Writing Craft for Authors: An Editor's Field Guide to Revision

Updated May 2026.

Most writing craft guides are written for students or for beginners learning to draft a first novel. This guide is for authors who already have a draft and want to make it stronger. It's written from the perspective of professional editors who have read hundreds of manuscripts and noticed the consistent patterns: the moves that work, the moves that backfire, the revisions that turn a competent draft into a finished book.

Craft, in this guide, means the choices an author makes at every level of the manuscript: figurative language, sentence rhythm, scene construction, dialogue, point of view, pacing, voice, and the dozens of smaller decisions that accumulate into the felt experience of reading the book. Each of these is improvable. None is mysterious. What separates a strong manuscript from a weak one is rarely talent in some abstract sense; it's the specific craft choices the author made and the revisions they did or didn't do to refine them.

This article maps the craft territory in four pillars, with deeper guides linked from each section. Read it as the overview. Then deep-dive into the specific pillars and spoke articles where your manuscript needs the most attention.

In This Guide

• What Writing Craft Actually Is
• Pillar 1: Figurative Language
• Pillar 2: Prose Mechanics
• Pillar 3: Scene Construction and Dialogue
• Pillar 4: Revision Strategy
• How to Diagnose What Your Manuscript Actually Needs
• What a Professional Editor Sees That Self-Editing Doesn't
• What to Do Next

What Writing Craft Actually Is

Craft is the difference between writing that does what it intends and writing that doesn't quite. It's not style, which is the surface texture of an individual author's voice. It's not technique in the narrow sense of grammar rules. It's the underlying set of skills that lets an author make deliberate choices about what to do at every level of the manuscript and execute those choices well.

A manuscript with strong craft accomplishes specific things consistently. The opening pulls the reader in. The voice stays consistent from chapter one to chapter forty. The pacing pulls forward through the slow middle. The dialogue sounds like the characters speaking, not the author writing. The descriptions land where they're supposed to land. The themes deepen across the book without becoming heavy-handed. None of this is accidental. All of it is craft.

A manuscript with weak craft typically has identifiable problems. The opening doesn't establish character or situation quickly enough. The voice wavers between chapters. The middle sags. The dialogue feels stilted, or every character sounds the same. Descriptions either over-explain or remain generic. The themes either don't develop or are stated rather than dramatized. These problems are not signs that the author lacks ability. They're signs that the manuscript hasn't been through enough revision focused on the right things.

Craft is improvable

The most useful thing to understand about craft is that it improves with deliberate practice. The most experienced novelists working today have nearly all said versions of the same thing: writing got easier and the prose got better because they kept writing and revising. Craft is not a fixed quality the author has or lacks. It's a skill that develops with attention.

What deliberate practice in writing craft means, specifically, is reading widely in the form, noticing what other writers do, trying things in your own work, and revising based on what does and doesn't work. Most authors who don't improve are not lacking talent. They're either not revising enough or revising without focus. The pillars below are how to focus.

Pillar 1: Figurative Language

Figurative language is the set of devices authors use to make abstract things concrete, ordinary things vivid, and specific things resonate beyond themselves. Metaphor, simile, imagery, personification, hyperbole, irony, symbolism, motif. The toolkit is large. The misuse of it is one of the most consistent problems editors see in first-draft manuscripts.

The most common figurative language problem

Most first-draft manuscripts have too much figurative language, not too little. Authors who care about their prose often over-decorate. Similes appear in every paragraph. Metaphors stack on metaphors. Personification gets reached for whenever a description is needed. The result is prose that calls attention to its own cleverness and slows the reader down without earning the slowdown.

The editor's perspective on figurative language is that each device should earn its place. A simile that doesn't tell the reader something specific about the comparison is doing decorative work, not craft work. A metaphor that mixes images (something cold blazing through a frozen room) is fighting itself. Hyperbole used everywhere stops being emphatic and starts being meaningless. The strongest manuscripts use figurative language deliberately, in specific places where the device is doing something the literal description couldn't do.

What earns its place

A simile that makes the reader see something they couldn't see otherwise. A metaphor that compresses a paragraph of explanation into one image. A piece of imagery that means more on the second read than the first. Foreshadowing that the reader doesn't recognize as foreshadowing until later. Symbolism that emerges from the work rather than being imposed on it. These are the moves figurative language can make that nothing else can.

