Seed Round

The stretch right after a product exists is the murkiest one. The screens work and a few dozen teams have signed up, but which part keeps them coming back, and who would pay, are still open questions. There is only one way to answer them: change the audience, change the price, add a feature, and do that several times over.

The trouble was that each attempt eats people and time. Two founders taking turns writing code and answering support tickets need two months per experiment. That's six tries a year, and if six isn't enough the money runs out first. A company that failed for lack of attempts and a company that failed because the direction was wrong looked identical from the outside.

A seed round is the money spent on that stretch. It doesn't go into building the product; it goes into finding out where the product fits. So what it buys is not a longer feature list but more attempts. Adding a few people so that four directions can be tried in the same window is what this round is for. Which is why the amount gets sized in experiments rather than in months of salary.

The defining condition is that the product exists and the market is still unknown. The amount, the contract, and the spending all follow from it.

1. The amount and the timeline step up. A wide range runs from roughly $400,000 to $2,300,000, with something near $750,000 being common, sized to cover eighteen to twenty-four months. The number of months current spending leaves before the cash runs out is runway, and since one experiment takes two or three months, the period has to allow at least eight of them. Cut the runway to twelve months and the number of attempts halves, which halves the chance of landing on the answer.
2. The round is usually priced. Revenue and usage numbers now exist, so there is something to value against, and the amounts are large enough that deferring terms gets uncomfortable. This is the round where the pre-money valuation — the value before the new money arrives — gets negotiated properly for the first time.
3. The money comes from institutions. Venture funds that specialize in the seed stage lead, and regular reporting and a board arrive with them. A few individual angels may join, but the center of the round is institutional.
4. Spending goes almost entirely to people. One or two engineers, a designer, and a first salesperson. A team of four becomes something like ten, and advertising budgets wait until it's clear what sells. Those hires are also what raises the monthly burn, so the runway figure has to be calculated on the team as it will be, not as it is.

If the stage has a single goal, it is finding product-market fit — the state where what you built meets real demand in a specific market. Evidence of having found it is the entry ticket to the next round.

Moving to a Series A means showing in numbers that the market has been found. Four things get checked.

1. The size of recurring revenue. Monthly recurring revenue (MRR) is the measure, and somewhere between $60,000 and $75,000 a month comes up often as the bar. One-off setup fees don't count toward it.
2. The rate of growth. The slope matters more than the amount. Ten to fifteen percent month over month, or a record of tripling in a year, clears this one.
3. Retention. Whether customers who arrived are still there. A business losing five percent a month has lost half of a cohort within twelve months, so revenue growth is pouring water into a leaking bucket.
4. Whether the acquisition path repeats. Twenty customers brought in through the founder's network is a method that stops at twenty. The question is whether the same approach carries to a hundred, and the usual check is whether a second channel has produced customers without the founder in the room.

Falling short of all four leaves options. A bridge round takes a smaller amount from existing investors to add a few months, or costs come down to extend runway while the team concentrates on the one item that's missing.

"We raised $900,000 seed and planned twenty-four months. The goal inside that is $75,000 of MRR."

The amount, the time it buys, and what has to be proven inside it, in one sentence.

"No SAFE this time — we're pricing it and issuing preferred."

The valuation negotiation isn't being deferred any further.

"Recurring revenue is where it needs to be, but every customer came through the founder, so a Series A is out of reach."

The numbers work and the method doesn't repeat, so the condition isn't met.

Pagedeck went out for seed meetings in month fourteen of its pre-seed with twelve paying teams and raised $900,000. They planned for twenty-four months and grew the team from four people to twelve.

Half of what they did in that window was narrowing the audience. At first they sold to teams of any size; after more than twenty churned, they found retention was noticeably higher only at companies between thirty and two hundred people. From then on they sold only into that band and reshaped the pricing around it. In month twenty-two they crossed $75,000 of MRR and started Series A meetings on that record.

The most common belief is that a finished product qualifies you for a seed round. Building was hard, so it feels like clearing that hurdle opens the next door. But what the people deciding look at is the behavior of users, not the polish of the product. Three features with forty teams coming back every week is far stronger than twenty features with nobody returning. Spend six months only on polish and the metrics stay where they were, which leaves nothing to say in a meeting. The habit worth building instead is shipping the smallest change that tests something and reading what users do next.

The other mistake is assuming a larger seed round is a better one. More money means more room, which seems obvious enough. But a large amount arrives with a high valuation attached, and that valuation becomes the starting line for the next round. A company that raised $2,250,000 at a stretched pre-money valuation has to reach considerably higher at its Series A, and failing that it raises at a lower price. It also gave up more ownership to get there. Counting the number of experiments actually needed, and raising the amount that funds them, is the safer approach.