your lead asks for a bring-down certificate.
What actually cascades
Here is the contrarian point founders and even some junior lawyers miss: MFN does not automatically turn every better term into universal treatment, but it creates a menu-rights problem. Prior investors may be able to elect into certain favorable terms selectively, and that selectivity can be more dangerous than a full amendment.
The practical question is not “Does MFN apply?” The practical question is “Which terms can earlier holders plausibly extract, and in what combinations?”
Category 1: Pure economic terms
These are the obvious ones:
If earlier investors can elect a lower cap, the cap table impact is immediate and modelable.
Category 2: Hybrid economic-rights terms
These look procedural but affect economics:
These do not change current conversion numbers directly, but they alter who gets to preserve ownership later. In a tight Series Seed allocation, that matters a lot.
Category 3: Governance and information rights
These include:
Founders often assume these are non-economic and therefore low risk. Wrong. Governance rights propagate cost, complexity, and leverage. A board observer right given to one strategic investor is often tolerable. MFN-spread observer demands from multiple investors are intolerable.
Category 4: Protective carve-outs and definitions
This is where drafting gets dangerous.
A side letter might narrow the company’s ability to amend documents without notice, alter definitions of “equity financing,” or create favorable treatment in edge cases like secondary sales, token distributions, IP transfers, or affiliate transactions. Those changes may not matter today, but they become very relevant in a future financing or exit.
Series A counsel often focuses here because definitional asymmetry can create unexpected closing conditions or investor claims.
Category 5: Procedural mechanics
Even something as boring as notice timing can matter.
If one investor gets enhanced notice before a financing closes, and earlier investors can claim the same right through MFN, the company may face a wider set of investors entitled to review documents, ask questions, or elect into rights before closing. That can slow execution materially.
How the cascade happens in practice
Let’s model a realistic pre-seed stack.
The initial stack
A startup raises capital over 10 months:
By the time the company reaches Cohort D, revenue is still light and the market has softened. The strategic investor wants sweeter terms to justify moving now.
The company grants Cohort D:
Founders tell themselves the damage is contained because the total amount is only $350,000.
That is where they get punished.
Step 1: Earlier holders review MFN language
Suppose Cohort A’s SAFE says investors may elect to amend their SAFE to include any terms offered in a subsequent convertible instrument that are more favorable than those in their own SAFE, provided the company determines in good faith that such terms are applicable to the investor.
Founders read “applicable” and relax.
Counsel for earlier investors reads “more favorable” and gets interested.
Now there is a negotiation over what is applicable:
Board observer rights or strategic commercial rights would be easier to ring-fence as inapplicable. But many side-letter concessions are not that bespoke.
Step 2: Only some investors elect in
This is where the cascade becomes messy, not just expensive.
Not every earlier investor will make the same election. Some may choose:
The result is not one revised SAFE template. It is a patchwork.
That patchwork increases legal review costs, cap table modeling complexity, and room for disagreement at conversion.
Step 3: The priced round surfaces the issue
Now assume the company raises a Series Seed at a $20 million pre-money valuation, selling $5 million of new preferred.
At conversion:
Since the cap price still beats the discount here, the real economic damage is the lower cap election.
Cap table math
Use a simplified fully diluted capitalization before the Series Seed of 10,000,000 founder/common-equivalent shares outstanding, ignoring option pool expansion for the moment.
Conversion share counts based on cap price are roughly proportional to investment divided by cap.
Illustrative ownership on conversion before the new round:
If nobody in Cohort A elects into the $8 million cap, aggregate SAFE ownership is about:
Now assume just two-thirds of Cohort A holders, representing $600,000 of the $900,000, validly elect into the $8 million cap.
Revised Cohort A economics become:
Now aggregate SAFE ownership becomes:
That 2.5 percentage point increase did not come from raising more money. It came from one concession cascading into prior paper.
And this is before option pool top-up.
What that means in the priced round
Assume the lead wants a 15% post-money option pool and is investing $5 million at a $20 million pre-money valuation.
In a clean model, maybe founders expected to own about 60%+ after the round.
But the MFN cascade pushes more pre-money dilution into the SAFE stack. That means one of three things happens:
All three are painful.
