Named human writers, one recorded expert interview minimum per piece, no model-written sentences, French written in French rather than translated.
About this service
Every article carries a named writer and at least one thirty-minute interview with somebody at your company. We publish nothing assembled from the pages already ranking, which is why we produce a small number of pieces at a high price rather than the reverse. A piece runs two to three weeks end to end, and most of that time is not writing.
Who writes:
French pieces are written by French writers in French. English pieces are written in English. We do not translate between them. Where a topic runs in both markets the two versions get different examples, different sources and often a different structure, because the two result sets reward different pages. That is two commissions, priced as two, and we say so before you have to ask.
How a piece is built:
Brief, then interview, then a source pass across product documentation, changelogs, regulator publications and your own support tickets, whatever the claim actually rests on. Then a draft that goes to your internal expert before it comes to you, so factual corrections happen while they are still cheap. Then two editing rounds with you. Screenshots, code samples and data tables are produced as part of the piece rather than requested from your team afterwards.
On models:
No language model writes a sentence that reaches the page. We use one for interview transcript cleanup and for nothing else, and we will put that in writing if your legal or brand function needs it on record. We also do not run drafts through detection or humanising tools, because there is nothing there to detect.
Developer-tools writing:
Code in an article is code we have run. If a piece includes a snippet, the writer executed it against the version named in the text, and the article states that version. When your library ships a breaking change we tell you which of our pieces stopped being true. That flagging continues for anything we wrote, indefinitely, whether or not the retainer is still running.
Editing rounds:
Two rounds are included and they are real rounds rather than typo passes. A third happens when the brief was wrong, and we do not charge for it, because the brief was ours. What we will not do is rewrite a finished piece toward a position the interview did not support. If the expert's claim is inconvenient, the answer is another interview rather than softer prose.
Outside scope:
No publishing into your CMS, no image licensing, no social copy, no email versions, and no ghostwritten bylines for executives who were not interviewed. We do not write in languages we cannot read, which rules out everything except English and French, and we will not broker a translator to work around it.
Do not buy this if:
You need pieces at a cadence with no room for interviews. You want a fixed monthly count more than you want these specific pieces. Your approvals run through a committee with no named owner, since three rounds of collective editing removes exactly the specificity you paid for, and we have watched that happen often enough to stop pretending otherwise. And if the topic can be written correctly without talking to anyone at your company, buy it somewhere cheaper. That is a recommendation, not a negotiating position.
Rights and handover:
Full transfer of rights on payment, with drafts and transcripts included. Every piece arrives with its sources listed and the interview date attached, so a reader two years from now can tell how old the claim is and your team can tell when to re-source it.