Back to the table

73 Beginner Be Beat Plot & Beats

Beat

One meaningful step in the story.

What it is

A beat is the smallest unit of change in a story. Everything useful about the word comes out of those two halves. Smallest means a beat sits below the scene. A scene is made of several beats, and a scene with only one usually reads as thin. Change means a beat is more than something that occurs. Rain occurs. A character wiping rain off a window occurs. Neither is a beat. Now say the character notices, through the rain, that the car parked across the street is the same car as yesterday. That's a beat, because the situation isn't what it was a line ago.

Robert McKee, in Story (1997), calls it an exchange of behavior in action and reaction, and he calls the beat the smallest element of structure there is. He wrote that for screenwriters, and it carries over to prose with one adjustment. In a novel, many beats have no visible behavior at all. A character works something out on the inside, and the working out is the beat. Nobody in the room can see it. The reader can.

Here's a test you can run on a paragraph. It has two parts:

Run those two questions along a page of dialogue and you'll find the lines where the scene moves and the lines that are marking time. Run them down a chapter and you'll find out whether the chapter is a chapter at all. The test doesn't ask for drama. The smallest change you can count in a book can be a character deciding not to say something. A beat is also smaller than the value that turns across a whole scene, which is a bigger, rougher measurement. Value Shift covers that, at the scale of the scene and above.

Now for the part nobody explains, which is why two capable people can argue about beats for an hour without noticing they're talking about different things. Four crafts use this one word for four different things:

So how do you tell which one someone means? Ask what the unit is made of. If the answer is a page or two, it's the fourth sense. If it's a moment, it's the first. If it's a tactic, you're talking to someone with theater training, and it's the second. If it's a comma's worth of silence, it's the third. That one question settles nearly every argument about beats, and asking it out loud in a meeting isn't rude, because everyone in the room has been quietly guessing. There's a fifth use too, so it doesn't trip you up. In journalism, a reporter's beat is the subject they're assigned to cover, and in music a beat is a pulse. Both are older than the craft senses, and neither has anything to do with fiction.

Story craft has a ladder of units, and it helps to know both the ladder and where it stops being exact. A beat is one change. A scene is an unbroken stretch in one place and time that adds up to something, and that page defines it properly. A sequence is a run of scenes that work as one movement, and the classic example is everything that happens between arriving somewhere and getting out again. A chapter is a packaging decision about where the reader is allowed to stop, and it's the one rung on the ladder that isn't a story unit at all. An act is a large division of the whole, marked by a change so big the book can't go back.

Here's what craft books rarely say plainly: these boundaries are habits people agreed on, not things found in nature. Nobody can say for certain where one beat ends and the next begins, the way you can say where a sentence ends. Two experienced editors reading the same page will count four beats and six, and both will be reading well. You could describe a long negotiation as one beat ("he refuses"), as three ("she asks, he refuses, she asks differently"), or as eleven. The ladder works like a zoom level on a map, and the right setting is whichever one makes your current problem visible. None of that is a reason to distrust the unit. Nobody thinks a mile is fake because you could have used kilometers. It does give you two habits to watch. The first is arguing about counts. If you and your editor disagree about whether a chapter has three beats or five, you're disagreeing about zoom, and you can settle it in ten seconds by agreeing on one. The second is treating a number as a target. Nobody can enforce "every scene needs three beats", because the person making the rule chose the zoom level where it's true. McKee's version of the ladder (beat, scene, sequence, act, story) is the cleanest statement of it in print, and it comes with a claim to take seriously: each rung turns something bigger than the rung below. A beat changes the moment, a scene changes the afternoon, and an act changes the book.

Most working writers write beats down in one of three informal ways. The formal version belongs to Beat Sheet.

Whichever you use, phrase the beats as changes, not topics. "The fight about the estate" is a topic and could be anything. "She finds out the estate was already sold" is a beat, with a before and an after.