A few points of unexpected SAFE dilution can easily shift hundreds of thousands, and sometimes millions, in implied value away from founders at the priced round.
A more aggressive scenario: rights, not just pricing
The bigger hidden problem is often pro rata.
Assume the same stack, but now the side-letter pro rata granted to Cohort D is also available through MFN to prior investors. Suppose:
Before the Series Seed, these investors convert into roughly 16% to 18% combined ownership, depending on the exact stack. In the next round, each may claim rights to maintain ownership.
If the new round is $5 million and the lead expected to take $4 million with $1 million reserved for insiders and selected existing investors, the math can break quickly.
Suppose MFN-enabled pro rata demand totals $1.8 million. Now the company has to choose:
This is why a single “harmless” side-letter pro rata can become a real financing problem. Pro rata is not merely etiquette. It is allocation economics.
How Series A leads and their counsel diligence MFN exposure
Experienced leads do not rely on the company’s summary.
They usually ask for:
Then they pressure-test three questions.
Question 1: Is the SAFE stack static or dynamic?
Static means the conversion economics are effectively known. Dynamic means rights remain exercisable or disputable.
Leads hate dynamic stacks.
If prior investors can still elect into more favorable terms, the lead may require all elections to be made before signing. In some cases they will require waivers of unexercised MFN rights as a closing condition.
Question 2: Are there hidden rights that reduce round flexibility?
Lead counsel will focus on rights that constrain execution:
Even if these rights do not change conversion price, they can reduce the company’s room to allocate the round and manage process.
Question 3: Can this create post-closing disputes?
If the answer is yes, the lead often discounts the company.
Investors pay for clean execution. They do not want a post-closing letter from a pre-seed angel claiming the company failed to honor an MFN election and therefore misissued shares.
Clean paper increases financing certainty. Messy paper reduces it.
How to model the cap table impact correctly
Most companies model SAFEs too simplistically. They assign each SAFE one cap and one conversion outcome. That is not enough when MFN is in play.
You need scenario modeling.
Build at least four cases
Case 1: Base case
No MFN elections beyond already documented amendments.
Case 2: Economic election case
All plausible holders elect into the best lower cap or discount available.
Case 3: Rights election case
All plausible holders elect into pro rata and related participation rights.
Case 4: Full stress case
All plausible holders elect any favorable applicable term and the company has to honor them.
Example stress model
Start with:
- $900,000 at $12 million cap with MFN
- $700,000 at $10 million cap with MFN
- $250,000 at $10 million cap with pro rata
- $350,000 at $8 million cap plus 20% discount and side-letter rights
#### Base conversion ownership
Approximate pre-round ownership equivalents:
Raw implied SAFE percentages:
#### Stress conversion if all of Cohort A and B can adopt $8 million cap
Then raw SAFE claim becomes:
That is a 5.5 percentage point increase in pre-money dilution versus the base raw claim.
On a $20 million pre-money priced round, 5.5% of pre-money value is economically equivalent to about $1.1 million of founder value transfer. That is not a rounding error. That is a financing issue.
Now add a 15% post-money option pool target and the founder hit grows because the pool expansion is usually borne pre-money.
Model rights separately from dilution
Do not stop at conversion share counts.
Create a second schedule showing by investor:
Then estimate the practical burden:
That second model is what your lead and counsel actually care about operationally.
Negotiation scenarios founders should expect
Scenario 1: Lead says “clean this up before we issue a term sheet”
This happens when the MFN stack is visibly messy.
The company may need to:
This is expensive and distracting, but still better than cleaning it up after signing.
Scenario 2: Lead uses MFN ambiguity to pressure valuation
This is common and rational.
A lead may say: “Because conversion economics are uncertain and there may be additional pro rata claims, our pre-money assumes the worst-case SAFE treatment.”
That is another way of saying the company bears the ambiguity discount.
If your internal model says expected dilution is 22% but the lead underwrites 27%, you are negotiating from different realities.
Scenario 3: Existing investors weaponize the financing timeline
Most angels are reasonable. Some are not.
An investor with an arguable MFN claim may wait until the company is in exclusivity with a lead, then demand side-letter parity or threaten to object. They may not ultimately win on the law, but they can create delay and legal spend.
This is why loose drafting is dangerous even when you think your interpretation is stronger.
