An editor is not a proofreader with opinions. Working well with one starts with knowing which of four jobs you are buying, then briefing for it, then reading the response as information rather than judgement. The relationship works when both people can tell the difference between a diagnosis and a prescription.
- Match the editor to the pass the draft is in, not to the budget you have.
- Brief in two paragraphs: reader, purpose, fixed constraints, and known weaknesses.
- Wait two days before responding to an edit letter you dislike.
- Separate the editor's diagnosis from their proposed remedy; trust the first more.
- Decline in writing, with a reason and a condition that would change your mind.
Four editors, four different jobs
A developmental editor works on what the book is: the argument, the structure, what is missing, what should go. A line editor works on how it reads: rhythm, clarity, the paragraph that takes four sentences to begin. Both change the text substantially, and neither is inspecting your commas.
A copy editor enforces correctness and consistency, queries facts, and applies a house style. A proofreader reads the set pages and catches whatever survived. The passes run in that order for a reason. Buying the last one first gets you an immaculate version of the wrong book.
Hire for the pass the draft is actually in
Most unhappy editing experiences begin as casting errors. A writer who suspects the structure is broken hires a copy editor instead, because copy editing is easier to specify and cheaper to buy. The copy edit returns spotless, and the structural fault is still there, now wearing polish.
Diagnose before you commission. If you cannot say in one sentence what the draft's biggest problem is, that uncertainty is the brief, and a developmental read is what answers it. Ask for a sample edit on ten pages before anyone signs for four hundred.
An editor who watches you concede fifty notes will listen hard to the one you defend.
Brief before the pages change hands
An editor without a brief edits toward their own taste, because taste is the only standard you left them. Give them a better one in two paragraphs: who the reader is, what the piece has to accomplish, and which constraints are genuinely fixed.
Include what you already know is weak. Naming your own bad chapter saves time and signals that you can hear a real note. Say which passes you have run and what you left rough on purpose, so nobody spends a week on a section you plan to cut.
How to read an edit letter
Read it once, then close the file. Whatever arrives in the first hour is a response to being seen, not an assessment of the argument. The letter will be just as accurate in two days, and by then you can read the sentences instead of the temperature.
On the second read, split diagnosis from prescription. That a chapter stalls is a diagnosis. That you should cut its history to two pages is one available remedy. Editors are more reliable about the first, because they felt the stall directly and inferred the cure.
Sort every note into accept, adapt, or decline
Do this before you change a word. Accept means the editor is right and you will do what they proposed. Adapt means the diagnosis holds but you have a better fix. Decline means the passage stays and you owe a reason in writing.
Most notes should land in the first two piles. When almost everything lands in decline, either the brief failed or the wrong editor took the job. Both are conversations worth having immediately, not at the third round when the schedule has already slipped.
Disagree with reasons, not feelings
A productive disagreement hands the editor something to work with. Say what the passage does for the reader, offer a smaller version of it, and name the condition under which you would give it up. That last part turns an objection into a shared test.
Do not litigate everything in the margins. Choose the two or three notes that genuinely matter, raise them together, and take the rest gracefully. Spending your credibility on a subheading leaves nothing for the chapter you actually need to protect.
What you owe in return
Send a draft you have stopped rewriting. Answer queries in the file rather than in a separate thread nobody can search later. Hold the date you agreed, because an editor books your weeks months ahead and a late manuscript takes somebody else's slot.
Then report back. A short note listing what you accepted, what you solved differently, and what you kept tells the editor how to pitch the next round. Silence leaves them guessing, and the usual guess is that the letter was ignored.
An edit letter note, and a reply that moved the book
From a developmental letter on a 40,000-word business book, on chapter four: "Chapter four stalls. You give eleven pages to the history of the procurement team before the reader learns why any of it matters, and the chapter's real argument, that the approvals process was designed for a company a third of your current size, does not surface until page twelve. I would open with the argument and cut the history to two pages. I would also lose the airline analogy in the middle section. It is charming, and it explains something the reader accepted four pages earlier. Right now the chapter asks for patience it never repays."
