Getting feedback that improves the draft.

Send a draft with "let me know what you think" and you will get back typo fixes and a compliment. Not because your reader is lazy / because you have not told them what kind of help you need, and line edits are the safest default.

Feedback quality is mostly set by the brief, before anyone reads anything.

Send three things with the draft

1. Your claim, in one sentence. The reader can then tell you whether the draft actually makes it. Half the time it does not, and that is the most valuable thing they can say.
2. The stage. "Structure draft, arguments not polished" stops people editing prose you are about to delete.
3. Two or three specific questions. This is the part people skip and it is the part that works.

Questions that get real answers

• Where did you stop believing me?
• Which paragraph could I cut without losing anything?
• Does the evidence in section three support the claim in the abstract, or am I overreaching?
• What will the obvious objection be?
• Read only the first and last paragraph / do they match?

"Where did you stop believing me" is the highest-yield question in that list. Readers can always locate the moment they became sceptical, even when they cannot say why, and the moment is nearly always the real problem.

Ask for the disagreement explicitly

People soften critique to protect the relationship. You can release them from that with one line: I would rather find the hole now than in the viva / be blunt. Say it and mean it, because if you argue with the first hard comment you will not receive a second one.

When the feedback arrives

Read it once, then leave it for a day. The instinct to defend is strongest in the first hour and it is not information.

Then sort every comment into three piles:

• Factual corrections. Just fix them. No deliberation.
• Comprehension failures. The reader misread something. This is almost always your fault, not theirs / one confused reader means the sentence is ambiguous. Rewrite rather than explain.
• Disagreements about the argument. These need a decision. You are allowed to keep your position, but write down why, in a sentence. If you cannot, that is your answer.

Reading between the comments

The signal is often in the pattern, not the individual notes. Three comments asking what you mean by one term means the term needs defining once, early. A cluster of edits in one section usually means that section is where your thinking is least resolved / the prose gets tangled where the argument does.

And if a reader spends their whole review on grammar in a draft you asked structural questions about, they probably could not follow the structure well enough to comment on it. That is data too.

Two readers, not five

One person who knows the field and one who does not is the ideal pair. The specialist catches the flawed inference; the outsider catches the paragraph that only makes sense if you already know the literature. More than three readers produces contradictory advice and a draft edited into blandness by committee.

---

Close the loop. Tell your reader which changes you made and which you did not, in three lines. It takes a minute, it is the reason people read your next draft, and explaining your refusals is often when you finally work out what your own argument is.