How to Get Through the Editing Phase Without Abandoning Your Book

The editing phase stalls because it has no natural finish line: drafting ends when you reach the last page, revision ends when you decide it does. The way through is to give it one. Work in passes, meaning a single read-through with one defined job, size the work in hours rather than in feelings, and decide in advance what “finished” will look like.

Why Does the Editing Phase Stall So Many Books?

Drafting has a built-in endpoint and revision does not. Finishing a draft announces itself; finishing a revision does not, because there is always something else a manuscript could be. Revision will absorb as much time as you are willing to give it, and unlike the draft it does not tell you when to stop. That structural difference does most of the damage, and it is the part you can actually fix.

A second problem follows from the first. Without a written plan for what each pass is for, revision collapses into one undifferentiated activity. You open chapter three to fix a scene, notice a clumsy sentence, rewrite the sentence, notice the chapter before it now reads oddly, and an afternoon goes by in which the book changed but nothing was finished. Do that often enough and the same chapters get worked repeatedly while whole sections at the back go untouched.

The third is simple over-familiarity. Reading the same sentences for the fifth time flattens them; you stop hearing the rhythm and start seeing only the machinery. What that calls for is practical rather than emotional: change how you are reading, and stop before you are worn out. The section on breaking up a revision covers both.

How Many Rounds of Editing Does a Self-Published Book Need?

Probably fewer rounds than you have been told a book needs. The widely repeated advice that every book needs a developmental edit, then a line edit, then a copy edit, then a proofread describes the workflow of a large trade publisher, and it assumes a publisher’s budget. A self-publishing author is paying for each of those personally, and the useful question is not how many stages a book “really needs” but which ones to buy with what you have.

It helps to know what the named stages actually are, because the labels are used loosely. The Editorial Freelancers Association’s service definitions break editing into four types. A marked-up manuscript, in the table below, means your own file returned with the editor’s changes and queries visible in it; a style sheet is the record an editor keeps of the choices made for your book, so that spellings, names and formats stay consistent across it.

Two things in that table are worth pausing on. The first is that the boundaries are genuinely soft, and two of the field’s professional bodies draw them differently. The EFA places a written assessment inside developmental editing, and says line editing “may be performed as a separate service, in conjunction with developmental editing after big-picture issues have been addressed, or in conjunction with copyediting”. The Chartered Institute of Editing and Proofreading notes that some publishers and editors combine copyediting and proofreading into a single service called proof-editing, in which “the proofreader is not working on the final draft”; that is a heavier pass done earlier, and it is a different arrangement rather than an exception to the ordering below. When two editors describe the same service differently, that is the field being imprecise rather than one of them being wrong. Ask what will come back, and at which stage, rather than trusting the label.

The second is the position of a true proofread. The CIEP states it plainly: “proofreading happens after a document has been designed.” A proofreader is checking the laid-out pages, which is where a large share of what a proofread exists to catch actually appears: bad line breaks, a running head on the wrong page, a contents entry pointing at the wrong page number, a caption that has drifted from its image. Running that pass on a Word file instead leaves those faults to be met in the printed proof. For a book going to print, then, buy the proofread after formatting rather than before it. An ebook-only release has no fixed pages to check, so there the final read sits wherever your production makes sense.

For a first book, then, a workable shape is: self-edit as far as you can take it, get reader feedback, buy one professional edit before formatting (most often a copy edit, or a lighter pass closer to a proofread), format the book, and have the laid-out pages proofread. That is two paid stages rather than four, and it is enough to produce a book that does not read as unedited. If the budget stretches to only one, the copy edit is the one that changes the most pages.

How Long Should the Editing Phase Take?

Editing a full-length book takes longer than most first-time authors expect, and the gap between the expectation and the reality is itself a reason people give up on one. The useful way to set expectations is in working hours rather than calendar weeks, because hours can be estimated from published figures and calendar weeks cannot. The figures below describe professionals working at professional speed; your own passes will be slower, because you are deciding as well as reading.

The EFA’s 2026 Rate Chart reports how many pages an hour its members get through, from a survey of more than 1,100 respondents run from November 2025 to mid-January 2026 about work done during the 2025 calendar year. A page there is 250 words. For fiction, members reported covering 6 to 10 pages an hour for copyediting and 10 to 15 pages an hour for proofreading. Run those against a 70,000-word novel, which is 280 pages at that measure:

• Copy edit: 280 pages at 6 to 10 pages an hour is roughly 28 to 47 hours of concentrated work.
• Proofread: 280 pages at 10 to 15 pages an hour is roughly 19 to 28 hours.

