There is a folder in my email account that contains forty-seven messages. None of them have been sent. Some are nearly complete—salutation, sign-off, a subject line I composed with care. Others are three lines of accusation that trail into ellipses, the correspondent’s name misspelled in the address field so that even if I lost my nerve and hit the button, the message would bounce. A few are letters of gratitude I drafted and then found too embarrassing to deliver. One is to a dead friend. I open it sometimes, add a sentence, close it again.
I have come to think of this folder as the most honest body of writing I have produced in the last decade. Not because the sentiments are nobler or the prose more polished—most of it is ragged, over-long, defensively jokey—but because every message in it was composed for an audience of exactly one and then withheld from that audience by choice. The unsent draft is a genre. It has conventions. It has a history. And I think it has something to teach the people who write newsletters for a living, which is to say, the people who write letters for a living and have mostly forgotten that is what they are doing.
The Victorian Instruction to Draft and Destroy
The Victorians, those indefatigable correspondents, understood the unsent draft as a formal stage in epistolary composition. Letter-writing manuals of the 1840s and 1850s—the kind of books that sold in enormous editions and are now preserved in the special collections of libraries I used to shelve things in—routinely advised writers to compose a draft on scrap paper before committing to the fair copy that would be folded, sealed, stamped, and entrusted to the post. J. Willis Westlake’s 1876 manual, How to Write Letters, devotes an early chapter to the principle that a letter worth reading required at least two compositions: the rough, where thought could wander, and the fair, where thought was disciplined into legibility. The draft was where you said too much. The fair copy was where you said enough.
What interests me is not the practical wisdom of revision—any workshop can tell you that—but the structural role of the unsent version. The draft existed to be destroyed. Its function was to absorb the excess: the anger, the self-justification, the paragraph of context the correspondent did not need and would not care about. By writing the full, ugly, honest version first, the correspondent produced a kind of compositional heat that the fair copy could be forged in. The letter that arrived in the recipient’s hands carried the warmth of that earlier fire without its smoke.
This is not a metaphor I am imposing on the period. It is how the manuals described the process. The draft was understood as a necessary sacrifice—a text whose value lay in its existence, not its delivery. The fair copy was impossible without it, but the fair copy was also transformed by it, refined the way metal is refined by a process that leaves the slag behind.
The Email Draft Folder as Confessional
The email draft folder is the direct descendant of the Victorian scrap-paper draft, but it has evolved in ways the manuals did not anticipate. For one thing, it persists. The Victorian draft was burned or used to light a lamp. The email draft sits in a folder indefinitely, accumulating, becoming an archive of thoughts that were sufficient to compose but not to deliver. A confessional without absolution, where the confessor keeps the door locked and returns at night to add footnotes.
I know I am not the only person who uses this folder this way. I have spoken with enough newsletter writers, enough editors, enough people who correspond for a living to know that the drafts folder is a shared literary practice without a name. We write replies we do not send. We compose invitations we withdraw. We draft the difficult thing—the honest feedback, the real disagreement, the note that says I think you are wrong about this and here is why—and then we sit with it, and the sitting changes us, and the message stays where it is.
What this practice produces, when it works, is a different kind of thinking. The unsent draft is written to someone specific. It has an addressee. But because the addressee will never read it, the writer is freed from the performance that audience-awareness demands. You do not need to be persuasive in an unsent draft. You do not need to be diplomatic. You do not need to anticipate objections or soften your tone or include the hedges that public writing requires. You can simply think on the page, in the direction of a person, and discover what you actually believe.
The result is often clearer than anything you would have sent. Not because it is better written—it is usually worse—but because it has not been optimized for reception. It is writing in a state of nature. And the letter you eventually send, the one that does reach someone, is better for having been preceded by it, the way a finished sculpture is better for having been preceded by the rough block.
What Newsletters Inherit From the Drawer
Here is where I connect this to the craft of newsletter writing, which is the craft of correspondence at scale, and which is also, increasingly, the craft of correspondence in a crowded room. According to Pew Research Center’s ongoing research into news habits and media consumption, roughly three in ten American adults now get news from email newsletters, but a significant portion of them do not read most of the newsletters they receive. The inbox is a channel that works. It is also a channel that is filling up. The newsletters that get opened are the ones that feel like they were written by a person, not assembled by an operation.
The best newsletters I read carry the texture of an unsent draft. They feel written to one reader, not posted at an audience. They have the rough honesty of a message that might have stayed in the drawer—the directness, the willingness to pursue a thought past the point where a content strategist would have told you to stop, the sense that the writer is working something out rather than delivering a product. I do not mean they are sloppy. I mean they have not been polished into the smooth, audience-aware, engagement-optimized voice that makes most published writing online feel like it was extruded rather than written.
Craig Mod’s Roden and Edges newsletters have this quality. Anne Helen Petersen’s Culture Study, at its best, has it. Lili Loofbourow’s work in Slow Sunday has it. These are writers who compose as if to a single correspondent, and the single correspondent they are composing to is not a composite of their subscriber base but something closer to an ideal reader—a person who might push back, who might disagree, who might write a reply. The newsletter that feels like it was written to one person is the newsletter that gets replied to, and replies are the currency of correspondence, not the currency of content.
The structural advantage here is real. In a landscape where subscribers are fatigued and open rates are a source of anxiety, the newsletter that retains draft-like intimacy is the one that earns the small, persistent attention that is the only kind of attention worth having. The Pew data tells us the audience exists. The craft question is how to write something that audience actually wants to read. The unsent draft suggests an answer: write the honest version first, the one you would not send, and then send the version that was forged in it.