The page-level beat, the third sense, is where the word meets line editing, and it goes wrong in one typical way. The beat is useful. Drop a bit of physical action between two speeches and you get a pause you can measure, with no word of explanation, and you can swap an adverb for evidence. Dialogue covers silence and interruption as tools. The problem is the beat used as punctuation. In a first draft, a writer sensibly avoids a wall of "said" and reaches for an action instead, but the actions come from a very short list. Characters nod and sigh. They shrug, blink, look away, run a hand through their hair, sip coffee they never finish, and take a deep breath. Each one does nothing but mark a rhythm, which is a comma's job. After three pages of it, a scene of two people negotiating a divorce reads like an aerobics class. Spotting it is easy and a little humbling: search your manuscript for nodded, sighed, shrugged and took a breath, and look at the count. Then fix it in three moves:

Notice where this meets the first sense. A page-level beat that reveals something the reader didn't know, or visibly changes a character's footing, has just become a story beat as well. The best action beats do both jobs, which is why they're worth writing well instead of reaching for the nearest sigh.

Why it matters

The beat is the unit you can count. You can count almost nothing in a novel. Not tension, not voice. Two readers can fairly disagree about whether the middle drags, and neither can show their work. So revision often feels like weather. You know something is wrong and have nothing to measure it with. You can count beats. The count isn't perfect, and it only works once you've picked a zoom level. Even so, the numbers mean something when you compare them with each other. Ten minutes with a printout and a pen gives you a count for each chapter, and a column of those numbers tells you things about your book that reading it won't, because when you read your own book you fill in the missing energy from memory. Don't use the word as theory, and don't use it as a plan. Use it as a measuring stick.

It finds the fault in a flat scene. The most common note on a draft is that a scene is flat, and the most common response is to rewrite the prose. Rewriting the prose is nearly always the wrong fix. Flat prose and a flat scene are different problems. Flat prose feels like bad sentences. A flat scene feels like nothing is happening, and the sentences in it may be your best. So mark the beats in the margin, one note for each change. You'll usually find one of four things. No beats: four pages and no note anywhere, people swapping information they have no reason to hide, or a character thinking in circles. Better writing won't help this scene. It needs a change put into it, or it needs to shrink to a two-line summary. One beat, arriving last: a page and a half of buildup, then the thing. Cut the buildup, because the beat was always the scene and you were clearing your throat. Scene covers where to start and stop. Beats that repeat: three notes down the margin all saying the same thing in different words, she asks again, he refuses again, she asks again. Repetition can be deliberate and powerful when each round costs more, and that's the try-fail cycle. When each round costs the same, the scene has stalled, and you can cut two of the three rounds for free. Beats in the wrong order: your character learns the deciding fact and then does the thing that fact makes pointless. Readers feel this as slackness but can't say why. Swap the two beats and the scene tightens without a single new word. Four findings, four fixes, and every one costs less than a rewrite.

It catches two beats fighting. Sometimes a scene has two strong beats in it, and they compete. A character quits her job, and in the same twenty minutes she's told her brother has been arrested. Both are real, both change everything, and readers come away remembering neither clearly, because each beat stole the other's oxygen. The resignation had nowhere to land, and the arrest showed up while the reader was still busy with it. You'll know the signs. Beta readers remember the scene happened but misremember what was in it. You keep having to re-explain one of the two beats later, because it didn't stick. The scene is oddly long and still feels rushed. The fix is almost always to separate them, and the interesting question is which way around. You can give the smaller beat its own scene earlier, where it has room to be felt, and let the bigger one have this scene to itself. Or you can keep the big one here and push the smaller one later, where it lands on top of consequences the reader now understands. Turning one of them down doesn't work. If you muffle a beat to stop it competing, you've paid for it and you're not using it. There's a book-sized version too: two subplot climaxes land in the same chapter as the main one, so three endings arrive at once and none of them gets felt. Same fault, same fix.

It explains most sagging middles. Go through a manuscript chapter by chapter and note the beats, and somewhere in the second half you'll find a chapter with nothing in the margin. It might be travel, or preparation, or a meal where the cast talks over what has already happened, or a character walking around a city, sad about it. The prose may be excellent, and often it's the best writing in the book, because there was no plot to worry about. A chapter with no beats isn't automatically one to cut, but it's always one to make a decision about. You have three options. Give it one: the character walking around the city can, halfway around, decide something, or see something that goes against what they believe, and one beat is enough to make a chapter a chapter. Summarize it: if the reader needs the information but not the experience, two paragraphs will carry what the chapter was carrying. Exposition covers how to tell without stalling, and Pacing covers what this does to the shape of the book. Or keep it on purpose: sometimes a chapter without beats does something a beat can't. It lets the reader breathe after a run of big changes, or it shows a state instead of a shift. That's a real choice. Make it knowingly, once or twice in a book. Doing it four times in the same fifty pages is what a sagging middle is.

