“Copyedit.” That is the entire edit summary left on a revision an agency was paid several thousand dollars to make — accurate, as far as it goes, and silent on everything Wikipedia's Terms of Use actually require it to say. The policy obliges anyone editing for pay to disclose the employer, the client, and the relationship behind the edit, and it does not leave the method open-ended. It names exactly three places that satisfy it: a statement on the editor's own user page, a statement on the talk page of the article touched, or a note inside the edit summary itself. Meeting any one of the three meets the letter of the rule. Which one an editor actually reaches for decides whether anyone besides the editor ever sees it.
The user page is the one most paid editors pick first, because it is written once and left alone — a line naming the employer, the client, the date, sitting on a page almost nobody opens unless they already suspect something is off. A reader working through a company's Wikipedia article, or a volunteer scanning its talk page for sourcing problems, has no reason to click through to a username, and no way to know there is anything behind it worth clicking through to.
The talk-page statement travels with the article rather than the editor. Convention places it in a template, {{connected contributor (paid)}}, pinned to the top of the article's own talk page, naming the employer and listing the edits made on its behalf. Anyone who opens that page for any reason — to propose a change, to question a source, to review what happened to a paragraph — finds the disclosure in the same place they would look for anything else concerning the article. Of the three, it is the version most likely to be read by someone who was not already looking for it.
The edit summary is the thinnest of the three, and the one most easily left blank or filled with a word like the one this piece opened with. It predates, for most editors, any awareness that paid work needs disclosing at all, which is exactly why it keeps showing up alone. It is also the first of the three an administrator or a new-page patroller is likely to see, because it prints directly into the revision history, with no click to a user page or a talk page required. A one-word summary sitting where a disclosure sentence belongs is usually the first gap a reviewer notices, not the last.
Disclosed where, exactly? Nothing on this talk page, nothing in the edit summaries.
An account can satisfy the policy on paper and still read, to a reviewer, as undisclosed. A statement sitting alone on a user page nobody consults does the same procedural work as no statement at all — until an investigation goes looking for it, at which point the silence everywhere else starts to look less like an oversight and more like a choice: disclosed quietly enough to avoid being found, rather than clearly enough to avoid the question.
We disclose in all three places, on every account, for every client, on every edit that touches an article. Not because the policy asks for it twice over — one place would satisfy it — but because the point was never to clear a legal minimum. It was to make the relationship impossible to miss, however a reader happens to arrive at the page.
