Your words are not your running time
A 75,000-word novel does not become a 75,000-word film. The film keeps a fraction of the novel's words as material that actually reaches the screen — screened words — and that screened material is then spoken at a spoken rate. The two numbers that matter are the keep ratio and the spoken rate, not the page count, the chapter count, or the thickness of the spine. Word count sets the starting point; keep ratio and spoken rate set the clock.
A disciplined feature keeps about 10 to 15 percent of source words. The tool uses 12 percent as the mid. That does not mean 12 percent of scenes survive; it means the words that survive — dialogue plus the screened fraction of description that must be shown — are about 12 percent of the source words. A 75,000-word manuscript at 12 percent is 9,000 screened words before genre pace. A 120,000-word epic at 12 percent is 14,400. The screened words are the material the spoken rate divides.
Spoken rate is the second number. Theatrical dialogue is typically delivered at about 130 to 140 spoken words per minute, and the tool uses 135 as the mid. Screened words divided by 135 is minutes. At 9,000 screened words the raw is 66 minutes before genre pace. At 14,400 it is 106. Those two raws are the same manuscript at the same keep ratio, and they land in different tiers. Word count alone moves the clock.
Genre pace moves the same words across tiers
Genre changes how much of the description must be shown rather than said. Thrillers are propulsive: dialogue carries the plot and description is thin, so the same source words produce more screened words per source word than a literary novel that carries interiority. The tool models this as a pace factor from 1.10 for literary to 1.35 for thriller, multiplied on screened words before dividing by spoken rate.
At 75,000 words the math shows the move. Thriller at 1.35: screened '75000 × 0.12 × 1.35' = 12,150, runtime '12150 ÷ 135' = 90 minutes. Fantasy at 1.15: screened '75000 × 0.12 × 1.15' = 10,350, runtime 76 minutes. Literary at 1.10: screened 9,900, runtime 73 minutes. The same 75,000 words can be a 90-minute feature or a 73-minute long short depending on what genre's description must survive. The genre you pick is the compression you expect.
The table in the output keeps the same word count across genres so you can see the range directly. At 120,000 words the same three genres produce 122 for fantasy at 1.15 and 90 for thriller as a 75k comparison, but the 120k thriller at 1.35 is '120000 × 0.12 × 1.35 = 19440' divided by 135 = 144, long feature, and the 120k literary at 1.10 is 117, also long feature. The feature window is 75 to 110, so both 120k cases exceed it even with the fast genres. The 120k book will not fit a single tight feature without discarding a major arc; the tier tells you that before the outline is written.
The tiers are the note you pitch, not just the minutes
Minutes alone are hard to pitch. Tiers are easy.
Short film: 25 to 45 minutes. Covers a single arc or 20 to 30 percent of the book. Best as a proof of concept, a casting sample, or a festival calling card. A 30,000-word memoir at 12 percent and memoir pace 1.20 is '30000 × 0.12 × 1.20 = 4320' divided by 135 = 32 minutes. The tier is short. The note is to select one chapter that demonstrates tone and casting, not to summarize the whole book. The checklist that follows a short estimate should list the one arc kept and the two that are not.
Long short / TV pilot: 45 to 75 minutes. Mid length that can be a festival featurette or a TV pilot that must compress the middle. A 60,000-word thriller at 1.35 is '60000 × 0.12 × 1.35 = 9720' divided by 135 = 72 minutes, long short. The note is that the middle must compress: two middle chapters should be kept as one sequence, and one subplot should be chosen, not two.
Feature: 75 to 110 minutes. The marketable single-film window that most 70 to 95k novels land in with a tight cut. A 75k thriller at 90 is feature; an 85k romance at 1.25 is '85000 × 0.12 × 1.25 = 12750' divided by 135 = 94, also feature. The note is that a tighter cut that lands inside this tier sells better than a looser one that runs to 125. The recommendation for a feature estimate should name the one chapter that can be cut entirely and the one scene that must be expanded to carry the act turn.
Long feature / limited: 110 to 160 minutes. Epics and manuscripts above 110 to 120k land here. A 120k fantasy at 1.15 is 122, long feature; a 150k literary at 1.10 is '150000 × 0.12 × 1.10 = 19800' divided by 135 = 146, long feature. The note is to plan an intermission or a two-act cut that uses the book's midpoint as the break. Audition the cut with that break; the single continuous cut that runs 145 minutes will feel long in the room.
Full uncut / multi-part: above 160 minutes. Uncut would be over two and a half hours even with tight compression. A 200k fantasy at 1.15 is '200000 × 0.12 × 1.15 = 27600' divided by 135 = 204, uncut. The note is to split into two films or a 6 to 8 episode limited. Quote runtime per part, not total; the total is the number that scares the buyer, the per-part is the number that sells the season.
The verdict the tool returns mirrors these tiers. Below 45 with a small word count is warning — short by design. Between 110 and 160 is warning — long feature that can work but needs an intermission strategy. Above 160 is fail — uncut would discard too much for a single film, so the recommendation is to split rather than cut further.
Where the minutes come from, step by step
The breakdown lists six steps so the estimate can be read, not just accepted.
