How much revenue do you need to raise a seed

Horus Labs, August 2026.

Everything you put in front of a fund sorts into two piles. Things a stranger can check without asking you, and things they'd have to take on trust.

The second pile gets discounted to near zero, whatever is sitting in it.

That's the whole mechanism, and it's why an amount can't be the criterion. The same number lands in either pile depending on what's behind it. Forty grand a month from signed contracts you're delivering is checkable. Forty grand a month you can only explain is not, and it gets treated like the forecast next to it.

So the useful question is which pile yours lands in.

Funds say this out loud more often than founders expect.

An offer priced against customers you haven't closed yet is a fund saying it likes the story and won't pay for the part it can't check. Same with a term sheet carrying a milestone, or a partner who wants to speak to two of your users before the second meeting.

That isn't doubt about you. They just can't pay for the half they'd have to take on trust.

Confidence arrives before evidence, and feels identical to it

Over fifty founders in the first half of 2026 asked us about raising a seed while three months or more from a first paid customer.

Almost none of them asked whether they were too early. Not one, some weeks.

The teams with money actually arriving ask us that unprompted.

That's backwards, and it has a cause. To sell a story to a fund you have to sell it to yourself first, and hard, because the alternative is folding by the end of the month. So you go and collect reasons. Then the reasons you collected convince you.

You decided, then you assembled the case, then you were persuaded by your own case.

By the first meeting the conviction and the evidence have been in one head so long they've stopped feeling like different objects. None of it is visible while it's happening. From the inside it just feels like you finally understand your own company.

The fund is running a different job. It's deciding whether a cheque now is worth more than the same cheque somewhere else. Your conviction doesn't help with that, because every founder it meets has plenty.

What survives being checked

Things with a counterparty attached:

• signed deals you're already delivering against
• invoices out of your own billing system
• a customer who will take a fifteen minute reference call
• a pilot somebody paid for, even badly

Things that don't survive:

• a forecast
• a waitlist
• a letter of intent
• a partnership you announced but never shipped
• a pipeline number
• strong interest from anyone

Anything with a counterparty attached survives contact with diligence. The rest reads as intent.

You can run a hundred user interviews and a fund reads a hundred user interviews. This is why founders quoting their own customers land harder than founders describing themselves.

A founder repeating what a customer told him last week has a person standing behind the sentence. You can't quote a customer you never had.

A paid pilot also beats a large vision, and credibility has nothing to do with it.

A vision invites the question "what if this doesn't work". The founder's answer to that question is what does the damage. Under pressure the scope widens. We'll also do this, and this. Now there are two more plans and nobody can check any of them.

A pilot that already happened doesn't leave room for the question.

Then whether it comes back

Checkable is the first question. Whether it happens again is the second, and that one gets skipped constantly.

Plenty of revenue goes from nearly nothing to a real number and then nobody involved can say why it would happen twice. One good month. One large customer. One launch. It landed, it's checkable, and it says almost nothing about next quarter.

Subscription makes this easier without settling it. Annual contracts, month to month, and a free trial nobody has renewed yet all show up on the same MRR line, and they get read as different companies.

The strong version is being able to describe how the money comes back. Renewal on a date you can name. Repeat purchase with some rhythm to it. The same accounts buying more than they did last time.

Even a rough account of that beats a larger number with nothing attached to it. Most founders report the total and skip the return path, and the return path is the thing being bought.

Who pays you decides what proof costs

This is about software and anything else that ships without a factory. If you're building robotics or hardware, nobody expects revenue next quarter and none of this applies to you.

Selling to individual consumers is where this hurts most, because the proof is expensive and the money to buy it is exactly what you don't have yet.

You need real spend to reach a scale that means anything. Your ability to raise that spend is capped by not having reached it yet. It's one wall and you're standing on both sides of it.

So the consumer pitch usually arrives asking for the biggest number in the room. The founder has correctly worked out what the plan costs. He worked it out at the one moment nobody will fund it.

Talking to your first users is how the product stops being wrong. It's the right thing to be doing. It also goes in the deck labelled traction, where it gets read as what it is.

Selling to companies, the same proof is cheap.

Three paid pilots across three months, all delivered, and you've answered three questions at once:

• somebody wanted it enough to pay
• you could deliver it, and delivering is a separate skill from selling
• you've been through onboarding more than once, so the next one costs less

Nobody is impressed by the amount. It's tiny. Three separate counterparties made a decision about you and it went the same way each time. That's the part being read.

Each one also leaves something behind. A signed document, an invoice, a person who will take a call. After three months you're holding most of a diligence pack without having sat down to write one.

So, the number

There isn't one. An amount was never the thing being measured, so two companies at the same monthly number get opposite answers.

Advisors who hand out thresholds are selling you a threshold. Most funds run stage guidance and ownership targets instead, and those move with the deal.

We had this wrong ourselves for a while. We used to weight how a project looked, the site, the presence, whether the marketing hung together. It counts for a lot less than we thought it did.

Being visible to funds matters and that's a separate job from having a brand in front of them. Early on nearly all of it should come from the founders.

A founder in front of investors and in front of customers does more than the company being in front of either, and costs less. If search is genuinely your channel the maths changes. Most teams building a brand at this stage aren't doing it for search.

That also changes what the quarter before a raise is for. Growing the number gets you there sometimes, when the growth happens to leave a trail behind it.

Going after the checkable things directly gets there more reliably, and it's usually cheaper.

Founders who ask whether their own revenue is enough have already started sorting their own piles. That's the same sorting the fund does, they just do it later and with your quarter.

Back to work.

TL;DR

• Everything you show a fund sorts into checkable and take-my-word. The second pile gets discounted to near zero
• No amount clears the bar, because the amount was never what was being measured
• Over fifty founders in H1 2026 asked us about a seed while three months or more from a first paid customer. Almost none asked whether they were too early
• You decided, then assembled the case, then were persuaded by your own case. That order is why certainty arrives before evidence
• Consumer products need the most expensive proof at the moment they can least afford it
• Three paid pilots in three months is three counterparties who decided the same way, and that's the part that gets read
• Checkable is the first question. Whether it repeats, and whether you can say how, is the second. That one gets skipped
• The quarter before a raise is best spent producing things a stranger can check. Building a brand is a different job and most teams do it too early