For the full treatment of how each device works, when it helps, and when it backfires, see our pillar guide on figurative language. Sub-articles deep-dive into specific devices: hyperbole, metaphor vs simile, personification, alliteration and assonance, foreshadowing, irony, symbolism, imagery, motif, anaphora and repetition, onomatopoeia, and oxymoron. Each is treated from the editor's perspective on what actually shows up in manuscripts and what to do about it.

Pillar 2: Prose Mechanics

Prose mechanics is the sentence-level and paragraph-level craft that makes prose feel alive or flat. It's not grammar in the narrow sense, though grammar matters. It's the broader set of choices about how sentences move, how paragraphs structure information, which words carry meaning and which dilute it, and how the prose reads aloud.

Sentence variety

The most common prose mechanics problem in first-draft manuscripts is sentence monotony. Most sentences are roughly the same length, follow roughly the same structure (subject, verb, object), and produce a numbing effect that the reader registers as flat prose without being able to name the cause.

Strong prose mixes sentence lengths and structures deliberately. Short sentences for emphasis. Long sentences for unfolding thought. Sentence fragments where appropriate. Compound sentences that build. The variety creates rhythm, and rhythm is the difference between prose that pulls the reader forward and prose that they have to push through.

Voice and consistency

Voice is the most discussed and least definable element of prose. From an editor's perspective, voice is the consistent set of choices an author makes about word selection, sentence rhythm, and narrative distance that distinguishes one author's prose from another's. Voice is also the element that most often wavers in a long manuscript, because chapters are typically drafted across months or years and the author's voice changes with experience and circumstance.

Consistency of voice across a manuscript is one of the things editors flag immediately. A chapter that suddenly sounds like a different writer, or a passage that shifts register without reason, breaks the reader's experience. Voice consistency is often what a manuscript gains in the second or third revision pass.

The words that weaken prose

Specific filler words appear in nearly every first draft. Just, really, very, quite, somewhat, basically, actually, literally. These words rarely add meaning. They dilute. Removing them tightens prose immediately. Editors track these patterns because once you see them, you see them everywhere.

Adverbs are the other consistent prose-mechanics problem. Stephen King's famous advice that the road to hell is paved with adverbs is not entirely right, but it's mostly right. Most adverbs in first drafts are signs that the verb isn't doing enough work or that the dialogue tag is over-explaining. Strong prose uses verbs and nouns rather than verb-adverb pairs.

For the full treatment of sentence variety, voice, word choice, active vs passive construction, paragraph length, and the consistent prose mechanics problems editors catch, see our pillar guide on prose mechanics. Sub-articles deep-dive into each topic, including show vs tell (the most misunderstood advice in writing), concrete vs abstract diction, and sentence rhythm.

Pillar 3: Scene Construction and Dialogue

Scene construction and dialogue are the two elements editors evaluate most actively when reading the first 50 pages of a manuscript. They're also the two elements where the gap between first-draft writing and published-quality writing is widest, and the gap that revision most reliably closes.

What a scene is supposed to do

Every scene in a novel has a job. It develops character, advances plot, deepens setting, establishes information that pays off later, builds tension, or accomplishes some combination of these. Scenes that don't have a clear job are scenes that should probably be cut or compressed.

First-draft manuscripts typically include too many scenes, and many of those scenes are doing too little. The author wrote them because something needed to happen between event A and event B, but the scene itself doesn't earn its place. Editors notice these scenes immediately. Most novels become tighter and faster in revision specifically because scenes are cut, compressed, or merged.

The difference between scene and summary

Some moments in a novel deserve to be dramatized as full scenes. Others should be compressed into a sentence or paragraph of summary. First-draft authors often dramatize everything, which produces a manuscript where every moment receives the same emphasis and the reader can't tell what matters. Strong revision involves identifying which moments deserve scenes and which deserve summary, then compressing accordingly.

Dialogue: the consistently weakest element

Dialogue is the element of fiction where editors most consistently see room for improvement. First-draft dialogue typically has several recognizable problems: characters explaining things to each other that they already know (for the reader's benefit), characters all sounding alike, dialogue that goes on too long without accomplishing anything, and dialogue tags that over-explain what the dialogue already shows.