Source words is your input. Keep ratio is the 12 percent mid. Pace factor is the genre value from 1.10 to 1.35. Screened words is the product 'words × 0.12 × pace'. Spoken rate is 135 words per minute. Runtime is 'screened ÷ 135'.
A worked walk makes the product visible. A 92,000-word mystery at mystery pace 1.30: keep screened '92000 × 0.12 × 1.30' = 14,352, runtime '14352 ÷ 135' = 106. That is feature, just under the 110 line. The same 92,000 at literary 1.10 would be '92000 × 0.12 × 1.10' = 12,144 divided by 135 = 89, also feature but shorter. The same words at thriller 1.35 would be '92000 × 0.12 × 1.35' = 14,904 divided by 135 = 110, right on the long feature line. The page count 'ceil(92000 / 275) = 335' is irrelevant to the runtime; the page count is a typesetting fact, the runtime is a compression times pacing fact.
Why the tool does not estimate from pages
A 335-page book and a 273-page book can have the same runtime if one keeps less. Pages are words divided by words per page, which is a typography and trim question. Runtime is screened words divided by spoken rate, which is an adaptation question. The tool reports pages at 275 words per page for reference and keeps the calculation on screened words so the two questions do not blur.
The same confusion appears in the opposite direction: a short book can be a long film if the keep ratio is loose, and a long book can be a short film if the adaptation selects one arc. The tool fixes keep at 12 percent so the genre pace is the variable that remains. In practice the keep can move from 10 to 15, but the tool does not expose it as an input because a move from 12 to 15 is a note about which additional subplot survived, not a number the author can pick before the outline exists.
A word count without genre is a range, not a time
A 75,000-word count without genre is 73 to 90 minutes at 135 wpm, a 17-minute range that spans the boundary between long short and feature. A 120,000-word count without genre is 102 to 144 minutes at 135 wpm, a 42-minute range that spans feature through uncut. The genre narrows the range to a point: thriller at the fast end, literary at the slow end. The time you pitch is the point, not the range; the genre is the choice that commits to the point.
What the increase metric would be if there were a plus
Unlike the large-print calculator that reports plus pages and the KDP calculator that reports print cost per page, this tool does not report an increase over a standard. There is no single standard film from which a book deviates; the standard is the tier itself. The screened words and the tier are the pluses. The 12,150 screened words of a 75k thriller at 90 minutes are the increase over a short, and the 27,600 of a 200k fantasy at 204 minutes are the increase over a feature. The minutes are the plus, not a delta.
Reading the genre table with the same words
The table keeps the same word count across six genres so you can see the same manuscript at different paces. A 75,000-word thriller at 90, a romance at 84, a mystery at 88, a fantasy at 76, and a literary at 73 are the same words at 1.35, 1.25, 1.30, 1.15, and 1.10. The five are within 17 minutes of each other. The same words at 120,000 span 102 to 144, a 42-minute range that makes the cut visibly harder for the longer manuscript. The table is not a menu of genres to choose from to land a shorter runtime; it is a map of the compression you must budget before the outline is written.
From minutes to a pitch line
The minutes alone do not sell the adaptation. The tier and the cut do. A development note expects a line like '90-minute thriller feature, 75k → 12% keep → 12,150 screened → 90 at 135 wpm, tier feature (75–110), scope: cut chapter 7 entirely and expand the confrontation in chapter 14 to carry the act two turn'. That line quotes the tier, the rounded minutes, the arithmetic that produced them, and the single cut that gets there. The same line for a long feature would be '122-minute fantasy long feature, 120k → 12% × 1.15 → 16,560 screened → 122, tier long feature (110–160), scope: split at chapter 18 midpoint for intermission'. The split is the planning that corresponds to the tier; the minutes without the cut or split are a number without a note.
Common mistakes and how to avoid them
Estimating from pages is the first. A 273-page book with 75k words at 275 words per page is not 273 minutes; it is the manuscript's typeset length, not its screened length. The same mistake in the opposite direction is estimating from reading time. A 75k manuscript that reads at 250 words per minute in 5 hours is a reading time, not a theatrical rate at 135 wpm with a keep. Reading time is words divided by reading speed; theatrical time is screened words divided by spoken speed. The two divides are on different numerators.
Choosing genre to shorten the estimate is the second. A thriller at 1.35 does not make the same fantasy shorter; it makes a thriller with the same word count that happens to be called fantasy. The pace factor is a note about how the book's description behaves on screen, not a setting to dial.
Carrying the keep ratio as a variable is the third. A keep of 15 instead of 12 makes a 75k book 11 minutes longer at thriller pace; a keep of 10 makes it 15 minutes shorter. Both are within the tier window, and the tier implies the scope that fills the keep, not the reverse. Fix keep at the mid and let the cut name the subplot that kept or was discarded.
Quoting total for a multi-part without per-part is the fourth. A 204-minute uncut at 200k that is pitched as a six-episode limited of 34 minutes each is the same material as a 204 total, but the per-part is the number the buyer can schedule. The tool reports the total and the recommendations say to quote per part when the verdict is fail.