Scenario 4: The company is forced into a SAFE exchange
Sometimes the cleanest fix is to exchange the messy stack into one harmonized pre-round instrument. That can work, but it is effectively a mini-recapitalization. It requires investor coordination and can trigger fresh negotiation over who gets what.
Founders should not assume a cleanup exchange is easy just because everyone wants the priced round to happen.
Founder tactics to ring-fence exceptions without poisoning the whole stack
Here is the practical part. If you need to make a concession, the goal is not wishful thinking. The goal is containment.
1. Narrow the MFN trigger itself
If you are issuing instruments over time, broad “any more favorable term” language is asking for trouble.
Work with counsel to define:
Narrow drafting is not anti-investor. It is anti-chaos.
2. Separate bespoke rights from the financing instrument
If a concession is truly strategic or personal, document it in a way that clearly ties it to that investor’s unique status.
Examples:
This does not guarantee ring-fencing, but it makes the “not generally applicable” argument much stronger than handing out generic sweeteners.
3. Explicitly exclude side-letter rights from MFN adoption
Do not rely on implication.
If the company grants a side letter, the documents should say whether prior MFN holders can or cannot elect into those rights. Silence creates litigation fuel.
4. Use sunset windows for elections
One of the worst features of MFN overhang is indefinite optionality.
A much cleaner structure is:
Without a deadline, you may carry election uncertainty into the next priced round.
5. Limit menu-style cherry-picking
If investors can adopt favorable terms, require package adoption rather than provision-by-provision cherry-picking where legally and commercially feasible.
Why? Because cherry-picking creates Frankenstein instruments. Package elections produce fewer variants and cleaner modeling.
6. Be careful with pro rata grants
This is where many stacks get poisoned.
Before granting pro rata to one late SAFE investor, ask:
If the answer is no, do not grant generic pro rata casually.
7. Keep a live rights matrix, not just a cap table
Most companies maintain a cap table and ignore the rights layer.
That is amateur behavior once you have multiple SAFE cohorts.
You should track, by investor:
Your Series A process will be faster and cheaper if you can hand over a clean matrix on day one.
8. Run a pre-diligence paper cleanup before fundraising
Do not wait for the lead to discover problems.
Three to six months before launching a priced round, ask counsel to perform a financing document audit focused on:
This is not glamorous legal work. It is high-ROI legal work.
9. When necessary, buy certainty
Sometimes the right answer is economic.
If one or two investors have colorable MFN arguments that could disrupt the round, it may be cheaper to negotiate a clean amendment, waiver, or exchange than to litigate interpretation during financing. Founders resist this because it feels unfair. Maybe it is. But financing decisions should be made on expected value, not indignation.
10. Stop doing rolling custom paper if you can avoid it
The more time-separated and customized your SAFE issuances are, the greater the drift and the greater the MFN danger.
If you can batch investors into cleaner closes with harmonized terms, do it. Administrative discipline is not cosmetic. It protects financing optionality.
The hard truth about “harmless” concessions
A concession is not measured by the size of the check receiving it. It is measured by the amount of paper it can reprice or re-right through MFN.
That is the key mental model founders need.
A $100,000 side letter can affect $2 million of prior SAFEs. A single pro rata grant can alter next-round allocation dynamics far beyond its face value. A notice right can turn a fast financing into a procedural maze. And a lower cap granted late can transfer real ownership from founders to prior investors who never negotiated for it directly but can claim it through MFN.
This is why sophisticated investors care less about whether a concession seems commercially small and more about whether it is legally contagious.
The core insight
MFN in stacked SAFEs should be treated as a propagation mechanism, not a fairness mechanism.
Once you think about it that way, the right founder behavior becomes obvious:
Founders who ignore this are often shocked when their “simple” pre-seed turns into a renegotiation event at Series Seed or Series A.
But that is exactly what happened. They just failed to notice the renegotiation when they created it.
Takeaways
If you are a founder, the practical rule is simple: before agreeing to any supposedly minor SAFE sweetener, ask your counsel one question in plain English, “If we give this to one investor, who else can claim it, and what does that do to our next round?”
If you do not have a precise answer, you are not making a small concession. You are opening a propagation channel.
RaiseSimpli Team
Author