How to Write a Blind Date With a Book Description (What 5,000 Books Taught Us)
Every blind date with a book lives or dies on its description. The cover is hidden. The title is hidden. The author is hidden. All the reader has is a handful of words on the front, and those words have to do the entire job that a cover, a blurb, and a browse through the first page normally do together.
We have written descriptions for more than 5,000 blind date books over the last two years, covering about 2,000 different titles. We have rewritten the method three times. This post is what we learned, written for anyone who has to do the same thing: a library running a February display, a teacher setting up a classroom shelf, a bookstore trying it for the first time, or a reader who wants to understand why some blind dates work and some land with a thud.
What a blind date with a book description has to do
It has to let a reader say yes or no in about ten seconds, without knowing what the book is.
That means it has to answer three questions, in order. Is this my kind of story? How is it going to make me feel? How does it move? Most descriptions we see, including our own early ones, only try to answer the first question, and they answer it badly.
There is also a fourth job that is easy to forget. The description has to protect the surprise. A clue that lets someone type three words into Google and find the title has given away the whole point.
Version one: copying the categories (and why it didn't work)
When we started, we did what most people do. We looked the book up, grabbed the categories from Goodreads and the retailer listing, and printed those.
Here is a real one from February 2025, for a very popular novel about a grumpy Swedish widower:
Genre Fiction, Humor & Satire, Sweden, Friendship, Movie
Every word is technically accurate. None of it helps. "Genre Fiction" is a shelving label. "Movie" tells you a studio bought the rights. "Sweden" is the only clue with any texture, and it is doing all the work alone.
The problem with categories is that they were built for sorting books, not for choosing between them. "Contemporary," "Literature & Fiction," and "United States" showed up on hundreds of our books. A description made of words that appear on hundreds of other books is not a description. It is a filing system.
Version two: writing the clues ourselves
The fix was to stop copying and start reading. Not the book, most of the time, but everything written about the book: professional reviews, reader reviews, interviews, the complaints and the raves. The jacket copy is marketing. The reviews are where you find out what a book is actually like to read.
Here is one thriller we have stocked over and over, tagged three times across sixteen months. Same book every time.
May 2025:
Psychological Thrillers, Friendship, Chick Lit, Mystery, Thriller & Suspense, Crime, Women's Fiction
April 2026:
Psychological Thriller, Wealth & Power, Gaslighting & Manipulation, Trapped At Sea, Below Deck Energy, Gold Digger Drama, Luxury Mediterranean Yacht
Mood: Suspenseful, Tense. Pace: Propulsive.
September 2026:
Glitzy Yacht Thriller, Toxic Girl Squad Gone Wrong, Eat-The-Rich Suspense, Billionaire Behaving Badly, Secrets & Score-Settling, Gaslighting By The Pool, Opulent Mediterranean Superyacht
Mood: Suspenseful, Escapist. Pace: Dual Timeline.
The first version could describe a hundred other thrillers. The second one tells you there is a yacht, that someone is being manipulated, and that the money is the point. The third one tells you all of that plus the tone: it is glitzy, it is a little mean, and it is fun. A reader who wants a grim procedural now knows to skip it. A reader who wants to watch rich people ruin each other on a boat will grab it immediately.
Notice what else changed between the second and third versions. The pace moved from Propulsive to Dual Timeline, because on a closer read the structure matters more than the speed. And a content warning appeared that the earlier passes had missed entirely. More on that below.
That is the test we use for every clue now: could this phrase appear on a hundred other books? If yes, cut it.
What makes a good subject or theme
Generic labels like "romance" or "thriller" are the starting point, not the description. A good clue is a compound: a trope or tone plus something concrete you could picture.
Some of ours, pulled straight from the shelf:
- Feisty Grandma Crime Ring
- Poison Pie Vigilante Justice
- Museum Curator Uncovers The Truth
- Gender-Flipped Lost Boys
- Corporate Girlboss Meets Grandma's Kitchen
- Sweltering 1980s Laredo
- Smoky '90s Madison Square Garden
Notice that none of them are single words. "Cozy Mystery" is fine as a first clue, because it names the shelf. But "Feisty Grandma Crime Ring" is what makes someone laugh and pick the book up.
We settled on seven clues per book, and the order matters. The first one names the sub-genre so the reader knows what shelf they are on: "Dark Gothic Romantasy," "YA Apocalypse Rom-Com," "Science-Based Murder Mystery." The next five are the hooks, the relationships, the stakes. The last one is always the setting, written with enough atmosphere that the reader can see it: "Rain-Soaked Vampire Beach Town," "Colonial Long Island Farmhouse," "Glittering Regency Port City."
Two rules keep this from going wrong. First, clues describe the premise, not the twist. Everything on the card should be true by the end of the first act. Second, no proper nouns that identify the book. Character names, invented place names, and anything that appears in the title are off limits.