Revision is where the concept earns its keep. Everything above is about checking a draft, and that's on purpose. The beat is a so-so tool for coming up with a story and an excellent tool for measuring one. Invention is where it's weak. Knowing that a change should happen here doesn't tell you what should change, and no list of slots will supply what goes in them. Writers who plan a book beat by beat before they know what the book is usually end up with something the right shape and lifeless. Plot sets out the same reason. As a measuring tool, it's the best thing in the toolbox, because the bigger the unit you're revising, the more the revision costs. Rewriting a book takes months. Rewriting a chapter takes days. Moving, cutting or sharpening a beat takes minutes to hours, and a surprising number of manuscript problems turn out to be beat-sized once you pin down where they are. Think about what a beat audit lets you do in a week, without drafting a single new chapter. You can find the two beats that are the same beat and delete one. You can move a reveal four chapters earlier, where it makes everything after it more interesting, instead of four chapters later, where all it does is explain. You can find the place a reader said they got bored and discover it's twelve pages before the place they named, because that's where the last change happened. You can take the chapter with no beats and fold its one useful paragraph into the chapter before it, losing three thousand words and nothing else. That makes the beat the right unit for a production schedule. If you write several books a year, you need faults you can fix in one sitting. This page argues for what the smallest pass should count. Revision is the page on what order to do your passes in. And the beat is the only fair way to answer "is this scene earning its length?" Don't go by word count or by whether you like it. Go by what changes in it, and whether that change needed this many pages.

It stops you having the wrong argument. Writers work with editors, cover designers, audiobook narrators, critique partners and, more and more, collaborators trained in film, TV or theater. All of them use "beat", and they don't all mean the same thing. A narrator marking beats in your manuscript is marking breaths. A screenwriter friend who reads your synopsis and says it's missing a beat means a structural landmark the size of a chapter. An actor in your writing group means a change of tactic inside a conversation. A line editor means a gesture between two speeches. If you don't notice, you end up doing the wrong work with total confidence. Picture a writer who's told her manuscript needs more beats. She adds three hundred nods and sighs, when the person who said it meant that the second act has no reversals in it, and she has spent a week making the book worse. Asking which beat someone means takes one sentence and saves that week.

Real-world examples

Bram Stoker's Dracula (1897): five beats in three hundred words

Dracula is free at Project Gutenberg, so you can check this one in about two minutes, which is why it was chosen. In Chapter II, in Jonathan Harker's journal entry dated May 8, a passage of roughly three hundred words holds five clean beats.

Harker has hung his shaving glass by the window and started to shave. A hand lands on his shoulder and the Count says good morning, and Harker starts and cuts himself. Beat one: he's bleeding, and something is wrong, because the mirror showed the whole room and no one in it.

He turns back to the glass to see how he was mistaken. The Count is close behind him, clearly visible over his shoulder, and there's no reflection. Beat two: what was a puzzle a line ago is now impossible, and Harker knows it. Nothing physical happens in this beat at all. It's a discovery, and it's the biggest change in the passage.

The Count sees the blood on Harker's chin, his eyes blaze, and he makes a grab at Harker's throat. Beat three: the host is a threat, and the relationship the whole chapter has been built on has just flipped.

Harker pulls away, and the Count's hand touches the string of beads holding the crucifix, which an old woman at the inn pressed on Harker in the opening chapter. The fury vanishes at once. Beat four: the threat has a weakness, and Harker is carrying it without understanding why.

The Count warns him to be careful how he cuts himself and calls the mirror "a foul bauble of man's vanity". Then he wrenches the window open, throws the mirror into the courtyard and leaves without another word. Beat five: the evidence is gone, and the pretense of a normal household is over.

Five changes in under a page, and each one leaves the situation different from the line before. That density is what this element is about, and it's why the passage is still frightening after more than a century of imitation. Try the counting exercise on it, then on a page of your own. The comparison usually teaches you something, and sometimes it's unpleasant. Two footnotes, for accuracy. The crucifix is a setup planted earlier, so it's foreshadowing doing its job, not a beat in this scene. And a reader could fairly count six beats here, by splitting the cut from the missing reflection. That disagreement is the zoom problem described above, and it changes nothing about the diagnosis.

Bits, beats, and a Russian accent: where the acting sense came from

No word history in story craft is better known than this one, and people repeat it everywhere as fact. They shouldn't.

