What good actually is.
Written for hosts of interview-style B2B podcasts who have accepted that the generic approach fails, and want to see what the alternative is made of.
The other pages on this site argue. This one shows. Below is what a pitch that lands is built from, part by part, and then the only kind of evidence that settles the argument: a message, the reply it got, and an episode that exists and can be listened to.
The parts, and which one is load-bearing.
A finished pitch carries a title, a premise, questions with their tips, a short line about the show, and an opening built on something the guest said. Four of those five are craft. They can be done well or badly and doing them well is worth the effort. None of them decides the outcome.
The title, the premise, the questions and their tips are the seven countable elements a finished pitch has to contain. They are the acceptance test for whether a pitch is complete. Completeness is not the same as landing, and the difference between the two sits entirely in the band at the top.
A verbatim sentence the guest said, with a link and roughly where in the recording it sits. It is the first thing they read and the only part that cannot be produced without doing the reading.
Specific enough that a stranger knows what they would learn. An invitation to something rather than to a subject area.
A short paragraph answering why these two parties in particular. This is the question a senior guest is actually asking, and it is not answered by describing the show.
Five questions, each with a note saying how you would handle it, so the guest can see before agreeing that answering costs them nothing.
One line. Credentials at the top read as pre-emptive defence, which tells the reader you expected to be doubted.
The part that decides it.
Everything else in a pitch can be produced without knowing anything about the person receiving it. A title, a premise and five questions can be written from a job title and a company description in about a minute. The opening line cannot, because a verbatim sentence with a link and a timestamp either exists or it does not.
That is why it goes first, before anything about you, and why it carries its source with it. The guest sees the work itself rather than a claim about it, in their own words, in the first two lines.
Which sentence you pick is a separate question from whether you have one, and it is where most of the difficulty lives. The Golden Thread play sets out how the sentence is found and how the rest of the pitch is built around it. This page is about what the finished object looks like.
Why the date is one of the slots.
The sentence can be genuinely theirs, genuinely relevant, and still fail, because how old it is decides whether finding it took any work. A quote from the last few weeks works because the thought is still live and they have not finished having it. A quote from far enough back works because reaching it took effort a search engine cannot do for you. Everything between those two is the first thing any search returns, which means every other message that month opened with it.
That is why the date sits in the slot list above rather than as a detail. The full rule, with both windows and the one-minute test for whether you are in the middle band, is on the Golden Thread play.
The part nobody in this category can copy.
A testimonial is somebody saying you are good. This is somebody saying yes, followed by a recording anyone can go and listen to. It is an outcome rather than an opinion, and the difference is that an opinion can be solicited and an episode cannot.
The three parts are not equal and should not be presented as though they were. The reply is redacted and therefore unverifiable, so it is the quietest of the three. The episode is public and checkable, so it carries the weight and it carries the link.
One · what was sent
The pitch as it went out, unedited, so the opening line and its source can be read against everything else on this page.
Awaiting the artefact. This slot holds the real message and is deliberately empty rather than filled with a written example, because a demonstration page cannot demonstrate with something invented.
Two · what came back
The words stay. The name, the handle and the photograph come out. We are declining to publish somebody's private message, which is a different act from concealing who they are.
Awaiting the artefact. Redaction applied at the point of publishing, never reconstructed from memory.
Three · what happened
Public, dated, and listenable by anyone who wants to check. This is the part that carries the proof, and the reason the redaction above is worth anything at all.
Awaiting the episode link. Until it resolves, the two panels beside it are not published either, because they do not stand on their own.
That last point is the one to hold on to. A blurred screenshot standing by itself is the oldest move in the category this site was built to argue with. Anyone can blur anything. It reads as truthful only because a person went to the trouble of protecting somebody real, and that reading collapses the moment there is nothing else to check. The episode is what makes the redaction evidence rather than theatre, so the chain ships complete or it does not ship.
The three panels above are structural and hold no content yet. The real message, the redacted reply and the episode link exist and are not in this repository. Publishing anything written in their place would be the exact failure this page exists to name, so the slots stay visibly empty and the page is not linked from anywhere until they are filled.
Everything above describes a finished object and none of it tells you how to produce one. The reading, the choice of sentence, the two windows, the questions that a guest can answer without exposure, and the judgement at every step are all set out in full on the play page, free to follow, with the services it takes named.
Hadef is in open beta. Applications are reviewed within a few days and small cohorts are admitted every few weeks.