CaddieOpen in CaddieDownload for Mac

How to cut and rewrite your own draft

Shared by Caddie Library · Updated
Chapters show their main points
Chapter 20 of 20 · Revising, Habits and Readers
← The growth course
1. Finding Product-Market Fit
Test for real customer pull earlyPick a narrow niche to start inLaunch a simple first version fastLearn from users and customersRead early signs of product-market fit
2. Onboarding and Activation
Structure onboarding as a storyAsk and use onboarding questionsGet users to the aha moment fastDecide how long onboarding should beBuild quiz funnels for paid traffic
3. Paywalls and Pricing
When to choose a hard paywall over freemiumDesign a paywall that convertsPrice annual and monthly plansSet, raise and localize pricesRun free trials and free tastes
4. Business Models and Revenue
Structure subscription tiersKeep a free tier that pays offCharge for usage and AI creditsPair hardware with a subscriptionAdd ads, affiliates and partners
5. Retention and Engagement
Measure retention honestlyBuild habit loops that bring users backUse notifications and lifecycle messagesReduce churn and handle cancellationsBuild social loops and community
6. Brand and Positioning
Position against bigger incumbentsName and describe your productEarn trust in high-stakes categoriesBorrow credibility from celebrities and experts
7. Organic Content and Creators
Find, vet and recruit creatorsPay creators fairly and safelyBrief and manage creatorsFind and copy viral formatsScale one format across many accountsRank in AI search and SEO
8. Paid, App Store and Viral Growth
Make ads that convertTurn organic winners into paid adsStructure and scale Meta campaignsTell if paid acquisition is profitableWin in the App StoreBuild viral and referral loops
9. Experimentation and Analytics
Run trustworthy experimentsJudge test results correctlyRun more experiments fasterWhat metrics to track and how to define them
10. Building AI Products
Build for the models of tomorrowDesign AI interfaces beyond chatDesign AI agents people can trustMeasure AI quality with evalsControl AI costs and pricingBuild products with AI coding tools
11. Teams, Hiring and Funding
Hire people who thrive on ambiguityOrganize small product teamsRun exploratory bets and kill themWork as a team with AI agentsWhen to raise money or sell your company
12. Founder Principles
Pick a mission worth building a company aroundWhy to trust your own judgment over the consensusTurn failure and setbacks into the next stepLearn from rivals and copy what works
13. Choosing Where to Compete
Make low cost your competitive advantageCreate a market instead of fighting for shareWhy scarcity and brand can beat volumeBuild an advantage that feeds itselfAttack a bigger rival where it cannot followWhen to stay focused on the core business
14. Building and Leading the Company
Hire great people and give them roomAlign pay and ownership with the missionKeep your culture and standards as you growBuild candor and fast decision-makingWhy leaders must stay close to the work and customersChoose and prepare a successor
15. Capital and Deals
Fund growth without losing controlAcquire companies and keep what makes them workWhat to do with cash: reinvest, buy back or holdWhy to keep a cash cushion and avoid heavy debtStructure ownership to protect your missionNegotiate deals that keep your leverage
16. Finding Ideas and Thinking
How to find ideas worth writing aboutWhy writing is how you find out what you thinkHow to research so the writing carries feelingHow to read like a writer
17. Structure and Argument
How to open so readers keep goingHow to find the shape before you draftHow to keep readers curious with questions and cluesHow to argue without preachingHow to end so the last page lingers
18. Clear Sentences and Voice
How to write plainly for a reader who knows nothingHow to choose specific details over generic onesHow to sound like yourself on the pageHow to shape rhythm and phrases people remember
19. Characters and Storytelling
How to build characters readers care aboutWhat makes a scene move: want, obstacle and changeHow to write dialogue that sounds real but stays shortHow to make readers feel without telling themHow to turn information into a story
20. Revising, Habits and Readers
How to cut and rewrite your own draftHow to get feedback you can useHow to build a writing routine that lastsHow to get unstuck when a draft stallsHow to use AI without losing your voiceHow to write for one specific reader

1 The bottom line

Revise in separate passes: draft loose, read aloud, cut what the piece does not need, and stop when you can only make it worse. Readers want the story, not everything you know, so cutting is how a draft becomes a piece.

