A Community Note on a brand post arrives without warning, and the first questions are practical ones. Who wrote it, what decides whether people see it, and is there anything the account can do? Most of what circulates about this is commentary. This post sticks to what X’s own Community Notes documentation says, and says plainly where it goes quiet.
Why a primary-source reading, and what could not be fetched
This is a desk reading of published documentation, checked on 2026-10-16. There was no testing, no accounts were examined, and no practitioners were consulted. The process may have changed since that date, and the ranking page itself says the algorithm is under continuous development.
Two access problems shape what follows. The X Help Center page on Community Notes (help.x.com/en/using-x/community-notes) returned HTTP 403 to automated fetching, in the research stage and again for the writer, so nothing below relies on it. The pages on communitynotes.x.com resolved with HTTP 200, but they are rendered in the browser and the HTML served to a script contains no article text. What could be read is the Markdown source that X publishes in its public repository, in the documentation folder that, per the repository README, is used to generate the documentation site. Every claim below is tied to one of those Markdown files, which resolved with HTTP 200 and were read in full. Where I could not find a statement, the text says so.
How a note comes to exist
The ranking page states that “Community Notes are submitted and rated by contributors.” The additional review page describes them as “volunteer contributors on X.” Nothing in the pages read here describes notes as written by X staff, so the safe reading is contributors, with X running the program and the algorithm.
The signing-up page lists three eligibility requirements. The account must have no recent notice of violations of X’s Rules, must have joined X at least 6 months ago, and must have a verified phone number from a trusted carrier that is not associated with other Community Notes accounts. The page adds that X may “remove or modify these criteria,” and that when applicants outnumber slots, accounts are admitted at random. It also says AI note writers enroll separately through an AI Note Writer API. The FAQ says any post can receive a note.
Whether the author of a post can also be a contributor and rate notes on it: the pages read here do not say. Contributions are anonymized under aliases, which the FAQ gives as a reason contributors can write and rate without sharing their usernames.
What decides whether a note is shown
The documented principle is bridging. The notes-on-X page says a note starts showing only “if enough contributors from different points of view rate a note as helpful.” The ranking page describes using matrix factorization to find notes “liked by people who normally disagree.” Ratings from contributors who usually agree with each other are not enough. The pages give thresholds and minimums for the technical version, and I have not reproduced them, because they change and the repository is the place to read them.
| Status as the documentation names it | What the documentation says it means |
|---|---|
| Needs More Ratings | Where every note starts. It is shown to contributors to gather ratings, and it is not displayed on the post. |
| Helpful | Enough contributors from different points of view rated it helpful. The ranking page says that at present only notes marking a post as “potentially misleading” that earn this status are eligible to display on posts. |
| Not Helpful | Enough contributors rated it not helpful. It is not shown on X. |
The ranking page also uses the abbreviations CRH and CRNH for the Currently Rated Helpful and Currently Rated Not Helpful states, so you may see those in the code and the data.
On losing visibility, the notes-on-X page is direct: statuses are updated as new ratings arrive, “so notes may show on X and then disappear,” and statuses are locked after two weeks. The ranking page adds that the status is stored once a note is two weeks old for notes decided by certain models, and that some models are exempt.
The ranking is published as open-source code at github.com/twitter/communitynotes. The latest commit on the main branch, as reported by GitHub on 2026-10-16, was dated 2026-10-01. The prose page and the code can differ in detail, and the code is the more current of the two.
What the author of the post can and cannot do
Options the documentation actually describes:
- Request additional review. The FAQ tells post authors with a note on their post that they “can request additional review or report notes.” Per the additional review page, this applies to notes already rated Helpful and being shown. The note goes back to contributors on the Community Notes site, and the page says there is “no guarantee that more contributors will review it or that the note’s rating will change.” It also says X does not determine the outcome, and that if new ratings move the note out of Helpful it stops showing on the post.
- Report a note. The FAQ and signing-up pages say a note believed to break X’s Rules can be reported from the note’s menu. Notes are subject to those Rules.
- Rate, propose, or respond. X’s pages read here do not say whether the author can rate notes on their own post, and do not describe a way to prevent a note being proposed. The FAQ says blocking or muting does not affect the visibility of notes, so that route is documented as closed.
- Delete the post. I did not find a statement on what deleting a post does to its notes. Not verified.
One inference, labelled as such and not tested advice: since the review route sends the note to contributors who judge sources and wording, a manager is better placed to read the note’s cited sources before deciding whether a review request has any grounds. For wider recovery questions, see what platforms officially recommend when a post goes wrong.
What X does not publish about reach
I searched the pages read here for any statement about a note changing a post’s reach, recommendations or ad eligibility. The closest is in the FAQ: notes rated helpful by enough contributors “will appear directly on posts. Beyond that, notes do not affect display of posts or enforcement of X’s Rules.” That sentence concerns display and rule enforcement. It does not mention distribution, recommendations or advertising, and I am not reading a reach claim into it. Otherwise, X’s documentation does not publish a distribution effect, and I cite no number about reach.
The recommendation code is a separate question. It is covered in what X’s open-sourced recommendation code does and does not show, and nothing about note ranking should be inferred from it.
Documented, inferred, claimed: sorting the statements
No commentator claim qualified here as named, dated and fetched, so this sort covers only what the documentation says and what it leaves out.
| Statement | Where it stands | Source |
|---|---|---|
| Contributors write and rate notes. | The documentation says contributors submit and rate notes. It does not describe X staff writing them. | ranking-notes.md |
| A note needs agreement across viewpoints. | The documentation says so. It does not give a simple count that a reader could check by hand. | notes-on-twitter.md |
| A shown note can disappear. | The documentation says so, and says statuses lock after two weeks. | notes-on-twitter.md |
| The author can ask for another look. | The documentation says so for notes rated Helpful. It does not promise a result. | additional-review.md |
| A note cuts the post’s reach. | The documentation does not say this. Inference from it would be unsupported. | faq.md |
| Deleting the post removes the note. | Not found in the pages read. Not verified. | faq.md |
What to check before you rely on this
This post reflects the pages as read on 2026-10-16. Before acting on it, read the current text of the ranking page and the signing-up page in a browser, which renders them, and look at the repository for recent commits. For recovery questions beyond the mechanism, the crisis post linked above is the place to start.
FAQ
Can a brand account stop a Community Note from being proposed on its post?
X’s pages read here do not describe such a control. The FAQ says any post can receive a note and that blocking or muting does not affect note visibility. What is documented is the option to request additional review of a note rated Helpful, or to report one that breaks X’s Rules.
Does X say a Community Note reduces reach?
Not in the pages read here. The FAQ says that beyond appearing on posts, notes do not affect display of posts or enforcement of X’s Rules, and X does not publish a distribution effect. Claims of a specific reach effect are not backed by this documentation.
Sources
- Note ranking algorithm (Markdown source), fetched 2026-10-16
- Signing up (Markdown source), fetched 2026-10-16
- Notes shown on X (Markdown source), fetched 2026-10-16
- Additional review (Markdown source), fetched 2026-10-16
- FAQs (Markdown source), fetched 2026-10-16
- github.com/twitter/communitynotes, fetched 2026-10-16