The claim goes like this. Konstantin Stanislavski broke a script into small pieces for rehearsal, and his Russian word for such a piece was kusok, an ordinary word that means a bit or a slice, as of bread. When his method reached America, the story goes, Russian émigré teachers told their students to break the script into little bits, and in a heavy accent "bits" came out as "beats". American actors wrote down what they heard and the word stuck. So a whole vocabulary of dramatic structure, including the beat sheets on this table, comes from a misheard vowel.

Here's what can be shown and what can't. The established part: kusok is the term Stanislavski himself preferred, and it means bit or piece. Neither the Russian nor the published English translations gave anyone "beat". The standard early translations, starting with Elizabeth Reynolds Hapgood's An Actor Prepares in 1936, used "unit". Jean Benedetti's complete translation, published by Routledge in 2008 as An Actor's Work, uses "bit". Its chapter on the subject is titled "Bits and Tasks", and its glossary lists the term as "Kusok, Bit, Unit". Meanwhile "beat" became the main term in American practice, so the word everyone uses today matches neither the original nor either translation. That much is on record. Scene Goal covers what novelists took from the technique behind it.

The suggested part is the accent story itself. Sharon Marie Carnicke, Professor of Theatre and Slavic Languages at the University of Southern California, gives the careful version in Stanislavsky in Focus (second edition, Routledge, 2009). She says the change from bits to beats may come from the way émigré Russian teachers in America pronounced it, perhaps helped along by the image of beads on a string. She also records the suggestion that Richard Boleslavsky's spoken teaching may have set off the mistranslation. Boleslavsky was a Moscow Art Theatre actor who founded the American Laboratory Theatre in New York in 1923 with Maria Ouspenskaya, and he taught Lee Strasberg, Stella Adler and Harold Clurman. Look at the verbs. May derive. It has been suggested.

Nobody is disputing it with a rival theory, and it would be convenient to invent one. The dispute is quieter than that, and it's about evidence. The scholars closest to the material hedge, and everyone who repeats them later doesn't. Carnicke offers the accent story as one possibility and the beads image as another, which is how someone writes after looking for a document and not finding one. Nobody has turned up a transcript, a letter or an account from the time in which a Russian teacher says "bits" and an American writes "beats". What survives is the result, a word everyone uses that matches neither the Russian nor the translations, plus a believable way it could have happened. Believable isn't the same as proven.

The same caution goes for the screenwriting question, which people ask all the time and which has no clean answer. Nobody has shown who first used "beat" in a screenwriting sense, or when the pause-on-the-page sense split off from the unit-of-intention sense. Both senses are in ordinary dictionaries now, which tells you they're settled usage and nothing about where they came from. Anyone who gives you a first user and a date for this is guessing with confidence.

So why tell you all this? The small reason is that craft writing is full of unsourced stories repeated as fact, and it's a good habit to stop repeating them. The bigger reason is that the story, true or not, holds a useful point. The unit was never about rhythm. It was about breaking a whole into pieces small enough to work on. Everything practical on this page follows from that, and none of it depends on how the word got its vowel.

Groundhog Day (1993): the same scene, and the only thing that changes

Groundhog Day was written by Danny Rubin and Harold Ramis from Rubin's story, and Ramis directed it. No popular film shows the beat as a unit more clearly, and it does it by accident, because the premise forces it.

Phil Connors lives the same day over and over, so the film replays the same scenes: the same sidewalk, the same insurance salesman, the same diner, the same woman at the same table. The audience has already seen them. What they're watching the second, fifth and twentieth time isn't the scene, because the scene stays fixed. What has changed is the beats inside it. And because everything else is held still, you can see the beat in a way you almost never can.

Take the courtship. Phil wants Rita, and he gets an unlimited number of tries at one evening. So he runs the evening as an experiment. He learns one fact, changes one beat, and watches what happens. In the January 1992 shooting draft, he offers a toast "to the groundhog" himself and gets a stare. He finds out she always drinks a toast to world peace, so on a later pass he proposes her toast before she can. He finds out she studied nineteenth-century French poetry, and his first reaction is to laugh out loud about what a waste that is. On a later pass, he has read the poets and can quote them. Each round is the same scene with a single beat swapped, and the film asks the audience to notice the swap and keep score.

So the sequence is funny, then uncomfortable, and then the moral center of the film. Manipulating another person one beat at a time is just what it looks like. In the end, the story's answer is that nobody can win the evening by putting the right beats together, because the thing being held back was never in the beats. That's as good an argument about the limits of technique as any craft book has made.

