The tenth piece on your subject repeats the second unless you plan against it

Each piece is fine and the set is not

Write about one area regularly and something happens that no individual piece reveals. The same three examples turn up everywhere. Two articles make the same argument from different starting points. A framing you found useful in March becomes the frame for everything after it, and the pieces begin to read as variations rather than as additions.

Nothing in the writing process surfaces this. Each piece is judged against its brief and against itself, both of which it passes. The reader who encounters four of them in a row is the one who notices, and by then it is published.

Drafting assistance sharpens the effect rather than causing it. Ask for a treatment of a topic in your area and you get the most standard account of it, which is exactly the account your earlier pieces have already been converging on.

The repetition is at the level of argument, not of wording

Checking for duplicate sentences catches almost none of this. Two pieces can share no phrasing at all and still make one point: that the obvious approach fails for a reason people underestimate, that the interesting decision happens earlier than anyone thinks, that the tool is fine and the process around it is not.

A subtler form is the section that quietly becomes an article. Something you covered in three paragraphs inside a longer piece reappears as a standalone treatment months later, and to a regular reader the second one adds a title and very little else.

So the unit to audit is the claim, not the text. What does this piece establish that no other piece of mine establishes? If the honest answer is that it establishes it more thoroughly, that may still be worth publishing, and it should be a decision rather than an accident.

Keep an index of arguments, not of titles

The practical instrument is a list, maintained as you go, with one line per published piece stating the claim it makes rather than its title. Titles disguise repetition; claims expose it, because two lines that say nearly the same thing sit next to each other on the page and cannot be missed.

Add a second column for the examples used. Examples are where writers repeat themselves first and notice last, since a good example feels earned each time you reach for it and its previous outings are invisible while you are drafting.

The list has to be consulted before writing rather than after, which is the part that requires discipline. Reading it at the outline stage costs two minutes and changes what you write; reading it before publishing only tells you the piece is a duplicate once the work is done.

Deliberate difference beats hoping for it

Once you can see the shape of what you have covered, the useful move is to choose a direction that is missing rather than a topic that is new. The set may be full of pieces about why something fails and empty of pieces about what to do when it fails anyway. It may be entirely about the beginning of a process, or entirely written for people who already agree.

Vary the axis explicitly: the stage of the work, the reader, the scale, whether the piece argues or explains, whether it treats the exception or the standard case. Each of those produces a genuinely different article about the same subject, because the substance changes rather than the vocabulary.

This is also where a subject that felt exhausted usually reopens. Most of the time the topic was not exhausted; one particular angle on it was, and that angle had become a habit.

A body of work is allowed to have a position

Consistency and repetition are not the same thing, and over-correcting produces its own problem, where a writer avoids their central claim because they have made it before. Returning to an argument is legitimate and often necessary. What matters is that each return adds a case, an objection, a complication or a change of mind.

Saying so explicitly helps the reader as well as you. A piece that acknowledges having argued this before and is now testing it against a harder example is doing something the first piece did not.

The failure to avoid is different: producing the same argument as though for the first time, repeatedly, with the examples reshuffled. That is what a body of work converges towards without intervention, and the intervention is nothing more than a list you actually read.