How Literary Agents Really Work

Literary agents function as independent business partners who screen manuscripts for market viability, help clear structural obstacles before submission, select and approach specific editors, and negotiate contracts and rights under a commission system that pays them only when books sell. Writers can evaluate agents in return by examining recent, relevant sales, concrete submission plans, and rights handling that support a long-term career.

Most writers first encounter agents through short bios on websites or names at the bottom of emails. Inside publishing, their role is narrower and more concrete. Agents sit between writers and publishers as independent intermediaries. They are not salaried editors, lawyers hired by a house, or publicists. They work for the author, are paid only when money flows from a publisher, and spend their time on two kinds of work: choosing which projects to represent and managing the business life of those projects across contracts, rights, and multiple books.

Agents are paid on commission. In many markets, the standard is around 15 percent of advances and royalties on domestic book deals they negotiate, and around 20–25 percent on foreign, translation, and some audio or film and television deals when co-agents are involved. Literary agents do not receive a salary from publishers. Reputable agents do not charge reading fees or retainer fees. Query reading, calls with writers, editorial work on manuscripts, proposal development, submission strategy, and contract negotiation are all unpaid until a deal closes. Most of that time will never be compensated because most projects do not sell. Under those conditions, “liking” a manuscript is not enough. Agents must decide whether it is realistic to recoup the hours required to represent it.

That decision begins long before a full manuscript is read. At the query stage, agents triage in minutes. They discard work that sits outside their stated categories, ignore basic guidelines, or carry obvious amateur signals in the first lines. Based on what remains, they request pages only where the category, length, and premise appear suitable for editors they know. When they read a partial or full manuscript, the first test is not whether it moves them personally but whether it could withstand the questions it will face in an acquisitions meeting. Is the category clear enough that sales and marketing can shelve it and pitch it without confusion? Does the length sit in the band where comparable books have sold in recent years? Can the premise be stated in a few lines to someone who has never heard of the author? Are there recent titles that show readers paying for this kind of work, rather than only a classic from decades ago or a recent highly publicized failure?

Some problems are almost always fatal at this stage. A debut adult novel at 200,000 words in a market where debuts in that lane typically run between about 80,000 and 110,000 words. A memoir that recounts a life in strict chronology with no external frame or question beyond “this happened to me.” A project pitched as both trade nonfiction and scholarly work, with neither audience clearly in view. In those cases, an agent may admire the writing and still decline because the mismatch between the work and the shelves they know is too drastic. Other issues, such as a slow opening, a confusing timeline, or a middle that repeats rather than escalates, are structurally repairable. Agents spend a large share of their time distinguishing between work that could be made viable with focused revision and work that will remain impossible to place, even after months of effort.

Editors rely on that filtering. At trade houses, editors acquire into specific imprints and seasons with a limited number of slots. Each slot carries an internal expectation, including the rough level of advance, the scale of marketing support, and the portion of the list it is intended to support. Editors answer not only to their own taste but to sales, marketing, and finance. Their reading time beyond existing authors is finite. To protect it, they narrow their attention to agents who consistently submit manuscripts that meet basic professional standards and belong to recognizable segments of the market. Over time, patterns form. An editor notices that a particular agent submits work only when they believe there is a strong chance of acquisition. An agent learns which editor can persuade a sales team to back a voice-driven literary novel, which editor has room for another debut in a given year, and which editor has just bought something close enough in topic or structure that another similar book will almost certainly be blocked.

Before a manuscript is ever shown to that group, most practical shaping occurs between the author and the agent. That editorial work is focused, not general craft teaching. An agent is not there to guide a writer from their first attempt at fiction through the fundamentals. They are working on manuscripts that already function at a basic level and trying to remove obstacles that would prompt early passes. That can mean tightening the first chapters so the central tension is visible quickly enough to support a sales pitch, cutting a subplot that blurs the core premise, or restructuring a nonfiction proposal so the central claim, chapter architecture, and market case are clear on the first page. It rarely means redesigning the book from the ground up. When fundamental issues of form, subject, or category make the work impossible to align with the agent’s sense of the market, the honest response is to decline rather than to attempt a rebuild.

