01 · The format
Three pieces, printed together
Every issue runs the same shape. It is not an interview and it is not a profile. It is a builder's account of their own work, our reading of it, and the wider argument it sits inside, published as one thing so a reader can hold all three at once.
The answers
Seventeen questions, written for that company and nobody else. They answer in their own words. We edit for length and clarity and nothing else, and anything marked off record never appears, including in our analysis.
The analysis
Our reading, printed beside their answers and clearly marked as ours. Where a claim can be checked we check it and say how. Where it cannot, we attribute it to them rather than presenting it as established fact.
The related piece
A longer article published alongside it, setting the company against the wider picture: the category, the money, the regulation, and what it means for anybody else building here. That piece is entirely ours.
02 · The rule
You approve your words. You do not approve ours.
Every subject sees the full piece before it is published.
They can correct any factual error in it, and corrections go in ahead of everything else. What they cannot do is remove our analysis because they dislike the conclusion. A piece the subject can veto is advertising, and readers can tell the difference. Saying that out loud, in the letter, before anybody has agreed to anything, is the cheapest possible way to find out whether a company is serious.
There is no fee in either direction. Nobody pays to appear here and we pay nobody to appear. That is worth stating plainly in a region where a great deal of what looks like coverage was bought.
Figures are attributed, not laundered. If a founder gives us a number we cannot independently verify, it is printed as their statement with their name on it. Where we can verify something, we do, and we say how. The difference between those two things is most of what this publication is.
Nothing is quietly edited after the fact. A published piece keeps its date. Corrections are added openly underneath it, because an archive you can silently rewrite is not a record.
The subject owns their words. We own the reading. Neither of us gets both.
03 · The register
In commission
Every issue is numbered by sector before it is written, so the count is never guessed at publication and nothing quietly disappears. An issue that is declined stays on this list marked declined, because a register that only shows successes tells you nothing.
Status is updated when a subject replies. Nothing on this list is published until the piece is.
04 · How a subject is chosen
Three things, and none of them is size
The work has to be real and openable. Same standard as the rest of this site. A product somebody can download, buy, sign into or read the documentation for. A deck is not a subject.
There has to be a hard question we genuinely want answered. Not a launch to announce. If the only honest question is what does it do, it belongs on the record rather than in a piece.
The builder has to be findable and willing to be named. Anonymous work can sit on the record with its evidence stated. It cannot carry an interview, because there is nobody to hold to it.
Being early counts for more here than being large. The register is a record of who was building when almost nobody was watching.
05 · Put yourself forward
Building something worth asking about.
Pitching yourself is not a conflict of interest. Hiding that you pitched would be. Tell us what you are building and, more importantly, what the hard question about it is. The pieces that work are the ones where the founder already knows what they would rather not be asked.
06 · Corrections
Something published here is wrong. Tell us.
Corrections are the most useful message this page can receive and they go in ahead of new work. They are added openly underneath the piece with the date, never by quietly editing what was already read.
eye@cedro.club