The SAFE was designed to make early fundraising fast and cheap, and it does exactly that. A founder can raise $250,000 in a week on a five-page document with no board approval, no interest rate, and no maturity date. The trouble is that the same speed hides the consequences. Nothing happens when you sign a SAFE. The dilution, the ownership math, and the arguments all arrive at the priced round, sometimes two years later, and by then the terms are fixed. Almost every painful cap table conversation in early hardware starts with a stack of SAFEs signed quickly by a founder who had not run the arithmetic.

If the instrument itself is new to you, start with the mechanics of how a SAFE agreement works and the comparison of a SAFE against a convertible note. What follows assumes you know the shape and are about to sign one.

Mistake 1: Not Knowing Whether It Is Pre-Money or Post-Money

This is the single most expensive misunderstanding in early-stage fundraising. The original SAFE used a pre-money valuation cap; the version most investors use today is post-money. The difference is not cosmetic.

Under a post-money SAFE, the investor's percentage is locked at signing and does not shrink when you sell more SAFEs. Every additional dollar you raise on a new SAFE dilutes the founders, not the earlier investors. Under the older pre-money version, later SAFEs diluted everyone including the earlier holders.

Work an example. You sell a $500,000 post-money SAFE at a $5M cap. That investor owns 10% at conversion, guaranteed. Three months later you sell another $500,000 at a $6M cap. That second investor owns 8.33%. The two SAFEs together are 18.33% before your Series A investor has bought anything and before the option pool is created. Founders who assumed the second raise would be dilutive to the first are usually shocked. The concept is worth learning cold through the pre-money and post-money valuation distinction, because it governs the entire calculation.

Mistake 2: Stacking SAFEs Without Modeling the Total

Hardware companies raise in bursts because tooling, certification, and first production runs arrive as lumpy expenses. So the pattern is three or four SAFEs at rising caps over eighteen months, each one signed in isolation. Nobody ever adds them up.

Build a conversion model before the second SAFE, not before the priced round. A minimal one has four inputs and one output:

  • Every SAFE: amount, cap, discount, and whether it is pre- or post-money.
  • The priced round size and pre-money valuation you expect.
  • The option pool the new investor will demand, and whether it comes out of the pre-money.
  • Your current fully diluted share count.

The output is founder ownership the day after the round closes. If that number is below 50% for two founders at a seed round, you have a problem that will only compound. Keeping a clean cap table from the first check makes this a ten-minute exercise instead of a crisis, and the underlying arithmetic is laid out in how much each round costs you in dilution.

Mistake 3: Taking an Uncapped SAFE Because It Sounds Founder-Friendly

An uncapped SAFE with no discount looks like the best possible deal: no valuation set, no ceiling on your upside. In practice, uncapped SAFEs are usually accompanied by a most-favored-nation clause, which entitles that holder to the best terms you give anyone later. Sell a $4M-cap SAFE to the next investor and the MFN holder converts at $4M too, having taken none of the risk of setting a number.

Worse, uncapped paper makes the next round harder. Institutional investors want to see what the early money paid. A stack of uncapped MFN SAFEs tells them nothing and signals that the founder avoided a hard conversation.

Mistake 4: Ignoring the Side Letter

The SAFE itself is standard. The side letter is not, and that is where the terms that matter get inserted. Watch for:

  • Pro rata rights that let a small early check hold a large allocation in your Series A, crowding out the lead you actually want.
  • Information rights requiring monthly financials, which is fine, or requiring them in a specific format that costs you bookkeeping hours every month.
  • Board observer seats granted to a $50,000 check. Observers attend everything and are extremely hard to remove. Think through what a board seat or observer right actually gives an investor before you concede one at this stage.
  • Right of first refusal on future rounds, which functionally lets one small investor slow or block your next raise.

Mistake 5: Selling SAFEs to People Who Cannot Legally Buy Them

A SAFE is a security. Selling one to a non-accredited investor without a valid exemption creates a rescission right, meaning that investor can demand their money back, potentially years later, and it is a diligence flag that can stall a priced round. This bites hardest in friends-and-family rounds, where the money often comes from people who do not meet the income or net worth thresholds. Verify accreditation and keep the documentation. If you need to take money from non-accredited backers, there are structured routes such as Reg CF and Reg A+ offerings that were built for exactly that.

Mistake 6: Treating Legal Review as Optional

Founders skip counsel on SAFEs because the document is short and the template is free. A startup attorney will review a SAFE and side letter for roughly $500–$2,000, which is trivial against the tens or hundreds of thousands of dollars of equity a single misread clause can cost. Budget for it as part of the legal costs of a funding round rather than treating it as an expense to avoid.

A Short Checklist Before You Sign

  1. Confirm in writing whether the cap is pre-money or post-money.
  2. Add this SAFE to the model with every prior one and read the founder ownership number at the priced round.
  3. Read the side letter twice. Strike anything granting control, board access, or blocking rights.
  4. Confirm the investor is accredited and file the documentation.
  5. Compare the cap to what a priced round would realistically value you at. If the cap is far below it, you are selling equity cheaply for speed.
  6. Have counsel read the final version, not the template.

Projects House works with hardware founders on the technical and cost evidence that supports a defensible valuation, so you are negotiating a cap from a position of proof rather than optimism. If you are about to raise, describe your product and stage through our contact form.