2 Key takeaways

  • Draft first, then edit in a different frame of mind. Judging while you write blocks the writing, and judging too soon blocks the editing.
    • Scott Dikkers50:09 leaves a draft for days, or up to two weeks, until he barely remembers writing it.
    • Wright Thompson20:35 wrote 280,000 words of The Barn fast, then cut 173,000, leaving about 107,000.
  • Read the draft aloud before you call it edited. Your ear catches repeats, stumbles and boredom that your eyes skip.
    • Peter Attia43:28 found sentences that looked fine on the page but were absurdly long when spoken, and rewrote them after typesetting.
    • Ben Thompson26:45 says his typo rate dropped to about one a month once he began reading every piece aloud.
  • Cut anything that does not serve the reader, even your best material. The reader does not care what a fact cost you to find.
    • David Grann58:58 cut 10,000 to 20,000 words on building a ship down to a few facts.
    • Johann Hari28:12 leaves out roughly 80% of his 200 to 300 interviews per book.
  • Give each pass one job. A single lens finds more than one read from top to bottom.
    • Michael Dean51:30 highlights one pattern in orange and ignores structure and thesis for that pass.
    • R.F. Kuang44:22 saves faces, architecture and furniture for a last, deliberate pass.
  • Know when to stop. Past a point, each round moves the draft away from what was good in it.
    • Matt Haig5:54 says diminishing returns arrive around the fourth or fifth round.

if you get rid of it and you don't notice it's gone, then turns out you didn't need it

Francine Prose, Reading Like a Writer

2.1 Step back before you edit

Make like a nurturing baby, edit like a harsh officer
MakingEditing
RoleNurturing babyHarsh officer
RuleDon't judge, get it on paperCut and polish every word
SettingDesk, morningCouch, evening
TimingDraft firstAfter days or weeks away
Separating the two phases, with rest in between, turns the writer into a reader of their own draft.

Make the draft in one mode and judge it in another, with time in between.

  • Keep making and editing in separate phases. Jerry Seinfeld35:54 says flipping between the two is the key to writing.
    • While you make material, be a nurturing baby. Do not judge, just get things on paper.
    • When you edit, be a harsh officer who cuts and polishes every word.
    • In that conversation, Perell asks which phase he is in each time he sits down. He creates at a desk in the morning and edits on a couch in the evening.
  • Let the draft rest so you can forget it. Distance turns you into a reader.
    • Michael Pollan15:30 steps away for a couple of days or weeks. What sticks is what he focuses on, and he says forgetting is how you edit.
    • Dikkers50:09 returns as "editor brain" and judges the piece almost as if someone else wrote it.
    • Ryan Holiday24:55 asked for an extra year on Right Thing, Right Now, let the draft sit, then spent his edit mostly cutting.
  • Write the first draft for yourself, then edit it for the reader. The second draft has a different owner.
    • Amor Towles7:13 lets the first draft hold every whim and digression. Then he asks what he owes a reader who spent money and time.
    • Towles7:13 cuts redundancies, cliches and anything that only satisfies his own fancy.
    • Richard Powers45:35 treats the first draft as a note to himself about the effects he wants. On the rewrite he removes the notes that announce his intentions.
  • Reread as someone who knows nothing about it. Steven Pressfield28:49 does this on every new draft.
    • He puts himself fully in the reader's seat and asks whether the story would pull in a stranger.
    • Sometimes he imagines a specific friend, Jim, and asks whether that person would be bored or interested.

2.2 Edit in focused passes

Ask one question per pass, and save the surface for last.