For a working writer, the lesson is mechanical and useful right away. If you want to see what a beat is, take one of your own scenes and write it twice, changing one thing only. Don't change the setting or the prose, just one thing someone does or learns. The difference between the two versions is one beat, and you'll be able to feel how big it is.

One caution about the film's own numbers, since people pass them around as if they were settled. The film never says on screen how long the loop lasts, and the people who made it have given different answers. Ramis guessed about ten years on the DVD commentary, then said in 2009 that, given how long it takes to get good at something, it had to be more like thirty or forty. Rubin's original concept draft had Phil estimating seventy or eighty years, which he tracked with books. Rubin said in 2005 that for him it was a lifetime. So treat any confident figure you see quoted as somebody's estimate.

The case against plotting by beats: Snyder (2005), Suderman (2013), and the answers

The criticism is real and well argued, and a page that dodged it would be worthless, so here it is at full strength.

Blake Snyder (1957 to 2009) published Save the Cat! The Last Book on Screenwriting You'll Ever Need with Michael Wiese Productions in 2005. What it added to earlier structure manuals was detail: a list of fifteen named beats, plus suggested page numbers for them, that could expand into a board of about forty cards. Jessica Brody adapted it for novelists as Save the Cat! Writes a Novel (Ten Speed Press, 2018), and now nearly everyone in commercial fiction uses its vocabulary. What matters here is what the method did to the word "beat". The Save the Cat page covers the book itself.

The prosecution first. In Slate on July 19, 2013, the critic Peter Suderman published "Save the Movie!". He argued that Snyder's beat sheet had taken over Hollywood screenwriting as a template, not a guide, with films hitting the named beats at the set minute. He went through a run of recent releases and showed the timings: a theme-stated beat around the seven-minute mark, a catalyst around twelve, an act break in the mid-twenties. He concluded that the beat sheet makes films that are structurally interchangeable, that it narrows whose stories get told, and that a book is pushing originality out of the market. Then he did something better than just say so. He built the essay itself to the beat sheet and reported what that cost him. He had to rearrange passages, leave material out, and invent new material only to fill a beat that had to be there.

The novel-side version of the complaint is sharper, because novels have no page-count habits to fall back on. Writing on Mythcreants in December 2022, Oren Ashkenazi argued that once the beats are softened enough to fit any novel in any genre, they stop saying anything. A beat that boils down to "characters react when bad things happen" doesn't teach you anything, because it tells you where a scene goes without telling you how to write one. It's too vague to help, and at the same time too strict for books that start mid-action or bring their protagonist into the plot slowly.

Both arguments land. Anyone who's read slush, judged a contest or run a critique group has seen beat-sheet fiction. Beat-sheet fiction is correctly proportioned and competently written, and it never once surprises you. That's a real way for books to fail, and it isn't rare.

Now the defense, from weakest to strongest. First, the quick reply, which Christopher Boone made at No Film School on July 25, 2013, six days after Suderman's essay. The fault Suderman found is unoriginal stories, Boone said, and structure isn't what makes a story unoriginal. He pointed to films that take completely standard structural pieces and rearrange them: Memento, Adaptation, Eternal Sunshine of the Spotless Mind. If the structure is laid out well, he argued, the beats fall where they belong, not where a page number says.

Second, Suderman's own demonstration cuts both ways, more than he admits. He fitted a magazine essay about the business of screenwriting to a screenplay structure. If the template is loose enough to hold that, then a film that matches it is weak evidence that the film was built from it. Matching a shape that fits almost anything doesn't prove copying. It may prove the shape tells you very little, which is a real criticism of the beat sheet, and a different, milder one than "the beat sheet is why films are the same".

Third comes the point this page exists to make. The critique is aimed straight at the fourth sense, the named commercial list, and people often quote it as if it had knocked out the first sense. It doesn't touch the first sense. "Something should change here, and someone should be able to tell" isn't a formula. It has no proportions, no page numbers and no names. A book that fails it isn't unconventional, just lifeless. You can reject every beat sheet ever printed and still need beats, the same way you can reject every sonnet form and still need lines.

