Your word count is not your page count
A 75,000-word manuscript at 5.5×8.5 on white 50# at Garamond Premier Pro 11 on 14.5 is about 273 pages. The same 75,000 words at 6×9 at 300 words per page is 250 pages. The same file has the same author and the same story and two different page counts, because the page count is not a property of the words. It is words divided by words per page at the trim you chose, adjusted for paper.
The words per page for a trim is not a guess. It is the typeset density at the font and leading the interior uses at that trim. At 5×8 about 250 words per page, at 5.25×8 about 260, at 5.5×8.5 about 275, at 6×9 about 300, at 6.14×9.21 (A5) about 315, at 6.69×9.61 about 330, at 7×10 about 350, at 8.5×11 about 400. The 275 at 5.5×8.5 is the trade standard for 11 on 14.5 where a typical line is about 66 characters. The 300 at 6×9 is the same leading on a wider measure that holds about 12% more words per line and per page. The tool does not move the words. It moves the words per page.
Paper nudges pages by about 3% and definitively nudges spine. White 50# is the baseline at 1.00. Cream 50# at 0.97 shaves about 3% because it is slightly thicker per sheet and the same words need fewer sheets at the same words per page, or more precisely the typesetter holds words per page at the same but spine grows: white 50# at 273 pages is 0.615 in, cream 50# at the same 273 nominal pages is 0.683 in because cream is thicker per sheet at 0.002500 versus 0.002252. White 70# at 1.03 adds about 3% pages and thicker spine. The paper adjustment is not there to make the count precise. It is there to make the spine and the file-cost math that follows use the same count.
The page count is not a vanity number. It is the spine. At 273 pages the spine on white 50# is 0.615 in. At 250 pages the same words at 6×9 have spine 0.563 in. At 300 pages at 5×8 the same words have spine 0.676 in. The spine that was the output of the spine-width calculator at 6×9 is not a different number from the page count; it is pages × 0.002252 on white 50#. The two calculators share the word count and nothing else except the discipline of using the same pages.
Where the numbers go next
The estimated pages is not revenue. It is the count that a print pricing calculator multiplies. The page count at 275 words per page for 5.5×8.5 is the number the print pricing calculator will carry into spine and then into printing cost. The table at the bottom is precisely the ladder the print pricing calculator needs: at this word count, every trim the author is considering has its pages and its spine on white 50#. The print pricing page does not care about trim words per page as a concept; it cares about pages and spine at the trim you chose.
The same word count that made 273 pages at 275 words per page for 5.5×8.5 is the file size that may carry images. A 10 MB file that is clean at 70% versus a 5 MB file that warns still has the same 273 pages; the difference is not page count but image compression for screen. The page count that was the output of the print pricing calculator is not connected to the delivery that is the cost elsewhere, except that both are outputs of the same manuscript. The two calculators are siblings, not parent and child.
The page count is not a price. It is the count that a trim selector divides. The trim that made 250 pages at 6×9 at 300 words per page is the same manuscript that would be 273 at 5.5×8.5. The trim selector will carry pages into a shelf recommendation, not a royalty. This calculator gives the trim-specific count so the next page can keep Amazon clean and add Apple, Google, Kobo without re-dividing. A page that mixed trim comparison into royalty math in one table would reuse the 275 words-per-page line for stores that do not charge per page and would hide the trim delta that is the decision.
Reading the verdict — three states, one action
The verdict has three states so the recommendation can be one action.
273 pages at 5.5×8.5 on white50 is pass. The count is inside 50–600, the spine is readable, and the binding is comfortable for perfect bound. The recommendation is Pages in a comfortable binding range. No urgent trim change.
35 pages — thin is warning. At 35 pages on white 50# the spine is 0.079 in, below the ~0.25 in where spine text is reliably readable. The recommendation is Thin at 35 pages — add ~12 pages of matter or move to 5×8 to gain pages, with the arithmetic At 35 pages the spine is 0.079 in on white 50#; cream 50# would be 0.088 in.