  • Pick one pattern and mark it everywhere. Michael Dean51:30 calls it a razor lens on one pattern at a time.
    • Highlight every place the pattern shows up, such as voice, in one color. Ignore structure and thesis for this pass.
    • Sections with no highlights show where the voice vanishes. Rewrite those.
    • Michael Jamin23:53 goes through after the first draft like a drinking game, finding repeated theme words such as "forgive" and swapping in synonyms.
  • Leave faces and furniture for the last layer. R.F. Kuang44:22 keeps early drafts to placeholders.
    • She skips character faces, architecture, furniture and the color of the drapes. She finds appearance the least interesting thing about a character.
    • She compares it to animation, "balls on sticks for a long time", before she makes those choices on purpose.
    • Details still need a check at the end. A copy editor told Elif Shafak38:43 that a 1943 hotel had pink curtains, not green.
  • Find the images that repeat and strengthen them. Andrew Sean Greer1:10 uses the second draft for this.
    • Note which images recur. If birds keep showing up, cut the porcupine stories and add more birds.
    • Readers enjoy a pattern they can feel but cannot quite name.
    • If you cannot say what the draft is about, start every sentence with "This is a story about." Tom Junod47:03 says it feels like second grade but forces the real subject out.
  • Mark where attention slips. Pacing problems are easier to see on the page than to feel.
    • Daniel Pink12:10 circles every laugh in a speech to see the pacing. Too many in one place feels like stand-up.
    • Andrew Stanton36:02 asks "when did we accidentally unclamp?" He looks for where the story let go, or never gripped.
    • In one play, Pink12:10 saw a long plot-heavy stretch with no laughs after four scenes of steady ones, and added some.
    • Morgan Housel41:20 says to find the exhaustion point in your own material and cut it off there. He pushes event planners from a 60-minute keynote to 30.
  • Print it and mark it up on paper. Michael Connelly5:59 starts every morning this way.
    • He prints the previous day's pages, marks them in pencil, then types the changes into the file. That is three to ten pages.
    • At the end of a book he prints and marks up the whole manuscript before touching the file.
    • By then he knows the full story. He can cut details or set up something that pays off a hundred pages later.

2.3 Read it aloud

Your ear finds what your eyes skip.

  • Make reading aloud your first line of editing. Susan Orlean32:19 calls it her biggest trick.
    • You hear sentences that do not work, repeated words and spots that are not clicking.
    • If you get bored while reading, that part is probably boring. Fix it or cut it.
    • Jodi Picoult37:29 says it also catches repeated words in the same sentence and helps with pacing and dialogue.
    • Peter Attia43:28 says Sam Harris told him silent reading does not count as editing.
  • Fix every sentence you stumble on. A stumble means something is wrong with the text.
    • Steven Pinker21:01 says that if you cannot say a sentence smoothly, your reader will not sound it out smoothly in their head either.
    • Brie Wolfson28:44 records herself reading and plays it back. Sentences that do not sound like her stand out.
    • Lulu Cheng Meservey1:40 says a phrase you would never say to a person will refuse to come out of your mouth. That is her defense against corporate speak.
    • Paul Millerd6:13 asks "Does this one feel true?" of each line and rewrote some chapters 50 to 60 times. Recording the audiobook gave him friction only about three times.
  • Get a voice to read it to you if you need to. A tool works as well as your own mouth.
    • Nick Bilton9:34 pastes a screenplay into Speechify or ElevenLabs and listens to it.
    • Taylor Tomlinson37:00 records audio of every show and listens back on planes or walks. Sets never sound how she thought they went.
  • Fix wherever the rhythm stalls. Rhythm tells you where a line is not working.
    • Mitch Albom1:02:56 rocks while he writes and stops when a line is not working. When he stops, he goes back and reworks it until the beat flows.
    • Jayne Anne Phillips21:41 says to edit with your ears, since language is music. A passage that goes off sounds like a flat note on a piano.
    • Phillips21:41 adds that this only works once there is something solid on the page.

2.4 Cut what the piece does not need

Test every line, scene and fact against what the reader came for.

