Publishing and professional handoff

How to write a book description that sells the read

August 16, 2026 · The StoryProp team

A book description is the shortest thing you will write about your book and the hardest, and the method fits in a single line: one person, one concrete cost, one question you refuse to answer, in the voice of the book. It is not a summarizing problem — which is why most descriptions read like a table of contents with adjectives. It is a seduction problem, and it has about a hundred and fifty words to do what the manuscript does in ninety thousand: make a stranger want the next page. The size of that job shows up in a published number. BookBub, which emails discounted titles to genre-sized reader lists, states on its Featured Deals pricing page that roughly 20 to 30 percent of the clicks a deal generates end in a purchase — and those readers had already seen the cover, the price and the category and clicked anyway. The description is most of what happens between the click and the decision.

The reader who encounters your description is not deciding whether your book is good. They cannot know that yet. They are deciding whether the promise you are making is a promise they want kept. Everything in the blurb serves that single decision, and anything that does not serve it — a second subplot, a third named character, the twist you are proud of — is taking up room that the promise needs. On a new title's product page there is usually nothing else to go on. Amazon's customer-review help page states that only customers who have spent at least $50 on Amazon in the last twelve months can submit ratings and reviews, and its community guidelines bar household members and close friends of the author from reviewing the book at all — so the pool of people permitted to review a new book excludes both the light Amazon shopper and everyone in the author's own house. Cover, title, description. That is the entire apparatus.

That apparatus also goes to work long before the book does. Amazon's KDP help pages let a Kindle title open for pre-order up to a year ahead of release, and state that pre-orders contribute toward sales rank and Kindle Store merchandising before the book is released — so for as long as twelve months, the description is the book. The same paragraph then does the same job on the Apple Books and Kobo listings, in the metadata IngramSpark pushes to the trade, in your newsletter, and on the back cover of the paperback. It is also the only part of the package that costs nothing but attention to make. Every other line item on a product page — the edit, the cover, the typesetting — is bought from somebody. The description is the one you write yourself, and the one you can rewrite on a Tuesday afternoon at no cost at all.

Blurb, synopsis, pitch: three different animals

Writers conflate these three constantly, and the conflation is the single most common cause of bad back cover copy. A synopsis is the whole plot, including the ending, written for someone whose job requires knowing how it turns out — an agent, an editor, a contest reader who asked for one. It is complete, chronological, and unglamorous. Nobody has ever been moved by a synopsis, and nobody is supposed to be. A pitch is a different thing again: a compressed comparison, spoken or written to a professional who reads a great many books, answering what kind of book this is, who it is for, and where it sits. It answers a business question, and it can be a single sentence — a logline is the screen trade's version of the same compression.

A blurb — back cover copy, the description on an Amazon or Apple Books product page, the paragraph in a newsletter — is a trailer. It is written for a reader with no obligation to finish it, and its only job is to create appetite. It ends before the interesting part. If you have written something that would spoil the book for someone who read it and then read the book, you have written a synopsis and mislabeled it.

The practical consequence: never write your description by cutting down your synopsis. Cutting a synopsis produces a shorter synopsis. Start over from the question a trailer asks, which is not what happens in this book but what would make someone need to know.

The first line carries more weight than the rest combined

Write the first line as though it is the only line you can count on being read. The reader owes your description nothing, and the second sentence is reached only by someone the first sentence held. So the first line is not a warm-up. It is the entire audition, and the remaining hundred and forty words are the callback for people who liked it.

This means the first line cannot be setup. It cannot be “In the small coastal town of Ashgrove, life has always moved slowly.” That is a sentence about weather. It is doing throat-clearing work that the second paragraph could do, if it needed doing at all.

What the first line can do is state an unstable situation. “Mara has not slept in the house since her mother died, but tonight the buyer is coming at eight and the kitchen still smells like her.” Now something is wrong, something is scheduled, and someone is in trouble. A reader can feel the shape of a story from that alone, and the shape is what they are buying.

