Editorial standard
How we decide what to publish.
This page exists to answer the question an analytical reader is silently asking: is this actual thinking, or is it generated filler with a founder's name on it. Here is the standard, so you can hold us to it.
What we publish
Arguments with a named claim, a named limit, and named sources. Every piece states what it believes in the first paragraph, gives the strongest version of the objection, and ends by naming where the argument stops applying.
We publish when we have something to say. There is no calendar. A month with nothing worth defending is a month with no essay.
What we do not publish
Trend summaries. Lists of things to know about a topic. Drafts an AI produced that a human then lightly edited. Anything a competitor could republish unchanged under its own name. That last one is the test we apply before anything ships: if the piece would work with another firm's logo on it, the argument is not sharp enough.
How we handle sources
Every statistic links to the most primary source we can find, and the link is checked before publication. When the best available number comes from an agency survey or a vendor benchmark rather than peer reviewed research, we say so beside the number. When we cannot find the original behind a widely repeated figure, we say that too, and we cite the secondary source that carries it.
We also verify our own specs before they become pages. When a source in a draft turned out not to say what the draft claimed, the claim was changed or cut. Several were.
How we handle being wrong
Corrections appear at the top of the essay with the date and what changed. We do not silently edit a number. If a reader shows us a better source, we update the piece and credit them if they want the credit.
Who writes
Real named authors with real bios and real accounts. No ghostwriting under a founder byline. When a team member drafts a piece that George reviews, the byline reads with George Stoff, not by George Stoff. Scientific and clinical content is reviewed by a named reviewer with the training to catch what the writer missed, and the reviewer is credited on the page.
AI runs scheduling, deployment, data collection, and research on this site. It does not write the arguments. That split is the same one we run for clients, and it has to be visible here first.
George Stoff, Founder and Lead Engineer
Thirty years building software, brands, and demand. On every account.