Most funders never say how much they invest: what 5,360 of their websites told us

Most funders never say how much they invest: what 5,360 of their websites told us

Our catalogue lists 7,697 venture capital firms, investors and investment funds. In the middle of September we read their websites, looking for one sentence: how much the firm puts into a company. Of the 5,360 sites that answered, we found that sentence for 360 firms — about one in fifteen. And where firms did say it, the words founders are taught to use for the size of a round turned out to mean almost nothing. Firms that tie a figure to "pre-seed" on their own pages range from $50,000 to up to $10 million.

This piece is about what those 360 statements say, what the other 5,000 sites do not, and how a founder should read a funder's website given both.

What we read, and how we counted

The research was run for our funder rankings on 16 September 2026. For each funder listing it fetched the website we hold, honoured the site's robots.txt, and read up to seven pages: the home page and up to six of the firm's own pages most likely to state its focus — approach, thesis, programme, apply, contact and about. 888 sites refused an automated reader and were not read at all. 1,449 did not answer or returned an error. That left 5,360 sites that answered, and on 1,992 of them we could read only the home page.

An automated extractor looked for statements of cheque size and found 289. For this piece we ran a second, broader pass over the same pages and it found 160 more that the extractor had missed, mostly figures laid out as labels on a page ("Check size $1M–$3M") rather than written as a sentence. We then read all 449 by hand.

Sixty-eight of them were not a cheque at all, and 21 were a second listing sharing a website and a statement with one we had already counted. That leaves 360 firms that state, in their own words, what they put in: 220 filed as venture capital firms, 139 as investors and one as an investment fund.

Two things about that number matter before anything else is said about it.

First, it is a floor. We read at most seven pages per site, and on more than a third of the sites only one. A firm that states its cheque in a downloadable deck, a fund page we did not reach, or a figure drawn on the page by a script is not in the 360. The true share is higher.

Second, it is still small. Even allowing generously for what we missed, most of the funders in our catalogue do not put the one number a founder most needs anywhere a founder would look first. The stage a firm invests at is far easier to find, because it is a word rather than a commitment.

No business we own is a funder listing, so none of our own listings is in any of these counts.

More than one figure in seven is not a cheque

The first thing the hand-reading taught us is the thing a founder skimming a funder's site most needs to know: the prominent number on the page is often not the one they want. Of the 449 candidate sentences, 68 — more than one in seven — contained a figure that was not what the firm invests in a company. They fell into a handful of kinds:

• History and anecdote (20). "Our initial investment" in a story about a portfolio company, or the year of a firm's first investment. The word "initial" appears, but the figure is about the past.
• Totals (13). The size of a fund, assets under management, the total a programme has committed across many companies, or the total a firm has invested since it began. A fund size is not a cheque, and a founder who reads it as one will badly misjudge the conversation.
• Minimums for the firm's own investors (10). The amount an individual needs to join an angel group or a syndicate, the smallest subscription to a retail fund. These are addressed to people giving money to the firm, not to companies receiving it.
• Round sizes (3). The size of round the firm likes to join, which is not what it will write into that round.
• Criteria about the company (6). Revenue, earnings or valuation thresholds, and in one case what the firm's co-investors might add.
• Unreadable or unclear (12). Counters that a script fills in after the page loads, so the text reads "€0M"; the options on a form; a sentence that is rhetorical rather than a statement of policy.
• Not the firm at all (4). Two listings whose websites now carry another firm's name, and two that are not funders: one is an airline and one domain now carries a betting blog.

That last group is about us rather than them, and we return to it below.

"Pre-seed" is not an amount

Founders are taught a ladder of round names — pre-seed, seed, Series A — as though each were a size. The 360 statements show how little the words carry.

Forty-eight of the firms tie a figure to pre-seed — or to a span of stages that starts there, or to an earlier stage they call idea, inception or pre-incorporation. At the bottom, 1517 Fund says its first cheques start at $50,000 at the idea stage, and Great Oaks and Geekdom both start at $50,000 for pre-seed and seed companies. At the top, Amity Ventures writes $1–5 million first cheques "usually pre-incorporation, pre-seed, or seed", and Valhalla Ventures invests up to $10 million from inception to seed. That is a factor of two hundred between two firms using the same word.