The CIEP arrives at a similar place from a different direction, giving “between 2,000 and 3,000 words an hour” as a rule of thumb for a professional proofreader on straightforward text, and noting that complex text takes longer. On the same novel that is about 23 to 35 hours. Two bodies, two methods, and answers that overlap between 23 and 28 hours: call it most of a working week of full-time attention for a proofread, and up to twice that for a copy edit.

The reason to look at the numbers at all is that “six weeks” means nothing you can act on, while “about thirty hours, so an hour a day on weekdays for six weeks, or two hours a day for three” is something you can put in a calendar and finish. Both bodies caution against reading the figures too tightly: the EFA states that it neither sets rates nor advises members on them, and that the number of responses varied widely between questions. Treat all of this as orders of magnitude.

How Do You Break a Large Revision into Manageable Steps?

Split the work into passes, and give each pass exactly one job: look for one kind of problem and deliberately ignore the rest. The reason this works is that the editorial trade is organised the same way, with structure, sentences and typography bought as separate services rather than done all at once. Trying to fix everything on one read is what turns an afternoon of work into no finished chapters.

A workable sequence for a novel:

1. Structural pass. Read the whole manuscript without editing. Note pacing, plot holes and weak scenes. Mark chapters to rewrite, cut or reorder. Fix no prose.
2. Scene-level rewrite. Work through the structural notes, rewriting and rearranging. Focus on story logic and emotional beats. Still ignore sentence polish.
3. Line pass. Now the prose. Tighten sentences, cut redundancy, sharpen dialogue, fix awkward phrasing. Work chapter by chapter.
4. Consistency check. Read specifically for facts: timeline, character descriptions, setting details, proper nouns. A simple spreadsheet of names, dates and physical details pays for itself here, and it is close to the style sheet a copyeditor would build for the same book.
5. Final read for typos and punctuation. Note that this is your own last check of the words, not the proofread described above: that one comes later, once the book has been laid out.

Each pass needs a start, an end and a scope you have written down. When a pass finishes, that kind of problem is closed. If you find yourself reopening structure after the line pass, treat it as a separate bounded job on the specific chapters concerned rather than as a reason to start the whole cycle again.

Three habits keep the passes moving when the manuscript has gone stale on you:

• Change the medium. Print it if you have been on screen; use text-to-speech if you have been reading silently. Shifting how the text reaches you surfaces errors your eyes have learned to skip over.
• Set session limits. Ninety minutes, or fifteen pages, then stop wherever you are. This is also what turns the hour estimates above into a plan: a fixed session is countable, so you can see the end from the start, and an open-ended one is neither.
• Put it down between major passes. Two to four weeks away from the manuscript restores the distance that repeated reading takes away. Build it into the schedule rather than counting it as time lost.

How Do You Handle Conflicting Feedback from Beta Readers and Editors?

Sort the comments into groups before you act on any of them. Conflicting feedback is the normal result of showing a book to several people, whether those are beta readers (readers who read the finished draft and report back as an ordinary reader would) or a paid editor, and the mistake that causes paralysis is treating every comment as equally binding.

Three groups cover most of it. A problem two or more readers hit independently is worth taking seriously, whatever each of them proposed doing about it. A suggestion that reflects one reader’s taste is information, not an instruction. A straight contradiction, where one reader wants a scene cut and another calls it their favourite, is decided by what you are trying to do with the book, not by counting votes.

Writing in Writer’s Digest in 2014, its then content editor James Duncan put the test well in a piece on soliciting feedback: “if everyone seems to get lost at the same point in the story … then you have a problem on your hands. You’re the final judge, but watch for patterns in the feedback.” On the contradictions he is equally useful: a character one reader hates and another loves “might be a flawed character who needs work, or just a divisive character”, and divisive characters “can be some of the best”.

The same article names the step that prevents much of the confusion in the first place, which is briefing readers before they start. Send them written questions, some about the book as a whole and some about particular chapters, or what comes back may be “That was great, thanks!” and nothing you can work with.