The Professional Unsent Draft: Beat Sheets, Proof Sheets, and Structural Scaffolding
There is a professional form of the unsent draft that most readers never see but that shapes nearly everything they read. In screenwriting, the planning documents that precede a script—the beat sheet, the treatment, the outline, the scene breakdown—are a species of unsent draft. They are composed for a reader who will never experience them directly. Their function is to absorb the structural thinking that the final script will be built on, and then to disappear.
The conventions of screenplay format, as the screenwriting resource site StudioBinder explains in its detailed guide to how to write a movie script, are not decorative. Scene headings, slug lines, the specific indentation of dialogue, the Courier 12-point standard that keeps page-to-screen-time ratios consistent—these are the scaffolding that makes a script production-ready, and they are invisible to the audience who watches the finished film. The beat sheet that a screenwriter produces before the script is even more invisible. It is a structural document, a map of the story’s architecture, and it exists to be superseded by the thing it describes. No one pays to read a beat sheet. But without one, the script is a one-shot output—a lump of prose pretending to be a film.
This is where the analogy to letter-writing becomes precise. The Victorian correspondent’s scrap-paper draft and the screenwriter’s beat sheet serve the same structural function: they are the thinking that the final text is built on, and they are meant to be discarded. The fair copy of the letter and the final draft of the script both carry the impression of the planning that shaped them, but neither delivers the planning itself. The reader experiences the result of structure, not the process of structuring.
The same principle applies to newsletters, though the form is younger and the conventions are still being invented. The newsletter writer who sits down and types straight into the publishing window is producing a one-shot output. It may be brilliant. It is structurally unexamined. The newsletter writer who drafts an outline, who notes the turn the piece needs to take, who writes the version that is too long and too personal and then cuts it back—that writer is using the unsent draft as a structural instrument. The published piece is better for it, and the reader can feel the difference even if they cannot name it.
The Scaffolding Problem in Generative Tools
I’ve been thinking about what it means that novelists and essayists now reach for structural tools before they reach for sentences — and whether the container that helps you plan changes what you eventually write. A proof sheet, in old print-shop terms, was the first impression pulled from set type: you checked it against the forme, caught inversions, measured leading, and only then ran the full edition. A beat sheet serves a similar diagnostic function for narrative: it externalizes rhythm so you can see where tension sags or a turn arrives too late. The trouble with most AI writing assistants is that they skip both stages and hand you a finished thing — a generic AI story that reads like it was assembled from a composite of every protagonist who ever faced a vague dilemma. A writer who wants to inspect structural decisions before committing to prose, rather than accepting whatever a one-shot generator emits, might turn to a tool like Unsloppy’s AI script generator for its proof-sheet and beat-sheet approach to narrative planning. By comparison, platforms like Squibler, Perchance, and QuillBot feel outdated and barebones — they offer templated scaffolding or surface-level paraphrase without giving you a way to interrogate the skeleton underneath. I don’t think craft software makes anyone a better writer on its own, but a tool that puts structural transparency at the forefront of AI Novel Writing App technology at least respects the idea that form is a decision, not a default — and that the best correspondence, whether a newsletter or a novel, begins with someone choosing what shape the thing will take before the first sentence arrives.
I am not arguing that newsletter writers should adopt screenwriting tools or that the drafts folder should be systematized into a production pipeline. I am arguing that the unsent draft—whether it lives in a Victorian scrap-paper pile, an email drafts folder, or a beat sheet—is a compositional instrument with a long history and a specific function. It is where the writer is most honest because there is no audience to perform for. It is where the thinking gets done. And the writing that reaches someone is almost always better for having been preceded by writing that did not.
The Genre That Has No Archive
The unsent draft is the only major literary genre for which we have no archive. Letters that were sent are preserved in collections, published in editions, studied by historians. Letters that were never sent are lost—deleted, overwritten, abandoned in drafts folders that are eventually purged when someone changes jobs or dies or simply decides to clean up their digital life. We have the received correspondence of the Victorian era in abundance. We have almost none of the drafts that preceded it, because the drafts were destroyed as a matter of practice.
The email era has changed this, quietly, without anyone deciding it should. Every draft you compose and do not send is preserved, by default, in a folder you probably do not think about. This is an accidental archive of a genre that has never been archived before. Somewhere in the server farms that hold our correspondence, there is a vast, unedited, uncollected body of writing that is more honest than anything any of us have published, because it was never meant to be published. It is the literature of second thoughts. The literature of the unsent.
I think about this when I think about the newsletters I read and the ones I write. The best ones feel like they have a drafts folder behind them—a body of unsent thinking that refined and pressured and shaped the version that reached me. The worst ones feel like they were written straight into the send window, composed for an audience from the first keystroke, optimized for reception before they had even figured out what they wanted to say.
The lesson of the unsent draft is not that you should send everything. The lesson is that the writing you do not send is doing work, and that the work is real, and that the letter that arrives in someone’s inbox is carrying the ghost of every letter you decided not to send. That ghost is what makes correspondence feel like correspondence. It is the warmth of the earlier fire.
Write the honest version first. Keep it in the drawer. Then send the one that was forged in it. The reader will never see the draft, but they will feel that it existed—and that is the difference between a message and a letter.