Take only the firms that give a figure for pre-seed alone, and the range is still wide. Sputnik ATX writes $100,000 at pre-seed. Catalyst Fund invests $200,000. Bee Partners puts its pre-seed at $300,000 to $500,000, Boost VC at $500,000, and SOSV's pre-seed cheque runs up to $550,000. Tidal Ventures writes $500,000 to $1.5 million at pre-seed, Necessary Ventures leads pre-seeds of $1–2 million, and Heavybit's typical first cheque in a pre-seed round is $500,000 to $3 million. Between the smallest and the largest pre-seed-only figure there is a factor of thirty.

Series A is no tidier. Cento Ventures puts its initial investment at US$1–2 million and describes that as normal for a Series A in Southeast Asia. TheVentureCity writes $1–4 million for Series A companies. EOS Venture Partners leads Series A rounds with initial cheques of $7–10 million.

None of these firms is being inconsistent. Each is describing its own market, its own fund and its own ownership target, and several say so. The point is that the label does not transfer. A founder who hears "we do pre-seed" has learned almost nothing about whether the firm writes a cheque the size of their round. A founder who has read the firm's own figure has learned a great deal.

The programmes are the only ones that name a price

One group of funders is far more specific than the rest: accelerators and structured programmes. Their statements are usually a single figure, often with the equity it buys stated beside it, and the same for every company.

Y Combinator says it invests $500,000, four times a year. Techstars invests $220,000. Entrepreneurs Roundtable Accelerator invests $150,000 on a post-money SAFE for 6% of the company, and states both numbers in the same sentence. AngelPad invests $120,000 in each company. Forum Ventures' accelerator invests $100,000. Bethnal Green Ventures invests £60,000 upfront in each company it selects. Antler's Indonesia page says it decides at the end of six weeks whether to invest $85,000. Hustle Fund's first cheque is $150,000.

This is the model the rest of the market could learn from. A programme can publish a price because it sells one product. A fund has more discretion, and some of it is real. But a range — "our first cheque is usually between $350,000 and $1.25 million", as Seedcamp puts it — costs a fund nothing to publish and saves every founder outside the range a meeting. The firms that do this are not giving anything away that a first call would not reveal.

Ceilings, lifetimes and first cheques

Even among the 360, a figure is not always the figure it looks like. Three distinctions change what a founder should take from it.

A ceiling is not a typical cheque. Forty-two of the 360 state only a maximum: "up to £10m in our first cheque", "seed checks of up to $10M", "up to $1M in companies". A ceiling tells a founder the most the firm has room for. It says nothing about what it normally writes, and a firm quoting a £10 million ceiling may write most of its cheques far below it. Treat a ceiling as a question to ask, not an answer.

A lifetime figure is not a first cheque. Nine firms give only the total they might invest in one company over its life. VTF Capital invests "$300,000 to $2 million in companies over their lifetimes"; Playfair Capital has the capacity to invest up to £5 million per company. Both are useful, but neither tells a founder what arrives in the first round. A further 54 firms mention follow-on money or reserves in the same sentence as their cheque, which is the most honest form of the statement. Correlation Ventures' is a model of it: anywhere from $100,000 up to $4 million in initial investments, and up to $10 million over the life of a company.

"First" and "initial" are the words to look for. 167 of the statements say explicitly that the figure is a first, initial or entry cheque. The rest leave it to the reader, and a founder should not assume.

The currency matters too, and it is usually but not always the firm's home currency. Of the 360, 279 quote a figure with a dollar sign, 59 in euros and 20 in pounds, with one each in Canadian dollars and Brazilian reais. A dollar sign on a page in Singapore, Sydney or Toronto may not be a US dollar, and one Canadian fund states its figures in Canadian dollars. Where the page does not say, ask.

What the reading told us about our own catalogue

Hand-reading 449 sentences also meant looking closely at the listings they came from, and some of what we found is our problem to fix.