Once a manuscript and, where relevant, a proposal can stand without obvious structural failure, strategy replaces triage. A submission list is not a mass email to everyone with “editor” in the job title. It is a short, ranked list of editors and imprints where the agent sees a real possibility of success. That list is built from current knowledge: which imprints are acquiring in the category, which have slowed or closed lists, which editors have a track record with comparable books, and where internal politics or list composition make a pass likely regardless of quality. Timing matters. Editors who have just bought a book very close to yours in topic, tone, or positioning will often have to pass on your work, even if they like it. Houses that have taken recent losses in a niche may be cool to anything that resembles it for a period of seasons.

A clumsy submission is hard to repair. Sending a promising project to an extensive spread of mismatched editors in the first round, targeting imprints that do not publish the category at all, or approaching an entire market with no sense of who is overloaded or closed, can burn through viable options quickly. Once a project has been declined by a large part of the relevant segment of the market, presenting it as fresh becomes difficult. Some avenues remain open, including smaller presses, different formats, or later reimagining, but the simple route into major trade houses may be gone for that version of the book.

If an editor wants to acquire the work, the agent’s focus moves to the contract. Publishing agreements are lengthy and slanted toward the publisher’s interests by default. Agents approach them with a checklist that repeats from deal to deal. They review the grant of rights: which territories and formats are included, whether world rights are being licensed to a house that actually exploits them, and which rights, if any, should be retained for separate sale. They review advances, royalty rates, royalty escalators, and the structure of deep-discount clauses, including their impact on income from bulk or special sales. They examine out-of-print clauses and rights-reversion triggers, which determine when a book can be reclaimed if the publisher stops selling it meaningfully. They read option clauses that can limit where and how the author can submit their next book, and non-compete language that can restrict related work for years. Small changes in those sections can have significant effects on a writer’s income and freedom.

Beyond the primary deal, agents track rights that can extend a book’s reach and earnings. Translation rights, separate UK or Commonwealth editions, audio, large-print, book-club, and some film or television rights can be sold separately. In other instances, granting broad rights to a single publisher makes sense because that publisher has a proven record of effectively exploiting those rights. The decision is not formulaic. It depends on the publisher’s track record, the agent’s subrights network, and the nature of the book. Agents monitor royalty statements and subrights reports over the years to determine whether those rights are being used and whether any reversion clauses are triggered.

Agents also think across books. They consider whether a proposed debut will position an author in a narrow way that constrains future work, whether a short-turn topical nonfiction project will affect how editors see a later novel, and whether a small, poorly handled early publication could make it harder to secure strong support for a subsequent book at a larger house. They see patterns where early decisions lead to stalled careers: a first deal at a marginal press for a book that could have sold to a stronger list, a niche project that dominates the writer’s profile while drawing too few readers to justify follow-up offers, a series of underperforming titles with the same house that sour the internal view of the author’s numbers. When agents decline work that reads well, it is sometimes because they do not see a path to sustainable publishing beyond the first contract.

There are also limits on what agents do. They are not in-house marketers or publicists; they can advocate for better positioning and support, but they do not control cover design, print runs, or advertising budgets. They are not therapists or general life managers. Their primary obligations are to evaluate materials, secure and negotiate contracts, oversee rights exploitation, and advise on career-level decisions within the parts of the industry they know well. Inside their own lists, they triage there as well. Active submissions, negotiations, and contracted work take precedence over new queries and speculative projects.

Agents themselves exist within different stages of their own careers. Senior agents with long lists and deep relationships may have more leverage in some rooms and less room to take on unproven work. Junior or newer agents at established agencies may have fewer sales on record but more time and appetite for pursuing debuts. Both can serve writers well. What matters is alignment between the agent’s connections and experience and the writer’s category, as well as clarity on who will do the day-to-day work and how decisions will be made.

For writers, seeing agents in this light changes the question from “Will someone choose me?” to “Which professional is in a position to represent this work and this career sensibly?” The evaluation runs both ways. Agents are deciding whether a manuscript can survive contact with the market they know. Writers should be determining whether an agent has recent, relevant sales, a concrete submission plan, a command of contract and rights details, and a view of the author’s second and third books that goes beyond general encouragement.