  • Ask of each line whether you would miss it. If you would not, delete it.
    • Francine Prose53:57 says to remove the line and see whether you notice it is gone. A character lighting a cigarette or refilling a coffee cup should reveal something or go.
    • Jonathan Franzen23:51 asks of each sentence "do I need this?", then whether it can use fewer words.
    • Gillian Flynn1:01:35 learned from David Fincher to say aloud why a scene exists. If you cannot, it probably should not be there.
    • Beginners write what they are picturing, such as walking across the room, turning the doorknob and opening the door Franzen23:51.
    • Cutting can overshoot. Flynn1:01:35 wrote her first Gone Girl script at exactly 120 pages, and Fincher's first note was to put more of the book back.
  • Cut the research you worked hardest to find. The reader only cares what a fact makes them see.
    • David Grann58:58 wrote 10,000 to 20,000 words on building an 18th century ship. His first reader groaned and he cut it to a few facts, such as the 4,000 trees one ship took.
    • David Baldacci1:08:52 keeps setting and period detail to three or four sentences. Beyond that is "ego writing", and he leaves out 99% of his research.
    • Johann Hari28:12 says what goes in the book is not the most interesting find. If it does not serve your questions and the journey, it goes.
  • Cut everything that is off your theme. Name the theme, then test each piece against it.
    • Rob Henderson10:36 asked of each piece whether it centered on family, foster care and class. His first manuscript ran about 100,000 words and he cut roughly 20 percent.
    • Robert Greene21:23 cut Harriet Tubman's whole life down to one thread, the voice she heard and her sense of fate.
    • Alex Hormozi40:55 cuts any passage if he cannot say what the reader would do differently because of it.
  • Set a number and trim to it. A target makes cutting concrete.
    • Scott Galloway50:21 adds metaphor and emotion, then removes whatever does not earn its place. He regularly turns an 1,800 word post into an 1,100 word post.
    • Sam Altman23:56 retells a teacher's exercise: cut one metaphor from every page and one unnecessary word from every sentence. Often students found there was no story, only pretty language.
    • David Zucker11:09 reads the page visually. Any speech over about six lines looks a quarter inch too long, so he cuts words until it fits.
    • Steven Levitan1:22:45 cuts half jokes, because a chuckle weakens the real joke.

2.5 Rewrite what stays

The hardest passages take dozens of drafts, not five
Simple Pixar sequence6 drafts
Pixar opening staff meeting60 drafts
Dear New York prologue180 drafts
Michael Arndt and Brandon Stanton show that draft counts rise with how much a passage has to carry.

The hard passages need many versions, and some cuts need to be undone.

  • Expect the most drafts where the most is going on. Heavy passages take far longer than simple ones.
    • Michael Arndt5:25 says a simple Pixar sequence took five or six drafts. The opening staff meeting, loaded with exposition, took about 60.
    • Brandon Stanton45:54 ran the 12,000-word prologue of Dear New York through 180 drafts to keep its many threads of one piece.
    • Harry Dry1:00:05 copies a paragraph and rewrites it a different way, ending with four or five versions. Showing versions gets him better feedback than one paragraph does.
  • If a passage reads wrong, delete it and run up again. Mark Forsyth51:47 compares it to a long jumper retaking the run-up.
    • He warns that fixing one sentence at a time can kill the rhythm that should have been there in the first place.
    • He also warns against changing one word, then another, because you get stuck.
  • Cut too much, then put the breathing room back. The overcorrection shows you where space belongs.
    • David Gelb44:56 finds his first assembly far too long. His second cut removes too much and goes too fast, and the right pace lands in the middle.
    • Ryan Holiday24:55 mended the seams after cutting and added back what the leaner version needed.
    • Holiday's24:55 manuscript went from about 85,000 words to roughly 65,000, then back up to 70,000.
    • To make cutting feel safe, Harlan Coben8:07 moves doubtful passages into a file called spare. It usually holds about 50 pages, and he has never put anything back in 37 or 38 novels.
  • Fix a flat draft with a real example or a twist. Flat drafts often copy something you already wrote.
    • Jack and Ryan of AJR29:14 say their failed songs sound like photocopies of their earlier songs. A draft comes alive when they find a real example or a twist.
    • Ocean Vuong37:57 searches his best lines online. His teacher Ben Lerner showed him "300,000 people beat you to it", and a flood of matches is a signal to go higher.