Twenty-one of the listings carrying a genuine statement shared a website, and the same statement, with a listing we had already counted. Some are firms that renamed themselves and were imported under both names: VegasTechFund and VTF Capital; Nokia Growth Partners and NGP Capital; Tengelmann Ventures and TEV, both of whose websites now lead to Cusp Capital. One is a programme listed separately from the firm that runs it. The rest are one website under two listings with slightly different names. Among the listings with a genuine cheque statement, one in eighteen duplicated another, and there is no reason to think the rate is lower across the rest of the funder catalogue.

Six more were not what they claimed to be. Two listings for predecessor firms of a large asset manager both led to the same retail-investor page. Two led to websites that now carry a different firm's name. One was an airline. One funder's domain now carries a betting blog.

We have flagged all of them for review. Nothing has been merged or archived yet, because merging two listings changes which page a link reaches and that is a decision for the catalogue, not for an article. In the meantime the counts above are of firms, not listings, which is why 381 listings with a genuine statement became 360 firms.

It is worth saying plainly why this matters. A directory's rankings are only as good as its rows. A firm listed twice appears twice; an airline listed as a funder takes up a place a real funder should have. Reading a sample closely is how you find out, and we would rather publish what we found than leave it to a founder to discover.

How a founder should read a funder's website

The research suggests a short method, and none of it depends on us.

Look for the sentence, not the stage. The stage a firm names is a starting point. The figure it names — if it names one — is what tells you whether your round is the right size for it. If a firm's site gives a stage and no figure, the first question on the first call is the figure.

Check which figure you have found. Before relying on a number, work out whether it is a first cheque, a typical cheque, a ceiling, a lifetime total, a round size or a fund size. More than one in seven of the figures we found were not a cheque at all. A fund size is the easiest to mistake and the most misleading.

Read the words "first", "initial" and "reserve". A firm that distinguishes its first cheque from its follow-on money is telling you how it plans to behave in your next round, which matters as much as the first cheque.

Match the market, not the label. A $1 million pre-seed in San Francisco and a $100,000 pre-seed in Austin are both real, and both firms may be right about their markets. Compare your round with the firm's own figure, not with the name of the round.

Treat a programme's price as a price. Accelerators state their terms because they are standard. If a programme's page says an amount and an equity share, that is what is on offer; the negotiation, if any, is elsewhere.

Confirm on the day. Funds change strategy between vintages, and several of the sites we read had changed name or owner. Every figure we quote here was re-read on the firm's own page on 22 September; a founder should do the same before a meeting.

Our checklist for doing that before a first meeting is at Checking an investor before you pitch, alongside What a pitch has to answer. The pre-seed venture capital firms and investor rankings list the firms whose stage we have evidence for, and our funding guides for the UK and the US, and for accelerators worldwide, set out the public schemes and programmes, whose terms are published by design.

Method and limits

The sample is every funder listing in our catalogue on 16 September 2026 — 5,485 venture capital firms, 2,161 investors and 51 investment funds — and the text is what their own websites said on 16 September. Every figure quoted by name was re-read on the firm's own page on 22 September and was still there. No review site, database or aggregator was used as a source, and no figure here comes from anywhere but the firm's own page.

A statement counts when a firm says, in its own words, what it invests in a company: a first or initial cheque, a typical cheque, a range, a ceiling, or a lifetime total, each counted and reported separately where it matters. It does not count when the figure is a fund size, a round size, a minimum for the firm's own investors, a company criterion or a piece of history. Where two listings shared a website and a statement, they were counted once. The judgement for each of the 449 candidate sentences was made by reading it, and where a sentence was ambiguous it was excluded rather than counted.

The count is a floor, for three reasons stated above: the robots.txt refusals, the seven-page limit, and figures that only appear once a script runs. We did not use a browser to render pages, which would have found more.

We deliberately did not record, compare or publish round sizes, fund sizes or deal counts, and nothing here should be read as a statement about how much is being invested in any market. It is a statement about what funders choose to tell founders, and most of them, so far, choose not to say.