A useful test: cover everything but your first sentence and ask a friend what the book is about. If they can guess the genre and sense a problem, the line works. If they say “it sounds atmospheric,” it does not.

Hook, stakes, and the turn

Under the first line, descriptions that work run the same three moves — hook, stakes, turn — and the structure is invisible when it is done well.

The hook establishes the character and the disruption — who this is and what has just gone wrong. Keep it to one person if you possibly can. Ensemble novels tempt you into naming four people, and four names in a hundred and fifty words means the reader remembers none of them. Pick the one whose trouble is most legible and let the others exist as roles: her brother, the buyer, the man who says he knew her mother.

The stakes name what it costs to fail. This is where descriptions most often go abstract — “she must confront her past,” “nothing will ever be the same.” Those phrases are placeholders that survived every draft. What is the concrete thing that is lost? The house. The custody. The forty years of silence between two sisters that will become permanent if this conversation goes wrong tonight. Concrete stakes are also genre stakes: what a thriller risks and what a family novel risks are different, and naming yours tells the reader which shelf they are standing at.

The turn is the last beat, and it should open rather than close. One sentence that introduces a complication the reader cannot resolve from the information given. “The buyer knows the layout of the house, and Mara has never met him.” Stop there. The temptation to explain is enormous and you must not.

Diction is genre signalling

Readers decide what kind of book this is from your word choices long before they parse the plot, and the signals are more specific than most writers realize. Short declarative sentences with hard consonants and a clock in them read as thriller. Sentence rhythms that lean on subordinate clauses, sensory detail, and the past perfect read as literary. Wry understatement and a domestic noun in an unexpected place read as comic. Capitalized proper nouns for systems and places — the Ninth Accord, the Hollow Meridian — read as fantasy, and readers who do not want fantasy will leave at the first capital.

This is why a description written in a voice that does not match the book is worse than a dull one. It brings the wrong readers, who then feel misled, and it repels the right ones. Write the blurb in the register of the book itself. If the novel is spare, the blurb is spare. If the novel is funny, one line of the blurb should actually be funny — not described as funny.

A reliable move: lift the diction, never the sentences. Do not paste a paragraph from your opening pages into the description. Out of context, prose written for a reader who is already inside the story tends to read as vague. Instead, notice which words your book keeps using and let two or three of them into the blurb. The chipped cup can be in the description. The paragraph the chipped cup came from should not.

What to withhold

Withholding is the craft. Here is a working list of what stays out. The ending, obviously, but also the mechanism of the ending — do not tell them there is a letter, if the letter is what resolves it. The twist, including hints artful enough to function as hints; readers are extremely good at decoding “but nothing is as it seems.” The second half generally: most descriptions should not travel past the point where the first act closes. Subplots entirely. Any character who does not appear in the first quarter. Backstory, which is the single largest category of blurb waste, because backstory explains why the situation is interesting instead of letting the situation be interesting.

What stays in is smaller than you expect: one person, one disruption, one cost, one unanswered question, in the voice of the book. That is genuinely all, and it holds whether the book is a debut, the fourth entry in a series, or a long novel with a cast of twenty. When a description feels thin, the problem is almost never that it needs more information. It is that the four elements are not specific enough, and adding a fifth element hides the vagueness rather than fixing it.

Write three versions and leave them a day. Then read only the first lines, back to back. One of them will be doing more work than the other two, and it usually is not the one you labored over. Build the winner out to length and let the others go. Titles are a separate problem with a separate logic and deserve their own sitting. And when the description finally locks — one person, one cost, one open question, in the book's own voice — it tells you something good about the manuscript underneath it: a promise you can state in a single unstable sentence is usually a promise the opening pages are already keeping. A hundred and fifty words is a small thing to get right, and it is entirely within reach.

Blurb work goes faster when the details are already written down — whose story it is, what the cost of failure actually is, which words the book keeps returning to. StoryProp keeps those decisions as project records your drafting partner reads before it writes, so you can select a description draft, say make the first line do more, and get that line redrafted while the rest stays yours.

← All posts