With a professional editor there is a distinction worth holding on to: the problem an editor identifies and the fix they propose are two separate things. If an editor says a chapter drags, take the observation seriously and treat the specific cut they suggest as one solution among several. Solving the stated problem your own way is a normal part of the process rather than a rejection of the edit. It also helps to read an editorial letter (the written summary of a developmental edit, sometimes called a revision letter) once, put it away, and come back to it before replying. The second reading is a noticeably different document from the first.

When Should You Stop Editing and Call the Book Finished?

Stop when further passes are producing changes rather than improvements. If you are rewriting sentences to be different instead of better, or switching between two versions of a paragraph and preferring whichever you read most recently, you have reached the point where more work stops paying.

Set the bar deliberately, and set it before you are exhausted. The CIEP puts the practical version of this to authors directly: “If you have a limited budget, be upfront about this and agree on what’s ‘good enough’ in terms of your text.” It also says the thing that settles the perfectionist version of the question, which is that “no proofread is ever 100% perfect”. Published books carry errors. The standard to aim for is a book whose errors do not interrupt a reader, not a book with none.

Practical signals that the phase is over:

• You have completed a self-edit pass and at least one professional round, and the proofread is on the formatted pages rather than the manuscript.
• You can open a chapter at random and read it without finding something that makes you wince.
• The last round of notes was mostly minor and optional rather than structural.
• Names, dates and facts have been checked across the whole manuscript.
• The opening, the ending and the book description (the sales copy that will appear on the retailer’s page) all describe the story the final draft actually tells, rather than the one you outlined.

One further test: if you handed the manuscript to a reader tomorrow with no more changes, would you be content with what they received? Not perfect. Content. If the answer is “almost, but chapter twelve still needs work”, then chapter twelve is a bounded, finite job. Finish that, and publish.

The kind of problem you are still finding tells you which way to go. The distinction between developmental editing, copy editing and proofreading is useful here: if you are still turning up plot holes in a novel, or arguments that do not follow in a work of non-fiction, you are in structural territory and further polishing will not help. If your changes are at the level of word choice and commas, you are near the end.

Frequently Asked Questions

How long should the editing phase take for a novel?

Estimate it in hours rather than weeks. Using the EFA’s 2026 Rate Chart figures for fiction, a 70,000-word novel is about 280 pages of 250 words, which is roughly 28 to 47 hours for a copy edit and 19 to 28 hours for a proofread at professional speeds. The CIEP’s rule of thumb of 2,000 to 3,000 words an hour for proofreading gives a similar answer. Your own passes will be slower, because you are deciding as well as reading.

Do I really need three or four rounds of professional editing?

No. The developmental, line, copy and proof sequence describes how large trade publishers work, on a publisher’s budget. A workable shape for a first self-published book is one paid edit before formatting, most often a copy edit or a lighter pass, and then a proofread of the laid-out pages: two paid stages rather than four. Choosing well between the stages matters more than collecting all of them.

Should I proofread before or after my book is formatted?

After, for a book going to print. The CIEP states that proofreading happens once a document has been designed. A proofread covers typographical errors and layout errors together, and the layout half of that list only exists once the pages are laid out: bad line breaks, misplaced running heads, contents entries pointing at the wrong page, captions that have drifted from their images. None of those are visible in a Word file. An ebook-only release has no fixed pages, so the constraint does not apply in the same way.

Should you self-edit before hiring a professional editor?

Yes. A thorough self-edit means the editor spends their time on what you cannot see yourself, such as inconsistencies, blind spots and genre conventions, rather than on problems you could have cleared. A cleaner manuscript is also quicker to work through, which on an hourly arrangement means a smaller bill.

How do you process an editorial letter without getting overwhelmed?

Read it once, put it aside, and come back to it before you respond. On the second reading, sort each note into structural, line-level, or optional. Work the structural notes first in a pass of their own, then the line-level notes in a second pass, and decide on the optional ones last. Taken in that order, one intimidating document becomes three ordinary jobs.

My manuscript feels worse every time I read it. What should I do?

Change how you are reading it and how long for. Print it if you have been working on screen, or listen to it through text-to-speech; fix your sessions at a set length or page count and stop at the end of them; and put the manuscript away for two to four weeks between major passes. Over-familiarity with your own sentences is an effect of repetition, and distance is what reverses it.