Scope note: the three passages below are the same kind of trade-off written in different settings, not a law of division. Who takes which part is arranged by the work, the experience with this kind of task and the time available; type does not enter the allocation.
An outsourced draft going back and forth: the second version is still wrong
(Fictional scene.) A brand manual's introduction goes to an outside writer. The first draft reads like a product description, the second turns lyrical, and the client says "neither is what I want". Five days to deadline, budget for one more round.
One approach brings all three into a short call: let the client say in their own words who should read this and what they should do afterwards, put the writer's concerns on the table, and settle three revisions — open with a scene, give one person's name in the second paragraph, end with a single call to action. The other approach puts the two drafts and the original brief side by side and finds what the client was pointing at every time they said "no": not the tone but the stance — the manual speaks for the company, and the client wants it to speak for the shop managers.
Each has something missing: talking face to face easily turns a disagreement into a compromise that satisfies nobody in the third draft; comparing text alone may stay too abstract for the writer to know how to change a "stance". A workable combination is to write one sentence — whose voice this text speaks in — confirm it with the client, then break it into three concrete revisions with examples. Either person can do that step, depending on who knows the client and this kind of text better.
What counts is whether the third draft settles whose voice it speaks in, not whether the tone matches. The conditions have to be written too: if the writer has a record of late delivery, acceptance terms belong in a supplementary agreement; if the writer is new, revisions should go down to sentence-level examples. Neither response is about type; both are about delivery risk and the other person's experience.
A way to verify: attach an acceptance standard and one example to every must-change comment, then check next round what was actually solved and what remains. Tiers still work — must change, could improve, optional — but the tiers only order this round's workload, and do not show who cares more about the text or about delivery.
A writer who keeps delivering late: replace or keep
(Fictional scene.) A writer the team has worked with for three months delivered three days late on each of the last two assignments, with no drop in quality. The project needs to scale up next month.
What can be compared are the shared conditions, not which type anyone resembles: whether the two delays had the same cause, whether they were flagged in advance, whether other collaborators were affected, what searching for and onboarding a replacement would cost, and how much error the scale-up period can absorb. One handling talks terms first — find out the cause, agree a schedule the writer can keep, and write down what happens if it recurs; the other costs out replacement first — assess how available candidates are and how long a handover takes, then decide whether to continue. Both types can choose either, and you can discuss revised terms while preparing a backup.
The disagreement is usually about the time scale: those who can wait will spend one more chance on quality, and those who are up against a launch would rather replace now. Writing those conditions down is more useful than arguing about who is more considerate.
A younger cousin choosing a university major
(Fictional scene.) A younger cousin has the grades for a strong university and has applied for palaeontology; her family is worried about the choice. One entry point is to talk to her alone first: why this one, has she met anyone who studied it, and then how to raise it at home — what to say first, and who says it. The other entry point is to read the programme and where its graduates go, and work out whether she is drawn to studying fossils, working outdoors or something else about the field.
Each carries a risk: the one who plans the conversation easily turns "persuade the family" into the goal, and she ends up pushed into changing her application; the one who checks the field easily treats a genuine interest as a conclusion and ignores where she will be in four years. Which appears more often depends on who is present, who understands the field, and whether the family has already argued about it; the same person may take a different entry point in a different family.
A way to verify: let her write three lines herself — why I chose it, which course in it will be hardest, and what I do if I cannot transfer. She writes them, and only then does the conversation about changing happen. The point is that she checks it herself, rather than someone else judging her motive for her.