ShelfCrunch
A brainstorming partner for the marketing copy your novel actually deserves.
Every quote verified against your text. Every comp checked against live Amazon data. Nothing invented and passed off as research.
A complete Full Pack for The Great Gatsby, published unedited. No email, no signup.
Upload your manuscript to get the logline, blurbs, taglines, comps, social posts and category strategy you’d normally spend two weeks scraping together — or get a free logline for your own book.
Your manuscript stays private — deleted from our systems after delivery, never used for training.
The productivity chasm
Before
You finished the book. Now you need a logline, two blurbs, five taglines, three comp titles, a category strategy, a keyword set, a launch newsletter sequence, and a BookTok script. None of which is the thing you trained your craft to do.
With ShelfCrunch
One upload, one pack. Concrete copy, grounded in your manuscript, ready to paste into Amazon, your newsletter, and your socials. The boring chasm becomes a 30-minute afternoon.
What’s in the pack
Check out a sample →Logline & blurbs
One sentence, then long and short back-cover copy.
Five taglines
Varied tones — literary, punchy, mysterious, emotional, provocative.
30 verified quotes
Every line string-matched against your manuscript. Anything that doesn’t match exactly is discarded, not flagged.
5 comps, validated
Each comparable title is looked up against live Amazon data — review count, price, and sales rank where the provider returns one — and stamped with the date we retrieved it.
Category & keywords
A category picked from Amazon’s real hierarchy, two backups, and five keyword hypotheses labelled as hypotheses.
Reader profile
Who they are, what they’re already reading, the mood they’re in.
Social posts
X post + thread, Instagram caption, Threads post, BookTok script with visual direction.
Newsletter subject lines
Five subject lines, varied: curiosity, urgency, intimacy, humor, declarative.
Elevator pitch
100 words in your voice — for events, DMs, and the moment a stranger asks.
Internal review
Candid editorial feedback for you alone: pacing, voice, positioning risks, and what this book is not. Deliberately not flattering.
External positioning brief
Media-kit material for reviewers and podcast bookers. Descriptive, never evaluative — it is not a review and says so.
One-time price. No subscription.
Pay once per book. Nothing recurring, nothing to cancel.
Teaser
Free
A logline and a tagline from a short questionnaire. No upload.
- · One logline
- · One tagline
- · No manuscript required
Hooks
$23
Refunded automatically if a job fails
The copy you need first, without committing to the full pack.
- · Logline
- · Back-cover blurb — long and short
- · Five taglines, tone-labelled
Full Pack
$47
Refunded automatically if a job fails
Everything you need to launch, with the checkable parts checked.
- · Everything in Hooks
- · 30 verified quotes — string-matched against your manuscript
- · 5 comps, looked up against live Amazon data and date-stamped
- · Category from Amazon’s real hierarchy, plus two backups
- · Five keyword hypotheses, labelled as hypotheses
- · Reader profile, social posts, newsletter subjects, elevator pitch
- · Internal review — candid editorial feedback
- · External positioning brief — media-kit material
The Full Pack sits above comparable report tools because of the verification, not the word count: the quotes are matched against your text and the comps are looked up against Amazon, so you’re not the one checking them.
Accuracy is a feature.
Quotes are matched, not generated
Every one of the 30 quotes appears in your manuscript exactly as printed. We string-match each candidate against your text and discard anything that doesn’t match — discarded, not flagged, because a flagged quote is one you have to check.
Comps are looked up, not guessed
Each comparable title is looked up against live Amazon data — review count, price, and sales rank where the provider returns one — and stamped with the date we retrieved it. Titles we can’t confirm at all are dropped and replaced, and a comp with a missing field says so rather than leaving you to assume.
We tell you what we don’t know
Keyword suggestions are labelled as hypotheses to test in Amazon Ads — we have no search-volume data and won’t pretend otherwise. If you need that, buy a keyword tool. Category data carries the date it was pulled, because Amazon retires categories without notice.
Before you buy
›How long can my manuscript be?
Between 5,000 and 250,000 words, in a file up to 25 MB. We count the words the moment you choose a file and show you the number before anything is paid for. Below the minimum there isn’t enough text to find comps, quotes or a category without guessing, so we refuse it rather than sell you a thin pack. Over the maximum, get in touch and we’ll do it by hand.
›What file formats do you accept?
DOCX, EPUB, and PDFs that have a real text layer. A scanned or image-only PDF is rejected at upload with an explanation — we don’t OCR it, because OCR errors would end up inside quotes we promise are verbatim.
›Does this work for books not written in English?
Not yet, and we’d rather say so than take your money. The pipeline is built and tested against English-language manuscripts, and the Amazon category tree and comp lookups it uses are the US marketplace’s. A manuscript in another language will produce output, but we can’t stand behind it the way we stand behind the rest of this page — so please don’t buy a pack for one.
›How long does it take?
Eight to ten minutes for a full pack, because we read the whole book rather than skimming it. You can close the tab — the link is emailed to you when it lands, and the free teaser takes about thirty seconds.
›What happens to my manuscript?
It's deleted from our systems after we deliver your pack, unless you tick the box asking us to keep it. We never train on your text. If you abandon a checkout, the copy we were holding is purged within two hours.
›What if the pack is bad, or the job fails?
A failed job is refunded in full automatically — you don’t have to ask, and you don’t have to notice. If a pack completes and you think it’s wrong, write to us: the things we promise are checkable (quotes match your text, comps exist on Amazon), so a disagreement about them is one we can settle rather than argue about.
Privacy is a feature.
Deleted after delivery
Your manuscript is purged from our systems once the pack is generated. No silent backups. No training. If a provider can't guarantee deletion, we won't send your manuscript to them. No customer’s book as an example on this site — the sample pack is a public-domain novel for exactly that reason. The one exception is yours to make: you can tick a box at upload to have us hold the text for 15 days so we can act on feedback you send. It is off unless you turn it on.
Your book isn't training data
A pack takes 8–10 minutes instead of seconds because we read your book end to end. We don't run on platforms that train on customer data or keep your text after delivery — that's the trade, and we make it on purpose.
Anonymity by default
No public customer list. We never use your book or your name without your explicit, separate opt-in.
Don’t take our word for it.
Read a complete Full Pack — every blurb, every comp, every verified quote — exactly as the pipeline produced it. Then decide.