Andy Weir spent twenty-five years as a software engineer. He worked at Blizzard Entertainment in the nineties, contributed to Warcraft II: Tides of Darkness, then cycled through AOL, Palm, and MobileIron. During those same years, he was also writing fiction. Serials, short stories, whatever he could post to his bare-bones personal website. Most of it went nowhere. His mailing list took ten years to reach 3,000 subscribers, which is a long time to stay interested in something that appears, by almost every external measure, to be failing.
The Martian was one of three serials running simultaneously on that website when he started posting it in 2011. He put up a chapter every couple of months. By the time the last chapter went up, readers were asking for a Kindle version, so he put one together and priced it at 99 cents. By December 2012 it was on the Kindle bestseller list. Crown Publishing called. A film adaptation followed. The whole arc from obscure freeware to bestseller to Ridley Scott production took roughly eighteen months. What made it possible was a decade and a half of writing evenings and weekends while holding down a programming job that paid the bills.
The Routine at a Glance
- Wake Time
- No documented early-morning ritual. Weir writes in the evenings, a habit formed during the years he held a software engineering day job and the mornings belonged to the employer.
- Writing Location
- At his computer, in OpenOffice Writer during The Martian years. Once he signed with Crown Publishing, the editorial process moved him to MS Word. The setup has always been spare: a programmer's desk, not a writer's studio.
- Daily Output
- Weir has described writing as a slog most days. He doesn't fetishize word counts. During the serial years, the pace was roughly one chapter every two months, fitting the work around a full-time job.
- Tools
- Word processor for drafts. Custom C++ code for fact-checking orbital mechanics and trajectory math. Research runs as deep as it needs to, which is often deeper than the book requires.
- Famous Ritual
- Publishing serially to a public audience and using reader feedback from scientists and engineers to correct errors before the next chapter went up. The audience was part of the editing process.
- Books Written This Way
- The Martian, Artemis, Project Hail Mary.
The Programmer Who Kept Writing
In his twenties, Weir took three years off from software engineering to try writing full-time. It didn't work. He went back to tech, took a job, and kept writing on the side. That second attempt is the interesting part of the story, because most people in that position conclude they're not really writers, fold the hobby, and move on. Weir folded the full-time-writer experiment but kept the writing, just relocating it to evenings and weekends where a day job couldn't reach it.
This is the part of his biography that tends to get compressed into a single sentence about self-publishing, but it deserves more attention. For roughly fifteen years after he went back to programming, Weir was building a catalog of online fiction with almost no commercial result. The mailing list grew slowly. A few thousand readers by the time he started The Martian. Nothing that would register as success by any conventional publishing measure. He kept going anyway, partly because he genuinely loved science and wanted to write about it, partly because the internet gave him a direct line to the readers who cared about the same things he did.
The day-job discipline shaped his approach in ways that are worth examining. A software engineer doesn't wait for inspiration to show up before opening the code editor. The work gets done because it's scheduled and because the job depends on it. Weir transferred the same logic to his fiction. He told Writer's Digest that if you wait for inspiration, you'll never get anything done. The writing was just work, subject to the same habits that made professional programming possible: show up, open the file, make progress. He also noted that when he looked back at pages he'd written weeks earlier, he couldn't distinguish the inspired days from the slogs. The output quality was more consistent than the process felt.
Getting the Science Right
The thing that separates Weir from most science fiction writers isn't the ambition to be accurate. Plenty of writers want that. It's the specific method he uses to get there. When he was writing The Martian, he wanted to calculate the trajectory of the Hermes spacecraft using constant-thrust mechanics. Rather than estimate or approximate, he wrote his own orbital simulator in C++. He ran the numbers. The trajectory in the novel is mathematically correct, not plausibly fictional.
Weir has described being terrified that someone like Neil deGrasse Tyson would read the book and publicly call out an error. That fear was productive. It pushed him past the level of research that would satisfy most writers and into territory that required original computation. His twenty-five years as a software engineer meant he had the skills to do it himself rather than ask a consultant. The novel and the code were made by the same person using the same habits of mind.
