When Andy Weir began posting chapters of a Mars survival story to his personal website in 2009, he wasn’t writing for an editor or an agent. He was writing for a small, self-selected readership of physicists, chemists, and orbital-mechanics enthusiasts who read each installment and wrote back with corrections. Weir revised earlier chapters to match what his readers told him, and by the time the story became The Martian, it had been shaped by thousands of small, unsolicited peer reviews conducted in public, in real time.
As he later told NPR, readers eventually pushed him to publish the finished manuscript as an ebook, which is how a serialized blog draft ended up as a bestselling novel and a film. The process looked collaborative from the outside. It wasn’t, not in the way that word usually gets used. Weir still sat down alone, chapter after chapter, and wrote every sentence himself.
The comment section is not a co-author
This distinction matters because a lot of commentary about online writing culture treats audience feedback as if it were authorship. It isn’t. In manuscripts submitted for editing, the writers who describe their process as “collaborative” because they post drafts to a subreddit or a Discord server are almost always doing something narrower and more useful: they are testing reactions to material they already wrote in private. The comment thread tells them where a scene drags or where a joke doesn’t land. It does not generate the next paragraph. Even in the most participatory corners of online fiction, the actual labor of choosing one word over another still happens off-screen, before the posting, in whatever solitary conditions the writer has arranged for themselves.
Anna Todd’s experience writing After on Wattpad is often cited as the clearest case of reader-driven fiction, and it is a genuine case, just a limited one. As the Globe and Mail reported in 2014, Todd’s novel started as One Direction fan fiction, and its male lead was renamed from Harry to Hardin in the print edition after a fan vote. That is a real instance of an audience altering a text. It is also a name change, not a rewritten scene or a restructured plot. The vote decided a label; Todd still had to write the 300-plus chapters that made “Hardin” a character worth voting to protect in the first place. Readers can push a decision. They cannot supply the prose that makes the decision matter.
What large-scale reading data actually shows
Research on platforms like Wattpad backs up the more modest version of this claim. A 2020 study by Gerhard Lauer of the Digital Humanities Lab at the University of Basel, published in PLOS ONE, analyzed how young readers write and comment on serialized fiction across thirteen languages. Lauer found a genuine reading culture built around comments in the margins of a text, with different emotional registers depending on genre — readers responding to teen fiction with affective, bonding-style comments, and to classic literature with more analytical, socially cognitive engagement. That is a meaningful finding about how reading has changed. It says nothing about writing changing in the same way. The study documents an active, sophisticated reading public, not a public that is drafting the books it comments on. Conflating the two is where a lot of “collaborative writing” claims fall apart on closer inspection.
What actually shifted is when the audience shows up
The real change, and it is a real one, is timing. Traditional solitary drafting kept the audience out of the room until the manuscript was finished, edited, and sold. Comment-driven and prompt-based writing culture moves the audience’s presence earlier, sometimes into the middle of the first draft. Kathleen Blake Yancey’s 2009 report for the National Council of Teachers of English, “Writing in the 21st Century,” made this case before most of today’s platforms existed: digital tools were producing new models of composition that required writers to think about an audience’s response as part of drafting, not as a separate stage that came after. Yancey’s argument was aimed at classrooms, but it describes exactly what has happened to online serial fiction, newsletter writing, and thread-based essays since. The writer still drafts alone. What changes is that the imagined reader arrives at the keyboard earlier, sometimes before the second paragraph is finished, because the writer has learned from experience roughly how that reader will react.
Editors notice this constantly in newer manuscripts: writers who have built an internal model of their audience so detailed that they no longer need the comments to arrive to write as if they had. That is a skill, and it is one that comment-driven platforms teach efficiently, because the feedback loop is fast and the volume is high. It is not the same as writing by committee.
Where the imagined conversation goes wrong
The complication is that an internalized audience can distort a draft as easily as it can sharpen one. Manuscripts that have been shaped too heavily by anticipated reaction — written, in effect, for the comment section rather than for the story — tend to develop a specific weakness: individual scenes land well in isolation, each engineered for a reaction, but the piece loses structural coherence when read start to finish.
Serialized formats reward the cliffhanger at the end of every chapter; a novel read in one sitting does not need one every ten pages, and manuscripts that carry over the habit read as exhausting rather than gripping.
Writers who move between serialized online formats and traditional long-form manuscripts have to unlearn some of what the comment section taught them, which is itself evidence that the two forms of drafting are not the same activity wearing different clothes.
The discipline underneath the conversation
None of this makes participatory culture a threat to solitary drafting, and treating it that way misreads what is actually happening. Writers who succeed in comment-driven and prompt-based spaces have simply gotten better at predicting a reaction before they need to see it, which lets them draft with an audience in mind without requiring that audience to be in the room.
The keyboard work is still solitary.
Weir wrote every line of The Martian by himself, in whatever hours he had after his day job, and the scientists correcting his orbital mechanics were responding to sentences he had already written and posted, not supplying new ones. The conversation runs alongside the drafting. It has never replaced it.