Why mood and pace matter as much as subject
This was the biggest change we made, and it is the one we would push hardest on anyone writing their own.
Two books can share every subject on the list and feel nothing alike. A WWII historical novel can be devastating or hopeful. A small-town mystery can be cozy or unsettling. If the card only lists subjects, the reader is guessing at the thing that matters most, which is how the book is going to make them feel.
So every book gets two moods, a primary and a secondary, from a fixed vocabulary of about thirty words: Cozy, Devastating, Witty, Unsettling, Heartwarming, Thought-Provoking, Swoony, Eerie, and so on. Two, because books are rarely one thing. "Heartwarming and Devastating" is a real combination, and it tells a reader exactly what they are signing up for.
Pace is the other one, and it is the one that prevents the most disappointment. A reader who wanted something propulsive and got something meditative is not going to blame the book. They are going to blame the blind date. So each book gets one pace word: Slow Burn, Fast Paced, Propulsive, Layered, Dual Timeline, Episodic, Meditative, Compulsive, Vignette-Driven. "Dual Timeline" alone has saved a lot of readers who do not enjoy that structure from a book they would otherwise have liked on paper.
If you only add one thing to your own descriptions after reading this, add pace.
Where the information comes from
Most of the metadata on our cards is not written by hand at all. When a book comes in we scan the ISBN, and a lookup through the ISBNdb API fills in the publisher, year, page count, binding, dimensions, and the publisher's synopsis. If you are cataloging more than a few dozen books, an ISBN lookup is the single biggest time saver there is, and it is the reason our cards can show page count and year published on every single book.
The rating comes from Goodreads, along with the number of ratings. We print both, because a 4.2 from 300 people and a 4.2 from 300,000 people are different kinds of information, and readers deserve to know which one they are looking at.
The clues, the moods, and the pace come from the reviews. Professional reviews tell you what the book is trying to do. Reader reviews tell you whether it worked, and, more usefully, what readers who disliked it were expecting instead. That gap between expectation and experience is where most of our mood and pace decisions come from. We also keep our own notes from the books we have read, and those override everything else.
Content warnings: the part people skip and shouldn't
Under our first system, about 3.5% of our books carried a content warning. Under the current one, it is about 22%.
The books did not get darker. The research got better. Jacket copy does not mention the miscarriage in chapter four or the suicide that the whole plot turns on. Reviews do. Once we started reading them properly, the warnings showed up. The yacht thriller above is a good example: we tagged it three times before a sexual assault warning made it onto the card, and it was there in the reviews the whole time. We just were not reading closely enough.
This matters more for a blind date than for a normal book purchase, because the reader cannot screen for themselves. They cannot read the first page or check a review. If you are wrapping the book, you are the screen. We keep a fixed list of warnings and print them on the card in plain words. For romance, we also print a spice rating, because "steamy" means something different to everyone and a number does not.
If you are doing this for a library or a classroom, this is the section to take most seriously. A wrong warning is embarrassing. A missing one is a betrayal of the trust that makes the whole idea work.
Series: when a book belongs in the mix and when it doesn't
Book three of a trilogy is a terrible blind date. The reader has no idea who anyone is, the author assumes they do, and there is no way to warn them without giving the game away.
Our rule is that a series book only goes in if it can be read on its own. We flag those specifically, and the card says so. The reader knows they are picking up something with more behind it, but they also know they will not be lost. Everything else in a series stays out, no matter how good the book is. We have turned away some excellent books for this reason, and every time it was the right call.
The checklist
If you are writing your own, here is everything on one of our cards:
- Seven clues: sub-genre first, setting last, no proper nouns, nothing past the first act
- Two moods, a primary and a secondary
- One pace word
- Rating and number of ratings, from wherever your readers already trust
- Page count and year published
- Content warnings in plain words, and a spice rating for romance
- A series note if it is one, and only if it stands alone
The clues are the part that takes the longest. The moods and the pace are the part that makes the difference.
If you would rather not write them yourself
We wrap books tagged exactly this way, in every genre from romance and thriller to nonfiction and memoir. Browse by genre on the blind date with a book collection, or let the Matchmaker ask you a few questions and find one for you. And if the suspense is too much, there is always the Book Revealer.
Next in this series: how we decide which books qualify in the first place.
Frequently asked questions
How long should a blind date with a book description be?
Short enough to read in ten seconds. Ours are seven short phrases plus two moods and a pace. If it needs a paragraph, it is doing the wrong job.
Can I just use the Goodreads description?
You can, but it will read like every other description. Publisher copy is written to sell the book with the cover next to it. On a wrapped book it has to work alone, and it usually doesn't.
Should I include the author's name?
No. It is the fastest way to give the book away, and readers who recognize the author lose the surprise entirely.
How many clues is too many?
Past seven, readers stop reading. Past ten, they can usually identify the book.
Do I need trigger warnings for a library program?
Yes, more than anywhere else. The reader cannot check for themselves, so the warning is the only protection they have.