The research process runs similarly deep across all his work. For Project Hail Mary, Weir has described going way down the rabbit hole on relativistic travel, time dilation, fuel consumption, and the behavior of specific stars. The book's premise requires only one departure from real physics, which is a lifeform capable of storing neutrinos. Everything else in the novel follows from rules that actually exist. That constraint is the entire point of hard SF as a genre, and Weir treats it the way a programmer treats a spec: as a set of requirements the story must satisfy, not as a loose guideline to gesture at when convenient.
There's a particular approach to research embedded in his process that comes directly from engineering culture. He doesn't just look up what he needs for the current scene and move on. He reads until he understands the system well enough to notice when something is wrong. That extra depth is what allowed serial readers, who included actual chemists and physicists, to send corrections that he could evaluate himself rather than take on faith. He'd fix the errors before the next chapter posted. The book got more accurate in real time, with a distributed technical review team who were also the audience.
Writing for an Audience You Can See
Most writers submit a manuscript to an agent, then an editor, and eventually to readers they'll never meet. Weir worked in the opposite direction. He posted each chapter as he finished it, received immediate feedback, and incorporated corrections into the chapters that followed. The audience was present during the writing, not waiting at the end of it.
This changes the psychology of the work in ways that are genuinely unusual. Serialization creates a deadline that has nothing to do with publishing contracts. If you said you'd post a chapter in two months, the readers know it. They're on your mailing list. They'll notice if you don't show up. That kind of low-stakes accountability is actually quite effective, especially for a writer who has no agent, no advance, and no formal obligation to finish. Weir finished because he'd told people he would, and they were watching.
The feedback loop also changed what the book became. Weir has said that he resists describing story ideas to friends and family because talking about a project saps the need to write it down. The serial model solved that problem neatly. The only way his readers could experience the story was to read it. He wasn't describing it over dinner; he was posting it, chapter by chapter, on a schedule, and the act of posting created a record that couldn't be revised by memory or enthusiasm. The work was the work, visible to anyone who wanted to see it.
What this meant in practice was that Weir was writing for an audience he knew cared about accuracy as much as he did. His readers weren't passive consumers catching a plot summary. They were engaged enough to email corrections. That expectation ran in both directions: Weir delivered the science because his readers would check it, and his readers kept reading because Weir kept delivering. The relationship between writer and audience for hard SF is different from most genres, and Weir understood it from the inside before he ever had a publishing contract. Take the genre quiz if you're trying to figure out whether the same reader dynamic applies to your own work.
Discipline from a Day-Job Mindset
Weir's writing advice tends to land harder than most, because it comes from someone who spent fifteen years doing the thing under conditions that offered no external reward. His standard advice is blunt: sit down and write even when it feels like a slog, because inspiration is unreliable and the work looks the same either way when you read it back weeks later. That's not the advice of someone who found writing easy. It's the conclusion of a programmer who's been debugging the same problem since his twenties.
One practical note from his process worth keeping: he sets rules for himself about what he can and can't do during working hours. Browsing the internet for non-research purposes is off the table. TV is off the table. The writing window has a structure to it that looks less like artistic ritual and more like an office policy. This is the day-job mindset applied to fiction, and it works for the same reason it works in software development. Clear constraints remove the decision about what to do next. During work hours, you work. The question of whether to open the file never comes up because the rules already answered it.
There's something instructive in the gap between how Weir describes writing and how the publishing story gets told. The publishing story is cinematic: free serial, 99-cent Kindle ebook, bestseller list, Ridley Scott, Matt Damon. The writing story is considerably less dramatic. A guy in the evenings, at a desk, posting chapters to a website that almost nobody read, for a decade, getting incrementally better at something he loved. The cinematic version happened because the boring version happened first and kept happening long after most people would have stopped.