The writer's reply, sent three days later: "Agreed on the opening. The argument moves to page one, and the history drops to two pages, kept only where it explains why the thresholds were set as they were. You are right that it stalls. "I want to keep the airline analogy, and here is the reasoning. Readers accept the point in the abstract by page eight, but the interviews suggest they do not believe it applies to their own team until they see it working outside procurement. That analogy is the only place the book leaves this industry. I will cut it from 600 words to 250 and move it to the end of the chapter, so it lands as a test rather than an explanation. If it still reads as decoration, it goes."
The reply concedes the larger note without hedging, then defends the smaller one with a reader-facing reason rather than an attachment. It offers a reduced version, moves it somewhere it does different work, and names the condition for dropping it.
Common mistakes
| The mistake | Why it costs you, and what to do instead |
|---|---|
| Buying a copy edit to avoid a developmental one | Line-level polish on an unresolved structure is the most expensive kind of wasted work, and it makes the real problem harder to see. Ask for a developmental read first, even a short one, whenever you cannot name the draft's main fault. |
| Replying to the letter the day it lands | The first-day reply argues with the tone rather than the content, and it sets an adversarial pattern for every round after. Read it, close it, wait two days. Nothing in the letter expires, and your judgement returns quickly. |
| Defending the prose instead of the purpose | Saying you like a sentence gives the editor nothing to weigh. Say what the passage does for the reader that nothing else does. If you cannot answer that, the note was probably right and the sentence was probably yours alone. |
| Accepting every note to stay agreeable | Total agreement is not cooperation; it hands the editor authorship they did not ask for and rarely want. An editor holds the standard, and you hold the intent. Declining a few notes with reasons is part of doing your half. |
Your checklist
- Name the draft's largest problem in one sentence before you commission anyone.
- Request a paid sample edit on ten representative pages first.
- Send a two-paragraph brief covering reader, purpose, constraints, and known weak spots.
- Read the edit letter once, wait two days, then sort notes into accept, adapt, decline.
- Answer every declined note in writing with a reason and a condition.
- Send a short summary of what changed once the round is finished.
Key terms
- Edit letter
- A standalone memo delivered with or before a marked manuscript, setting out the editor's overall diagnosis, the main problems in order of severity, and the recommended remedies.
- Sample edit
- A short paid edit of a representative extract, commissioned before the full engagement, used to test whether an editor works at the level and in the register you need.
- Stet
- A mark meaning let it stand: an instruction to ignore a proposed change and restore the original wording. The traditional way a writer declines a note on the page.
Getting a draft read by someone else?
Frequently asked questions
How do I tell which kind of editor I need?
+
Describe the symptom, not the cure. Readers losing the thread points to developmental work. Readers agreeing but finding it heavy going points to a line edit. Readers happy but tripping over errors points to copy editing. Two of those can be one person, in sequence.
What should a sample edit tell me?
+
Whether they read at the level you asked for, whether their notes name consequences rather than preferences, and whether they can hear your voice. A sample that rewrites your sentences into theirs is a genuine answer, just not the one you were hoping for.
How many rounds should a long project take?
+
One developmental round with a follow-up read, one line pass, one copy edit, one proofread. Extra developmental rounds usually mean the brief never settled. If you find yourself opening a fourth structural round, stop and rewrite the brief before anyone edits anything further.
What if I disagree with almost the whole letter?
+
Treat that as a signal about fit or briefing, not about the editor's competence. Ask for a call and put your intent on the table in plain terms. If the gap survives that conversation, part company early and pay for the work done.
Can Strut AI do an editor's job?
+
Strut AI can carry the mechanical load: consistency sweeps, missing support for a claim, an outline of what the draft currently does. It cannot hold a standard or take responsibility for what ships. An editor is a second professional judgement, and that is what you are paying for.
Who has final say on the text?
+
Under most arrangements the writer does, and the editor's job is to make sure a refusal is informed. Contracts vary, and legal or factual corrections are not negotiable. Settle this question in the first conversation rather than discovering it at proof stage.