Raise Your First Pre-Seed Checks

Pre-seed is not a smaller seed. It is money that buys you time to find a reason to raise a seed. The check is smaller, the bar is lower, and the story is allowed to be mostly team and insight. What it is not allowed to be is vague. "We will figure out the product with this capital" is how you raise a pre-seed you cannot convert into a seed.

A useful pre-seed is 12 to 18 months of runway for two or three people, pointed at one question: is there a customer who will pay for this, repeatedly, for a reason that compounds? If you already know the answer, skip this round and go to seed.

What the first $400K to $800K is for

Typical pre-seed for a B2B software company in 2026 is $400K to $800K on a post-money SAFE, enough for a small team to live and build for a year plus. Less than $300K usually means you are still in friends-and-family territory. More than $1.5M is a seed wearing a pre-seed label, and you will be judged on seed metrics.

Spend it on the few things that produce evidence:

• Founder living costs so you can work on this full-time
• One complementary hire, not five
• Cloud, tools, and the boring security baseline
• Customer conversations, travel, and the first implementation work

Do not spend it on a brand agency, a rented WeWork for eight, or a salesperson before a founder has closed ten customers. Pre-seed sales is founder-led or it is not happening.

Who writes first checks

Ranked by how early they will actually say yes:

The highest-leverage first checks are operator angels who have felt the problem. They decide in one meeting, they intro you to the next five, and they do not need a partner meeting. Pre-seed funds are useful when they lead and set a clean SAFE. They are expensive when they "want to follow a lead" that does not exist yet.

Accelerators are a trade: money, a network, and a deadline, against dilution and a very public demo day. Take one if the network is specific to your buyer. Skip one if you already have customers and a path to a seed list.

What you actually need

Less than you think, more than a slide.

• A Delaware C-corp if you want US institutional money later. Flip before the first SAFE if you can. The US tech startup guide covers the entity decision.
• A one-line description someone can repeat. "We do [X] for [Y]."
• A reason this is possible now. The same why now that carries a seed deck, just shorter.
• Some evidence of pull. A waiting list of strangers is weak. Five conversations with people who have the budget and a date they would buy is strong. A paid pilot is enough to stop calling it an idea.
• A clean cap table. No 15% advisor, no forgotten co-founder, no uncapped note from an uncle.

You do not need revenue. You do not need a 20-slide deck. You do need to be able to say who hurts, why now, why you, and what this money buys, out loud, in sixty seconds.

How the money actually arrives

Pre-seed is usually a rolling close on a post-money SAFE. You pick a cap, you take checks as they come, and you stop when you have the runway. That is a feature. It lets you start building before the last $100K lands.

Rules that keep a rolling close from becoming a mess:

• One instrument, one cap. Do not stack three different SAFEs because each angel asked for a tweak.
• A minimum check, stated out loud. $25K is a common floor. Smaller checks cost the same legal time.
• A date you stop. Rolling forever is a quiet seed raise you never admitted to.
• A lead or a first-close announcement once you have enough to be real. "We have $400K in from operators who have run this function" is a useful sentence. "We're raising" is not.

Friends-and-family money goes on the same SAFE, not on a side letter you will regret in diligence. If someone cannot sign a standard Y Combinator post-money SAFE, they are not the right first check.

When not to raise

Do not raise a pre-seed if you have a salary you can keep, a product you can sell without it, and no use of funds except "optional validation." Dilution you take to feel official is the most expensive confidence purchase in the company.

Do not raise if the only proof you have is that other people are raising. Pre-seed is cheap compared to seed, and still expensive compared to a year of nights and weekends that produces a real customer.

Do not raise from someone whose terms you do not understand because you need the money this month. That is how odd liquidation preferences and advisor-sized grants show up in a seed data room.

What "good" looks like 12 months later

You can point to a customer who pays, a reason they stay, and a path to ten more that does not require you to invent a new product each time. You have a monthly update with the same three numbers on it. You know whether the next raise is a seed or more time at this size.

If twelve months later you have a prettier deck and the same amount of evidence, the pre-seed did not work. Raise again only if the story changed.

Failure modes

1. Treating pre-seed like a tiny Series A. You do not need a TAM slide from Gartner. You need a customer who flinches when you describe the problem.
2. Hiring ahead of evidence. The first extra salary is the one that forces a desperate seed.
3. Custom terms for the first three angels. Your future lead will make you unwind them.
4. Raising from a "strategic" who is actually a customer you have not sold. Sell them first. Then take the check if it is still clean.
5. No end date on the rolling close. You will still be raising in month nine.
6. Skipping the update list. The people who wrote $25K are the warmest path into the seed fund you want next year.