Fourth, and least comfortable for both sides, defenders usually claim too much about what's out there. Brody's position is that the fifteen beats turn up in almost every successful novel, whether the author meant them to or not. As a claim about real books, that's much weaker than it sounds, because the beats are defined so loosely that you could find them almost anywhere and nobody could prove you wrong. The fair defense of a beat list is narrower, and still useful. A beat list is a checklist of questions commercial readers tend to expect answers to, and a fast way to find out which ones your draft hasn't answered. Used that way, it checks a draft, like everything else on this page. Used as a plan, it produces the books its critics describe.

What this means for your afternoon: if you're using a beat sheet and the book is dead, adding beats won't bring it back, and neither will switching methods. Go back to what each beat is supposed to change and ask whether anything does. If you've never used one and feel guilty about it, don't. Nothing in the definition at the top of this page needs a list, and the counting exercise works just as well on a finished draft.

An everyday scenario: the chapter with two beats fighting

Say you have a contemporary mystery in its second draft, and chapter nineteen is your favorite chapter. Three beta readers have now told you the book loses them around there, and none of them can say why, though one said "I had to go back".

You mark the beats in the margin, one note per change, and there are two of them. Cleo Marchetti, your investigator, tells her boss she's dropping the case. That's a real beat, because she has been driving the book for eighteen chapters and she has just stopped. Four pages later, in the same office, her sister calls to say their mother has been taken to the hospital. That's a real beat too, and the last third of the book runs on it.

Both beats are good, and together they cancel out. The resignation never gets felt, because the reader gets ninety seconds with it before something bigger arrives. The hospital call never gets felt either, because the reader is still taking in the resignation and is also, fairly enough, wondering whether Cleo is about to take it back. Reading your own notes, you see that you've re-explained the resignation twice in later chapters, and that's the giveaway. If you have to keep reminding the reader about a beat, it didn't land.

You have three options, and they're not equal. You could turn the resignation down and make it a threat instead of a fact. That's cheap, and it costs you the strongest character beat in the act. You could move the hospital call to chapter twenty-one, after Cleo has spent two chapters finding out what it's like to be off the case. Now the call lands on a woman with nothing to do and no excuse, and it means something quite different from what it meant in chapter nineteen. Or you could move the resignation earlier, to chapter sixteen, and let chapter nineteen be the hospital call alone. This version changes the least prose, and you should weigh what it costs before you choose.

You take the second option. You move about eight hundred words and write one new paragraph to bridge the gap. Chapter nineteen is now shorter and hits harder, and chapter twenty-one, which used to be the weakest in the book, now has the call in it. Two readers who reread it say the middle is fixed. Nothing was rewritten. You moved one beat. That's the whole case for this element, at the scale you'll actually use it.

Try it yourself

Here's an audit of your own manuscript that takes an evening. You need a printout or a PDF you can comment on, and you need to be willing to write in the margin instead of in the text.

  1. Pick your zoom level and write it down. Before you count anything, decide what size of change counts as one beat, and note that at the top of the page. A default that works for a novel: a beat is a change you could state in one short sentence with a verb in it, after which a character or the reader knows or has done something new. Sticking to one zoom level matters far more than which one you pick.
  2. Mark one chapter. Take a chapter you feel sure about and go through it, making a margin note at every change. Each note is three to eight words, verb first: refuses the loan, sees the second car, decides not to tell him. Don't tidy the prose while you're in there. You're only measuring on this pass.
  3. Count, and read the notes on their own. Copy the margin notes into a list, without the chapter, and read the list. If the list is a story, the chapter works. If it reads as one thing said four ways, you've found a stall. If it has one item on it, you've found a scene that starts too early.
  4. Do the same for eight chapters across the book. Take two from the opening, four from the middle and two from the last quarter, and write the counts in a column. You're not looking for a target number, just at the shape of the column. If your middle chapters drop to one beat or none while the opening runs at five or six, that tells you what's wrong, and it's the most common finding there is.
  5. Find your emptiest chapter and decide about it. You have three choices, and you should make one on purpose: give it a beat, summarize it, or keep it as breathing room. Write down which one and why, in one line, so you don't quietly leave it as it is.
  6. Look for two beats fighting. Scan your notes for a chapter with two big changes close together. Ask whether you've re-explained either one later in the book, and search the manuscript for the second mention to check. If you have, separate them, and weigh both directions before you choose which one moves.
  7. Run the swap exercise on one scene. Take a scene that feels lifeless and rewrite it, changing one beat only: one different thing learned, decided or refused. Keep the rest of the prose, then read both versions. No exercise gives you a faster feel for how much weight a single beat carries, and it takes twenty minutes.
  8. Do the filler-beat search. Search for nodded, sighed, shrugged, took a deep breath and looked away, and note the totals. Cut half, then replace a handful of the ones left with something specific to that character in that room, and leave the rest as bare dialogue. Then read one of those conversations out loud.
  9. Ask which beat anyone means, once. The next time a reader, editor or critique partner tells you a scene needs more beats, ask what size of thing they have in mind. One sentence, and now and then it'll save you a week.
  10. Keep the list. Typed up, the margin notes become a beat list for your book, and it fits on two or three pages. For the rest of the draft, you can check a wandering chapter against it. It's also a far better planning document for the next book than anything you could write before you start. See Outline.