2.6 Know when to stop

Quit while the draft is still alive, and hand it over only when you have done your part.

  • Stop when you can only make it worse. Over-editing drifts from what was good.
    • Matt Haig5:54 says each draft improves things only up to a point, around the fourth or fifth round. After that the book gets worse.
    • Jason Fried11:56 finds his drafts get wordier the longer he works them. He stops as soon as he feels he could only make it worse, and fixes a typo later.
  • Exhaust yourself on the draft before an editor sees it. Then the editor's time goes to what you cannot see.
    • David Sedaris30:30 would rather give his editor the 15th draft of a 17-draft piece than the third. Having worked it hard, he can also defend what matters, like his biggest laugh.
    • Tish Rabe12:53 writes several versions by hand alone. She still has an editor read everything, because you are too close to your own work.
    • Ana Lorena Fabrega32:29 accepted her editors' calls to cut paragraphs she loved that added nothing. Her husband flagged the most.
  • Leave time for cutting instead of a final sprint. A rushed end leaves loose prose.
    • Dan Wang19:20 wrote his letters in a hasty last-ten-day sprint and wished they had been better edited and trimmed.
    • He is about 85% satisfied with the book and wanted another month to refine the links between sentences and paragraphs. New Yorker writers told him no writer gets above 85%.

2.7 What the books say

Facts that share nothing support no conclusion
From Barbara Minto's Pyramid Principle: being true is not enough reason for a fact to stay.

The book says to fix the order and the content of your ideas first, and to cut any fact that does not belong with its neighbors.

  • Fix structure before you polish sentences. Barbara Minto, The Minto Pyramid Principle
    • Most unclear writing comes from the order of ideas, not from style.
    • Structure is far easier to change than style, which takes long, repeated practice.
    • When the structure is broken, treat it as a thinking problem to solve, not a wording problem.
  • Cut facts that share nothing with their neighbors. Barbara Minto, The Minto Pyramid Principle
    • Facts with no common subject or predicate support no inference. They are news, not thinking.
    • Being true is not a reason to keep something in.
    • Minto's example: Japanese drives into China, Iceland and Peru share nothing and support no insight. Drives into China, Indonesia and Australia support moving into Southeast Asia.

2.8 Where writers disagree

Writers split on when to polish, how to fix a bad passage and how many rounds are enough.

  • Polish as you go, or draft fast and cut later. Both routes end in a clean book.
    • Dean Koontz6:06 perfects each page before the next, at 20 or 30 drafts per page. He says that spares him a big revision pass and a draft whose ending he dislikes.
    • Wright Thompson20:35 races through a first draft, then edits slowly and cuts hard. He has tried to slow down and it does not suit him.
    • Elif Shafak38:43 edits as she drafts and then keeps jumping back and forth between sections.
  • Edit in place, or delete and rewrite. The answer depends on what is alive in the draft.
    • Mark Forsyth51:47 dislikes editing sentence by sentence and prefers to delete a passage that reads wrong and write it again.
    • Henry Shukman43:45 gets back into the mindset the piece came from before rewriting, so he does not "cut live flesh". Rewriting usually takes him much longer than the first writing.
    • Shukman will tinker with a poem for a year and find it almost the same as at the start.
  • Stop early, or keep going for dozens of drafts. The stakes of the passage seem to decide it.
    • Jason Fried11:56 thinks the short first version is usually the best.
    • Michael Arndt5:25 wrote about 100 drafts of Little Miss Sunshine before it went into production.
    • Robert Greene21:23 says he rewrites a story hundreds of times, which he admits may be an exaggeration, until it gives the feeling he wants.

Sources

Part of Revising, Habits and Readers in the growth library.