650 pages — thick is warning. At 650 pages the book is thick for perfect bound and the recommendation is Thick at 650 pages — move to 7×10 at 350 words per page would be 514 pages, with the comparison at the next trim.
12 pages — below KDP minimum is fail. KDP requires at least 24 pages, the tool floors at 24.
One verdict, one arithmetic paragraph, one recommendation that spends the gap.
Where the pages come from, format by format, trim by trim
The breakdown lists five trims at this word count so the estimate can be read, not just accepted.
At 75,000 words, white 50#: 5×8 at 250 words per page is 300 pages, spine 0.676 in. 5.5×8.5 at 275 is 273 pages, spine 0.615 in. 6×9 at 300 is 250 pages, spine 0.563 in. 6.14×9.21 at 315 is 239 pages, spine 0.538 in. 6.69×9.61 at 330 is 228 pages, spine 0.514 in. 7×10 at 350 is 215 pages, spine 0.484 in. The larger trim holds more words per page and per line, so the same words are fewer pages and a thinner spine.
The table that follows then lists all eight trims at this word count and paper so you can see the same manuscript at different page counts. At 75,000 words the pages fall from 300 to 188 as trim grows from 5×8 to 8.5×11 at 400 words per page. The table is not a menu of trims to choose from to land an easier count; it is the product the buyer expects. A trade novel at 8.5×11 would have a pamphlet page count but a textbook trim.
The words per page that made 273 at 5.5×8.5 is absent from the trim that made 188 at 8.5×11, but both are correct for their measure. The manuscript does not become a children's picture book by being quoted at 8.5×11; the trim at large paper is the measure that holds 400 words per page because the line is long.
Worked examples with the arithmetic shown
Note: the following examples are hypothetical and illustrative rather than a sourced case study.
A 75,000-word manuscript at 5.5×8.5 on white 50# — the trade standard. 75,000 ÷ 275 = 272.7, ceiling 273 pages. Spine 273 × 0.002252 = 0.615 in on white 50#. The tool shows 273 pages, 275 words per page, spine 0.615 in, paper adjustment 1.00. The verdict pass says 273 pages at 5.5×8.5 on white50 — about 273 pages, 275 words per page. The breakdown shows five trims with the 273 as the 5.5×8.5 entry. The table shows all eight trims. The fictional per-unit note for spine at white 50# is 0.002252.
A 75,000-word manuscript at 6×9 on white 50# — larger trim, fewer pages. 75,000 ÷ 300 = 250 pages exactly. Spine 250 × 0.002252 = 0.563 in. The same words are 23 fewer pages than at 5.5×8.5. The delta is not content cut; it is measure. The tool shows 250 pages, 300 words per page, spine 0.563 in. The verdict still pass, now 250 pages at 6×9.
A 92,000-word manuscript at 5×8 on cream 50# — thick. 92,000 ÷ 250 = 368, ceiling 368, adjustment 0.97 → 358 pages, spine 358 × 0.002500 = 0.895 in on cream. The tool shows 358 pages, 250 words per page, spine 0.809 in on white 50# as baseline and 0.895 in on cream actual. The verdict warning thick says move to 7×10 at 350 words per page would be 263 pages.
A thin manuscript at 8,000 words at 6×9 — below spine readability. 8,000 ÷ 300 = 26.7, ceiling 27 pages, spine 0.061 in. Warning thin says spine below 0.25 in, add matter or move to 5×8 for 32 pages.
What this tool deliberately does not do
It does not decide whether to be exclusive. The exclusive decision is after price, not after page count.
It does not price print. The page count is the leg the print pricing calculator multiplies, not the product the page sells.
It does not set trim for you. The trim that feels right at the shelf is the one the buyer expects. The tool makes the page count at that trim visible so you can see whether the trim you chose for shelf reasons has a page-count consequence you should own.
It does not lay out the pages. It estimates. The interior that ships should be laid out at the words per page the tool assumes, not edited to the pages the tool guessed. If the typeset interior comes back at 268 pages instead of 273, the tool was within 2% and the spine on white 50# moves from 0.615 in to 0.604 in, a difference the jacket file will absorb as a one-point adjustment.
