Choosing comp titles

What a comp actually signals

Agents work in a commercial ecosystem. When they pitch your book to editors, they will use comps to establish the kind of reader your book has and the kind of publishing precedent that makes it sellable. Your query comps are a test run. An agent reading "RECENT LITERARY THRILLER by A and RECENT LITERARY THRILLER by B" gets three things immediately: the genre, the market position, and proof that you read in your category. Comps that fail on any of those three don't do their job.

Comps are not a claim that your book is as good as the comp. They are a claim that your book lives near the comp on a shelf. Plenty of writers cringe at naming a bestseller they admire, as if that were boasting. It isn't — so long as the comp is recent and reasonable in scale.

The three rules

There are three working constraints on a good comp:

1. Recent. Published in the last three to five years. The market shifts. A comp to a 2002 breakout tells an agent you read the bestseller list fifteen years ago. A comp to a book from last spring tells an agent you read the bestseller list last spring.
2. Real in scale. Not a book that sold a million copies and spawned a film franchise, and not a book from a tiny indie press that sold four hundred copies. A mid-list book from a major publisher, or a successful indie from a well-regarded small press, is the sweet spot. "If my book performs like this one, we will all be pleased" is the right register.
3. Close in fit. Genre, tone, structure, audience. A comp that's in the right category but wrong tone confuses more than it helps. "Dark, quiet literary horror" needs a comp that is dark, quiet, and literary — not a splashy commercial horror novel with the same subgenre tag.

Two comps is standard. Three is fine if they are genuinely doing different work (one for tone, one for structure, one for audience, say). More than three is too many.

How to find comps

There is no shortcut for actually reading in your category. Agents can tell when a comp has been Google-selected from a list of "books like X." But some concrete methods:

• Look at the acknowledgments and bibliographies of books you love in your genre. Writers and editors often thank each other, which gives you a network of adjacent titles.
• Read the Publishers Weekly category roundups. Their quarterly previews flag the notable upcoming and recent titles in most categories.
• Read "best of the year" and "underrated" roundups in places like The Millions, Electric Literature, and genre-specific outlets. The underrated lists are particularly useful because they surface books sold at the right scale for a comp.
• Check bookseller recommendations. Independent bookstores' staff picks for your category are a signal of what's selling now.
• Talk to librarians. Readers' advisory is their trade. Ask which novels have been moving in your category this year.

Make a short-list of eight to twelve possibilities, then narrow to two that genuinely share a sensibility with your book. You should be able to articulate, in one sentence, what each comp shares with your novel ("This book shares its domestic-horror register and its rural New England setting"). If you can't say it, the comp is wrong.

Common comp mistakes

• Mega-bestsellers. A debut writer comping a book that sold two million copies signals they don't understand market scale. Imagine an agent pitching an editor with that comp — the pitch reads as overconfident. No matter how much you loved it, don't comp that book.
• Dead classics. Comping a 19th-century novel or a decades-old literary touchstone is a signal that you haven't read anything in the category in years. "Dickensian" is not a comp. "Nabokovian" is not a comp. They are adjectives. Comps are books.
• Author-to-author comps. "Fans of Toni Morrison will love this." Don't. Comp books, not writers.
• Only one comp. Two lets an agent triangulate. One, alone, is either too narrow or too vague.
• Film and TV comps. Occasionally acceptable as a third, flavor comp ("with the ensemble feel of [TV series]"), but two book comps should do the main work. Film comps alone tell an agent you pitch visually and don't read publishing.
• Your own previous books. If you have a backlist, an agent can see it. Don't comp yourself.
• Books that don't exist yet. "My book is like a mix of X and Y, a novel no one has quite written." That's not a comp, that's a claim of uniqueness, and agents regard it with suspicion. Books come from traditions. Show the tradition.

Beyond novels

For nonfiction, the rules are the same but the stakes are even clearer: a nonfiction proposal that comps the wrong books is instantly dismissed. Publishers are buying the market, and the comps are the market. For memoir, comp recent literary memoirs with a similar register. For prescriptive nonfiction, comp recent books that solved a similar problem for a similar audience.

For short fiction collections, comps can be recent short-story collections, but some writers also comp a novel that shares the collection's sensibility, noting the difference in form. For poetry, comps usually don't appear in queries at all — most poetry submissions don't have query letters in the trade-fiction sense.