Here's a template for the audit, one line per chapter:

Chapter __, beats: __ Beat one: __________ (verb first) Beat two: __________ Beat three: __________ Largest change in the chapter: __________ Re-explained later? yes / no Verdict: works / stalls / starts late / two beats fighting / empty

Common mistakes and misunderstandings

Four meanings, one word, and nobody checks. This is the costliest mistake on this page. A note saying a scene needs more beats can mean a reversal, a tactic, a gesture or a structural landmark the size of three chapters, and the four fixes have nothing in common. Ask which one, because everyone in the conversation has been guessing.

A beat count treated as a quality target. There's no correct number of beats per scene, because the number depends entirely on the zoom level the counter chose. Nobody can enforce "three beats per scene", and the advice falls apart the moment it meets a good novel. Use counts to compare chapters within one manuscript, counted the same way. Don't use them to grade a chapter against a rule.

A beat sheet's items mistaken for beats. The named steps of a commercial method are structural landmarks, and each one needs several actual beats to deliver. If you write one scene per named beat, your book gallops through fifteen things and develops none of them. The beat sheet page has the document. The individual landmarks, like the inciting incident, first plot point, rising action and midpoint, have their own pages, and each takes more than a beat to do.

The belief that a beat has to be visible. In prose it doesn't. A character silently working out that her colleague is lying is a beat, and often the biggest one in the chapter. A beat needs a witness, and in fiction the reader counts as one. Here's one of the main differences between the prose beat and the screen beat, so keep it in mind before you bring screenwriting advice over whole.

Confusing the beat with the value that turns across a scene. They're related, and they work at different sizes. When a whole scene moves its charge from one end of a pair to the other, that's the subject of Value Shift, measured at the scale of the scene and above. A beat is smaller and more local, and a scene will hold several beats that each move very little. Checking at the wrong scale is why some flat scenes won't come right: the value is turning, and nothing inside the scene is happening.

Events counted as beats. This is the most common way counts get inflated. A car chase isn't a beat, just a place where some beats can happen. The question is never how much activity is on the page, only what's different at the end of it. The busiest chapters are so often the emptiest for that reason. Plot makes the same point at the scale of the whole book.

Beat notes phrased as topics. "The conversation about the money" tells you nothing, and it'll let a dead scene pass an audit. Write the note as a change with a verb in it and the audit starts working. A topic can cover four pages of nothing. She refuses to lend it can't.

Filler beats where punctuation was wanted. Sighing, nodding, shrugging and sipping between every pair of lines, where each one takes a split second of the reader's attention and gives nothing back. Cut most of them, and make the ones you keep specific. Dialogue covers pauses and interruption more widely.

Planning a book beat by beat before you know what it is. The beat is a poor tool for coming up with a story, because a slot marked reversal won't tell you what reverses, and a plan made only of slots produces the same interchangeable fiction the method's critics describe. Plan in decisions and consequences. Audit in beats.

The critique of beat sheets applied to beats. Peter Suderman's argument, and others like it, are about a commercial template with page numbers attached. They say nothing against the idea that something in a scene ought to change. You can reasonably reject the template. If you reject change, you're describing a manuscript nobody will finish.

Arguing about the count instead of agreeing on the zoom. If you say four and your editor says six, you're almost certainly both reading well and measuring differently. Settle the zoom level first, which takes one exchange, and the disagreement usually goes away.

The audit postponed until the book is finished. A beat pass at the fifty-thousand-word mark costs an evening, and it changes what you write next. The same pass on a finished draft finds the same faults and costs weeks, because by then the chapters exist and you're fond of them. Run a cheap one early. Revision is about the expensive version.

Related elements