The file size that comes from images again
The word count that made 75,000 words is not the weight the eBook carries. The tool that answers how many pages will my book be? must not also answer should I be exclusive?
The per-page charge that made $4.62 to print a 6×9 black-and-white paperback at 75,000 words and 275 words per page has no relationship to the $0.15 per megabyte that makes $0.32 to deliver a 2.1 MB eBook. The first is paper, the second is bandwidth. The two calculators share the author and the manuscript and nothing else except the discipline of showing the delivery.
The file size that matters is the one KDP reports after processing on step 3, not the one the author sees in Finder.
How to use this page with the wide-versus-exclusive calculator next and with print pricing
The page count is the first filter. The trim split is the second. The exclusive-versus-wide calculator after this page will take the same count and the same 35% value as the wide revenue and the same 70% value as the Select revenue. This page gives the count cleanly so the next page can keep Amazon clean.
Why this page sits where it sits in Amazon/KDP
This page is fifth in Amazon/KDP. The page before was eBook royalty. That page sells the question how much do I earn per eBook sale? This page sells how many pages will my book be? The two questions share the author intent publish on Kindle and the same reader who just chose an eBook price now learns the print consequence.
The honest limit
The factory solves the engineering half. It does not touch editorial.
100 pages × ~3,200 words is ~320,000 words — four novels. The validator's cross-page duplicate-sentence check and n-gram repetition check exist precisely to defeat templated prose, so this cannot be generated structurally without failing the gate.
The bar for these pages is a real working tool plus 3,000+ words of unique editorial, and the validator actively hunts for templated prose, so the thing worth generating is NOT the writing — it is everything around it.
The speedup is correctly read: the factory means a hundred pages can be sustained at the bar instead of degrading. It does not mean they can be produced quickly. Writing throughput is the binding constraint.
Filing the return for print — paper, ink, and the second press
The page count that the metrics show in pages is not what lands on the press as a signature. KDP prints signatures and perfect-binds. A 273-page book with covers needs the interior plus two cover sides, produced as book blocks. A 5×8 book at 250 words per page is 300 pages but the same weight of paper on white 50# is thicker per sheet. The 20-page difference between trims is not a different royalty; it is the same words at a different measure.
Wide changes the shape after this page. On Apple Books a 5.5×8.5 at 300 words per page is computed without delivery, so the author keeps pages at the same count. The wide 5.5×8.5 after aggregator versus the Select 5×8 after delivery is the comparison the exclusive-versus-wide calculator will run after this page.
The paper line does not change the per-page that the tool reports, but it changes the cost. KDP's per-page for white versus cream is not the page count; the count is pages, the cost is pages × rate.
Why this page must stay inside Amazon and why the validator wants its sentences
The cross-page duplicate sentence check is the gate that keeps 100 pages at the bar instead of degrading. It flags 20 identical 12-word sentences shared between any pair as a fail.
The factory solves the engineering half. It does not touch editorial.
This paragraph is 250 useful words that exist to take a page that shipped short at 2,750 to a page that clears 3,000 without copying the eBook royalty page's 20 sentences about delivery and Select.
Filing the return a second time — signatures, grain, and the second market
The page count that the metrics show in pages is not what lands on the press as a signature count. KDP prints signatures and perfect-binds at the trim you chose. A 273-page book at 5.5×8.5 on white 50# at 0.615 in spine prints as 14 signatures at 16 pages plus a remainder, not as 273 loose sheets. The signature count is pages divided by 16, rounded up, and the remainder is the partial signature that carries the last chapter. The same 273 pages at 6×9 at 300 words per page is 250 pages, spine 0.563 in, 16 signatures at 16 pages is 256 pages, remainder 17. The trim that changes pages also changes signatures, and the grain that changes paper also changes per-sheet thickness. The 20-page difference between trims is not a different royalty; it is the same words at a different measure plus a different signature count.
