The 7% Assumption: Why Founders Overpay in Equity for a Number Nobody Told Them Was Negotiable

Is 7% equity standard for accelerators?

No. It's not a standard, a tax, or a law of physics. It's one number, from one very famous program, that got repeated so often it turned into folklore. Somewhere along the way, "YC takes 7%" quietly mutated into "accelerators take 7%," and first-time founders started writing it into their mental model before they'd read a single term sheet.

That's the whole problem in one sentence: founders are anchoring on a number nobody actually offered them.

Why the 7% number spread in the first place

Accelerator terms aren't posted like a restaurant menu. There's no public price list where you can see that Program A wants 6% for a $125k check and Program B wants 4% for a $100k check with a lower valuation cap. Founders applying for the first time have no comparison set, so they borrow one, usually from whichever accelerator has the most Twitter threads written about it.

That's not a knock on the founders doing this. It's a rational response to an opaque market. If the only data point you have is the most-discussed program in the industry, you generalize from it. The problem is that generalization is doing a lot of unearned work: across the 105 programs in Accelerator Atlas's directory, spanning 14 verticals, equity and check size vary by tier, by vertical, and sometimes by cohort. There is no single going rate, and treating one like it exists means founders stop comparing before they've started.

What actually varies (and why it isn't random)

Four things move independently, and none of them move because of some industry-wide standard: they move because of program tier, vertical, and how the deal is actually structured.

Check size

The cash the program puts in. This ranges by vertical: biotech and climate programs often write larger checks because the underlying R&D or hardware costs more to get to a fundable milestone, while some fintech and generalist seed programs write smaller checks against a lower cap. Check size and equity percentage should always be read together, not separately. A bigger check for the same equity is a better deal on its face. A smaller check for the same equity is worse, and that's the comparison a lot of founders skip.

Equity percentage

The actual ownership stake, usually taken via a SAFE (Simple Agreement for Future Equity), a document that converts to equity at your next priced round instead of pricing your company today. Equity percentage is the number most founders anchor to 7% without checking, and it's the one most worth verifying program by program.

SAFE terms

Not all SAFEs are the same instrument. Some carry a valuation cap (a ceiling on the price your equity converts at), some carry a discount (a percentage off the next round's price), and some carry both. A lower headline equity percentage attached to a low valuation cap can end up costing you more dilution than a higher percentage with a generous cap. This is exactly the kind of clause a technical founder can misread by skimming instead of comparing.

MFN clauses

MFN (Most Favored Nation) means that if you give a later investor better terms than the accelerator got, the accelerator's terms automatically upgrade to match. It's a protective clause for the accelerator, and it's worth knowing it exists before you're three fundraises deep and realize an early SAFE quietly reset itself in someone else's favor.

None of this is a scandal. It's just fine print that varies by program, and fine print you can't compare if you don't know it's there.

The real cost of not checking

Here's the part founders underweight: 2 to 3 percentage points of equity doesn't feel like much in a Zoom call with a program partner. It feels like a rounding error next to the mentor access and the demo day. But equity compounds. That 2 to 3 points, diluted again at your seed round and again at your Series A, isn't a rounding error by the time you exit. It's real money, and in the earlier, more practical sense, it's real runway: months of payroll you didn't have to raise for because you didn't give it away in year one.

The fix isn't suspicion. It's verification. Before you assume a program's terms are "standard," check whether they're actually competitive against others in the same vertical and tier. If a fintech accelerator in your shortlist wants 6% and another wants 4.5% for a comparable check size and a mentor network that's actually real and not just a demo-day mixer, that gap is worth a real conversation, not a shrug.

How to verify terms instead of assuming them

This is where most founders are stuck reading secondhand Twitter threads or a blog post from two cohorts ago. Terms change. Programs update their SAFE templates. A number that was true last year might not be true for the cohort applying this fall.

Accelerator Atlas exists specifically to close that gap. Of the 105 programs in the directory, 102 are verified against the accelerator's own primary documents, not scraped summaries or forum hearsay. Every program shows its source and the date it was last checked, and every one carries a confidence tier: Verified, Secondary, or Unverified, so you always know how solid the number in front of you actually is. Where we don't have a verified figure, we show a blank. A blank you can trust beats a number you can't.

If you're also weighing whether a big-name program's terms are worth the equity premium against a smaller one, that's a related but separate question, and worth reading through before you assume brand name buys you a discount on scrutiny.

What to do before you accept any offer

Don't sign a term sheet because 7% sounded familiar. Pull up every program in your vertical, check size against check size, equity against equity, and read the SAFE and MFN terms side by side. If a lower-equity program offers comparable mentor depth and network access, that 2 to 3 point gap is real leverage, and real leverage is worth a conversation with the program before you sign anything.

Compare check size, equity, and terms side-by-side across all 105 verified programs at Accelerator Atlas before you accept any offer at face value.