Strong dialogue has specific qualities. It sounds like the way characters would actually speak, not the way they'd write. It accomplishes more than one thing at a time (revealing character while advancing plot, deepening tension while delivering information). It varies in rhythm and length. Each character has a recognizable voice. The dialogue tags are mostly invisible (said, asked), with action beats doing the work that overwritten tags try to do.

Point of view and consistency

Point of view (POV) is the single most consequential structural decision in a novel after the choice of story itself. First person, third limited, third omniscient. Each has its conventions, and each has its consistency requirements. Editors notice POV violations immediately: a third-limited narrator suddenly knowing what a non-POV character is thinking, a first-person narrator describing their own face in a mirror in ways no person does naturally, an omniscient narrator inconsistently dipping into and out of characters' minds.

For the full treatment of scene construction, scene vs summary, dialogue, dialogue tags, point of view, voice, and interior monologue, see our pillar guide on scene construction and dialogue. Sub-articles deep-dive into each topic.

Pillar 4: Revision Strategy

Revision is where books are made. First drafts are raw material. Second drafts begin to shape that material into a book. Third drafts and beyond are where the craft work happens that separates published novels from unpublished ones. Most successful novelists revise more than they draft.

The problem with revising everything at once

The most common revision mistake first-time authors make is trying to fix everything in one pass. They open the manuscript and start rewriting sentences while also noticing scene structure problems while also wondering whether the pacing is working while also catching typos. The result is a revision that addresses nothing thoroughly. Each layer of concern interferes with the others, and the author finishes the pass exhausted without having actually fixed anything.

Strong revision is layered. One pass for structural problems (scenes that need cutting, chapters that need reordering, character arcs that need work). A separate pass for prose-level problems (sentence variety, voice consistency, filler words). A separate pass for dialogue. A separate pass for opening pages, which need disproportionate attention because they're what agents and readers see first. A final pass for proofreading after the structural work is complete.

When revision is productive vs when it's tinkering

Revision has diminishing returns. The first revision pass typically produces dramatic improvements. The second pass produces meaningful improvements. The third pass produces refinements. After a certain point, additional revision produces marginal changes that may be improvements or may not be, and the author can't reliably tell the difference. Editors recognize this stage. Authors often don't, and continue revising indefinitely.

The signal that productive revision has ended is when the author finds themselves making different choices on different days without being able to articulate why one is better than another. At that point, additional revision is tinkering, and the manuscript needs either external feedback (beta readers, professional editing) or to move forward to its next stage.

When to stop revising and start submitting

The hardest part of revision is knowing when to stop. Authors who can't stop revising can't finish books. Authors who stop too early send weak manuscripts out before they're ready. The honest answer is that the manuscript is ready when the author has done every revision they know how to do and the remaining problems are either invisible to them or beyond their current craft level. At that point, external help becomes useful: critique partners, beta readers, professional editors.

For the full treatment of revision sequencing, multi-pass strategy, self-editing techniques, and when to stop, see our pillar guide on how to revise a novel. Sub-articles deep-dive into self-editing checklists, reading your own manuscript like an editor, multi-pass revision, and how to cut words from a manuscript without losing the story.

How to Diagnose What Your Manuscript Actually Needs

Most authors approaching revision know something is off but can't identify what. Diagnosis is the first step. The pillars above each connect to specific symptoms.

If readers tell you the prose feels flat

This is usually a prose mechanics issue. Sentence monotony, voice inconsistency, filler-word saturation, or generic word choice. The fix is a targeted prose pass focused on sentence variety, removing filler, sharpening verbs and nouns, and bringing the voice into consistent register.

If readers tell you the middle drags

This is usually a scene-and-pacing issue. Scenes that don't earn their place, summary that's been dramatized when it should be compressed, subplot threads that lose energy, or character arcs that flatten. The fix is a structural revision pass focused on identifying scenes that don't have jobs and cutting or compressing them.

If readers tell you the characters don't feel real

This is usually a dialogue and interior issue. Dialogue that doesn't sound like real speech, characters who all sound alike, interior monologue that explains rather than reveals, or character actions that don't match their established traits. The fix is character-focused revision: dialogue pass for distinct voices, interior pass for show vs tell balance, and consistency check across the manuscript.

