Method
Most investor databases print what funds say about themselves. We print what we watched them do, and we would rather tell you the size of our blind spot than let you assume there isn't one.
We read 14 funding publications every morning, extract the company, stage, amount, date and every named investor from the article text, and keep the link. When several outlets report the same round, it collapses into one record with several sources — so a fund that made one investment never looks like it made four. Our record starts Sept 2024.
Some outlets block automated reading. We do not work around that, so their rounds only reach us when another publication also covers them.
We measure it rather than guess. Take two publications covering the same country: if one reported 500 rounds, another 400, and 100 of those are the same round, each is catching a knowable fraction of the whole — the arithmetic used to count wildlife populations. Measured per country, because outlets specialise regionally.
Read that as the ceiling, not the floor: roughly half of the deals these outlets themselves cover, and the outlets do not cover everything. Rounds that were never announced — most family-office, LP and private-credit activity — are invisible to everyone, us included.
Silence in our record is usually our blind spot rather than a fund's inactivity, so one missing signal is never enough. A firm is only marked as appearing to have stopped when three independent things agree: we have observed at least four cheques from it before (so its deals do get announced), we have seen nothing for more than eighteen months, and its website is dead or parked.
For the types that almost never announce — family offices, institutional LPs, wealth and asset managers, private credit, hedge funds, funds of funds, investment banks, government vehicles — we never infer dormancy at all. Their pages say so in as many words.
Purchased and third-party lists are treated as leads, not facts. We fetch every listed website ourselves and record what we found, then read the site to decide what the firm actually is — which regularly contradicts the list it came from.
Where a website never mentions the firm it is attached to, we treat the link as wrong, publish nothing read from it, and say the website is unconfirmed rather than quietly showing a stranger's homepage.
We do not publish or sell personal contact data — names, emails and phone numbers stay out of this product entirely. We do not resell third-party lists as they shipped. We do not fill blanks with estimates: an unknown is printed as unknown, and a figure we did not observe is not printed at all.
Every figure on this page is generated from the live database, not written by hand.