VEN Blog

Round Size Should Decide SAFE or Priced, Not Story

The SAFE-or-priced call is math, not narrative. Once a raise crosses roughly $3M to $5M, pricing the round costs less than pretending it's still a SAFE.

A founder and advisor reviewing SAFE and priced round term sheets together.

TL;DR: the instrument question on a pre-seed or seed round isn't a narrative choice, it's a math problem. Once the raise crosses roughly $3M to $5M, a SAFE stops being the cheap, fast option and a priced round starts being the honest one, regardless of how the deck frames the stage.

Here's what actually gets decided wrong, over and over: a CEO picks the instrument first, based on what sounds better to the board or the next investor, and backs into the round size to match.

Run that in reverse.

Know the dollar figure you need, then let the instrument follow.

Decision one: size the round before you name the instrument

A seed-stage CEO running a 14-person infrastructure company is raising $4M, eighteen months after her last SAFE closed. Her instinct is to keep calling it a SAFE, because "priced round" sounds like a down-round risk to a board that remembers the last valuation conversation going badly. But a $4M raise is not a small top-up. It is a full round that happens to be wearing SAFE paperwork, and the paperwork is the wrong choice for the number underneath it.

a founder and a lawyer reviewing term sheets side by side at a small table The instrument should follow the number, not the other way around.

A SAFE round costs a few thousand dollars in legal fees and closes in days. A priced round runs a full document set, roughly two hundred to two hundred fifty pages, takes four or more weeks, and costs both sides something in the range of $100,000 to $150,000, paid by the company. That gap is exactly why small rounds should stay on SAFEs and why rounds in the $3M to $5M range start to make more sense priced: the legal cost stops being the deciding factor once the round itself is that large, and the protections a priced round gives every investor on the cap table start to matter more than the speed.

  1. Write down the actual dollar figure you're raising before you touch a term sheet template.
  2. Check that figure against the range where pricing starts to make sense, roughly $3M to $5M.
  3. Model your cap at the size you're actually raising, not at the cap your last round used.
  4. Ask counsel for the real cost and timeline difference on both instruments, in writing, before you choose.
  5. Pick the instrument the math supports, then build the investor story around that choice.

Decision two: watch what a bigger number does to the cap

The part that catches CEOs off guard isn't the legal bill. It's what a bigger check does to the cap math on paper. A $1M SAFE against a $10M post-money cap, which is roughly the median pre-seed cap investors are seeing right now, is a clean trade that nobody thinks twice about. Raise $3M against that same $10M cap and the trade stops being clean. You've sold a much larger piece of the company for the same ceiling, which is why a SAFE raise that size typically needs a cap closer to $18M to $20M just to keep the round from quietly becoming the most expensive money the company ever took.

A founder who keeps the cap flat while the round grows isn't being conservative. She's giving away more of the company to avoid a conversation about price, and the conversation doesn't disappear, it just shows up later, bigger, at the priced round that eventually has to true everything up.

A $4M SAFE is not a smaller round. It is a priced round wearing a disguise.

Decision three: know when "still a SAFE" becomes the expensive lie

Stacking SAFEs past the point where the round should have priced creates a second problem that doesn't show up until the first priced round forces a reckoning: every uncapped assumption, every oversized option pool set aside "to be safe," and every stacked instrument converts at once, against a cap table nobody has actually looked at closely in over a year. The CEO who treats each SAFE as a separate, small decision ends up owing more of the company than she modeled, because she modeled each round in isolation instead of modeling where they all land together.

This is the same discipline behind lining up a lead investor before a seed round opens: somebody has to set the real price, on purpose, instead of letting a string of SAFEs set it by accident. And it only works if the company has already done the work of qualifying which investors can actually write the check size the round needs, so the round size itself is grounded in something real before the instrument question even comes up.

Before you pick a term sheet, sit with these

  • Could you state the exact dollar figure you're raising without rounding it to a familiar round-label number?
  • Have you modeled the cap at the size of this raise, not the cap from your last one?
  • Does your board understand the actual legal cost and timeline difference between a SAFE and a priced round?
  • If you stacked every SAFE you've issued and converted them today, do you know who owns what?
  • Are you avoiding a priced round because the math says SAFE, or because pricing feels like admitting something?

The pattern underneath

The CEOs who get this right size the round first and let the instrument follow without apology. The ones who struggle pick the instrument that sounds better in the room, then spend the next raise untangling a cap table that grew by accident. Neither path is about confidence or storytelling skill. It's about whether the number on the page matches the paperwork wrapped around it.

If this is the decision in front of your leadership team, that is what VEN is for: small groups and 1:1 conversations with peers at your own level of experience, focused on the call you are carrying. Can we help? →

Next step

What are you working through?

Have a question before applying? Tell us what you’re working through. If your company qualifies, we’ll reply personally within 24 hours.

For companies that have raised at least $1M from venture firms.