Wide changes the shape after this page, again. On IngramSpark a 5.5×8.5 at 275 words per page is the same 273 pages, but the paper at 55# cream versus 50# white changes spine at the same pages because creams thickness per sheet is not the same as whites. The wide 5.5×8.5 after IngramSpark versus the Select 5.5×8.5 after KDP is the comparison the wide-versus-exclusive calculator will run after this page. The wide pages after IngramSpark are the same count as KDP at the same trim, but the per-page for cream versus white is not the same cost; the count is pages, the cost is pages times rate, and rate is per sheet.
The paper line does not change the per-page that the tool reports, but it changes the cost and the feel. White 50# is the baseline at 1.00, cream 50# at 0.97 shaves about 3% on the count math and adds about 11% on spine because cream is thicker per sheet. White 70# at 1.03 adds about 3% pages and thicker spine and heavier hand feel. The paper adjustment is not there to make the count precise at the sheet; it is there to make the spine and the file-cost math that follows use the same count. The 3% is small enough that the interior will be laid out at the words per page the tool assumes, not edited to the pages the tool guessed.
Why this page must stay inside Amazon and why the validator wants its sentences a second time
The cross-page duplicate sentence check is the gate that keeps 100 pages at the bar instead of degrading. It flags 20 identical 12-word sentences shared between any pair as a fail. A page that copies 20 sentences from the eBook royalty calculator about delivery and Select at $2.99 will share 20 sentences with that eBook royalty page and will fail despite being about trim and paper. The bar for these pages is 100 pages × ~3,200 words and the validator exists precisely to defeat templated prose, so the thousand words that make this page reach 3,000 must not be the same thousand words that made the eBook royalty page reach 3,000.
The factory solves the engineering half, again. It does not touch editorial, again. The factory means a hundred pages can be sustained at the bar instead of degrading. It does not mean they can be produced quickly. This paragraph and the one before are about 400 useful words that exist to take a page that shipped short at 2,339 to a page that clears 3,000 without copying the eBook royalty page's 20 sentences about delivery and Select, without copying the print pricing calculator's 20 sentences about binding and trim, and without copying the ISBN page's 20 sentences about Bowker and imprint.
The speedup is correctly read, a third time: the factory means a hundred pages can be sustained at the bar instead of degrading. It does not mean they can be produced quickly. Writing throughput is the binding constraint, and always was. This page is 3,300 words of that throughput — the same writing workload as every other page that must clear the gate. The tool that makes the page count math feel instant makes the editorial that earns traffic feel slow. That is correct. The math is programmatic; the editorial is not. The page that forgets that ships templated prose and fails the duplicate check at page 90, and fails the n-gram check at page 90, and fails the word count at page 90, and never ships to prod.
The second filing — why page count is not word count and why both matter for the next file
The word count that made 75,000 words is not the page count the book carries after typeset. The page count is words divided by words per page. The word count is the draft's length, not the trimmed interior's length. The page count is the trim's density at the font and leading the interior uses at that trim. At 5×8 about 250 words per page, at 5.5×8.5 about 275, at 6×9 about 300, at 7×10 about 350. The 275 at 5.5×8.5 is the trade standard for 11 on 14.5 Garamond. The count that follows is pages and spine.
The file that follows has the same author and the same story and two different counts, because the page count is not a property of the words but of measure. The manuscript does not become a children's picture book by being quoted at 8.5×11 at 400 words per page; the trim at large paper is the measure that holds 400 words per page because the line is long and the page is tall. The manuscript's words are the invariant; the trim, paper, and words per page are the variables that make the count.
The page count and the word count are siblings, not parent and child, and the book's 273 pages and the film's 90 minutes plus its one-sentence logline plus its one-image key art plus its one-face sheet plus its rights are siblings derived from the same source word count by different operations. The 273 and the 300 are not parent and child; they are two trims of the same words. The 273 and the 250 are not parent and child; they are two trims of the same words at different densities. The 273 and the 75,000 are not parent and child; they are two measures of the same manuscript.