Final checks before you commit and upload
Run the tool twice with two genres that bookend your manuscript — for example, thriller and literary for a thriller that has lyrical passages, or business and memoir for a business book with a strong personal arc. The estimate should bracket the true runtime within a 17-minute window for 75k and a 42-minute window for 120k. If the bracket crosses a tier boundary, the single minute to pitch is the tier you prefer, and the cut or split you promise should be the one that narrows the bracket.
Read the breakdown with the genre note. 'Thriller: fast, lean dialogue and short scenes compress well' is not a compliment; it is a reminder that a thriller cut that keeps the same amount of description as a fantasy cut will run thinner. 'Literary: interiority and prose texture compress least' is not a complaint; it is a reminder that the adaptation that keeps interiority must choose which interior scenes survive.
Compare the tier to the cover. A short film cover that sells a short tier should not promise the full book; a multi-part cover that sells an uncut total should not hide the per-part. The cover, the tier, and the minutes are three views of the same book, and the tier is the one the audience uses to decide whether the object fits the shelf.
Run the tool once with the true word count and once with the word count plus front and back matter removed — typically 800 to 1,200 words for fiction. The screened words that carry front matter are zero; front matter does not screen. The front-matter trimmed count is the one whose minutes matter.
A note on spoken rate — why 135
Spoken rate is not reading speed. Reading speed for an adult is 200 to 300 words per minute eyes to brain; spoken rate for film is 130 to 140 words per minute mouth to ear, because the actor must breathe and the camera must cut. A novel that reads at 250 words per minute for the reader does not speak at 250 for the actor. The tool uses 135 as the mid so the minutes are the minutes the audience sits, not the minutes the reader turns. A faster spoken rate — 145 for rapid screwball dialogue — would make the same screened words shorter; a slower rate — 125 for stately period drama — would make the same screened words longer. The 10 wpm range is about 7 percent of the total; the genre pace range from 1.10 to 1.35 is about 23 percent, so word count and genre dominate over spoken rate within small moves.
Two masters again — film and book are siblings, not zoom
As with large print, the large-print interior and the standard interior are two masters. The film and the book are also two masters. The words per page inside the book and the minutes per minute inside the film are sibling values derived from the same source manuscript by different operations — one by typesetting at a leading, one by keep and spoken rate. The book's 273 pages and the film's 90 minutes are siblings, not parent and child. The tool reports both so the author can see the book as a book and the film as a film from the same words.
How long the pass really takes to decide
The runtime question is often answered by an executive in a room with a single word count. The tool answers it with three numbers they can repeat: genre pace at 1.35 or 1.10, keep at 12 percent, and spoken at 135. The three numbers are the note that explains why 75k thriller is 90 and 75k literary is 73, why 120k fantasy is 122, and why 30k memoir is 32. The three numbers are also the budget that tells the writer which chapter to cut and where to split. That is the estimate that survives the pitch, not the raw word count the room started with.
Checklist before you quote the minutes
The checklist that follows the runtime is the verify that prevents the minutes from being an orphan. First, name the genre you used to pitch. Second, quote the screened words that carry that genre, not just the minutes. Third, quote the tier the minutes land in and the tier hint that the hint carries. Fourth, name the single cut or split that gets the tier — the chapter cut or the episode break. Fifth, compare the tier to the cover the book already has. Sixth, run the tool again with the genre at the opposite end of the pace band to bracket the true runtime and confirm the single minute to pitch is the tier you prefer. Seventh, subtract front and back matter and confirm the word count you quoted is the text that screens.
Engineering note — the table is the map
The table that shows the same word count across genres is the map of the compression the outline must budget. It is not a menu of genres to choose from to land a shorter runtime. The same manuscript does not become a thriller by being quoted as a thriller; it is a thriller by its description's behavior on screen, and the pace factor at 1.35 or 1.10 records that behavior. The table's five rows for 75k and for 120k are the range the writer must plan within.
Five traps that make the minutes wrong
Estimating from pages is trap one. Estimating from reading time is trap two. Choosing genre to shorten the estimate is trap three. Carrying the keep ratio as a variable the room can dial is trap four — a keep of 15 instead of 12 makes a 75k book 11 minutes longer at thriller pace, a keep of 10 makes it 15 minutes shorter. Both are within the tier window. Fix keep at the mid and let the cut name the subplot that kept or was discarded. Quoting total for a multi-part without per-part is trap five — a 204-minute total pitched as a six-episode limited of 34 minutes each is the same material as a 204 total, but the per-part is the number the buyer can schedule.
Why the verdict is pass, warning, or fail
The verdict pass means the estimate fits the chosen form without a note. Warning means the estimate can work but needs an intermission or a named cut that the room must approve. Fail means uncut would discard too much for a single film, so the recommendation is to split rather than cut further. The verdict is the note the room can action, not just the minutes the spreadsheet produced.
Final pass — run the tool on front-matter-trimmed count
Front matter does not screen. Subtract the 800 to 1,200 words of title page, copyright, dedication, and table of contents, and confirm the word count you quoted is the text that screens. The front-matter trimmed count is the one whose minutes matter, not the file's total that includes front matter.