If readers tell you the prose is overwritten or purple

This is usually a figurative language issue. Too many similes, mixed metaphors, decorative description that doesn't earn its place, or symbolism that's too heavy-handed. The fix is a figurative language audit: keep the devices that do work the literal description couldn't do, cut the ones that are decorative.

If readers tell you the opening doesn't grab them

This is usually a combination of issues concentrated in the first 50 pages: slow scene construction, weak character introduction, opening voice that hasn't found itself, or backstory that's been front-loaded when it should be threaded later. The opening gets disproportionate attention in revision because it's what agents and readers see first.

What a Professional Editor Sees That Self-Editing Doesn't

There's a limit to how much an author can improve their own manuscript through self-editing, and the limit isn't talent or effort. It's familiarity. After months or years with a manuscript, an author reads what they intended to write rather than what's on the page. The brain fills in the missing word, corrects the grammatical error, and resolves the ambiguous pronoun reference before consciously noticing any of them. This isn't a failure of attention. It's how human cognition works with familiar material.

A professional editor reads the manuscript the way the reader will. As a cold reader, with no prior knowledge of the author's intentions, catching every place where the text fails to communicate what was meant. This is the gap self-editing cannot close, regardless of how skilled the author is at their own craft.

Different editing services address different layers of craft. Developmental editing addresses structure, scene construction, character arcs, and the foundational craft questions covered in pillars 3 and 4. Line editing addresses prose mechanics, sentence rhythm, voice consistency, and figurative language balance, covering pillars 1 and 2 at the prose level. Copy editing addresses grammar, consistency, and technical correctness. Proofreading catches what remains after everything else.

Authors who invest in professional editing typically do so at one of two points: after their best self-revision is complete and they want external feedback before querying or self-publishing, or after they've received initial feedback from beta readers and want professional perspective on the issues raised. Both timings work. The earlier investment usually produces more dramatic improvement because the editor catches structural and prose-level issues that affect every subsequent revision.

For more on the editing question and on which service fits which manuscript stage, see our guide on whether you need a book editor either way.

What to Do Next

If you've finished a draft and are trying to figure out what to revise, the order of operations is straightforward.

1. Read the full manuscript in as few sittings as possible. Two or three days, ideally. This is the only way to see the book as a whole rather than as a sequence of chapters.
2. Identify which pillar your manuscript most needs work in. Figurative language, prose mechanics, scene construction, or revision strategy. The diagnosis section above is the starting point.
3. Read the relevant pillar guide and 2-3 spoke articles. Each pillar has deeper guides on the specific elements within it.
4. Do a single focused revision pass on that pillar. Don't try to fix everything at once. One layer of craft, all the way through the manuscript, then stop and assess.
5. Move to the next pillar. After a successful single-pillar pass, identify what the manuscript needs next and revise accordingly.
6. Bring in external readers and editors when self-editing reaches its limit. Beta readers, critique partners, professional editors. The right combination depends on what you can identify yourself and what you need someone else to see.

Editing Support at Every Craft Layer

Editor World connects authors with native English editors who have genre-specific experience and verified client ratings. Different editing services address different layers of craft, from structural developmental editing for foundational scene-and-character issues to line editing for prose-level revision to copy editing and proofreading for technical correctness.

You browse editor profiles by genre experience and category match, select the editor whose background fits your manuscript, and message them directly before submitting to discuss what kind of editorial work would most help your draft. A free sample edit is available on request. Every editor is a native English speaker from the United States, the United Kingdom, or Canada. No AI tools are used at any stage. A certificate of editing confirming human-only editing is available as an optional add-on.

Visit the book editing services page for full details on what's included at each editorial stage. Use the instant price calculator to see your exact cost before committing. Or browse available editors to find the right match for your manuscript and the layer of craft you're working on.

Frequently Asked Questions About Writing Craft

This article was reviewed by the Editor World editorial team. Editor World, founded in 2010 by Patti Fisher, PhD, provides professional editing and proofreading services for academic researchers, graduate students, businesses, and authors worldwide. BBB A+ accredited since 2010 with 5.0/5 Google Reviews and 5.0/5 Facebook Reviews. More than 100 million words edited for over 8,000 clients in 65+ countries. No AI tools are used at any stage.