Last week I had a job interview at a company that distributes books for independent authors. (I’m super stoked, wish me luck). As with any interview, they wanted to ask me questions to make sure I knew my stuff. As a last question, my interviewer asked me: “Where do you think publishing is headed?”
I probably answered more hastily than I should have:
Faster releases
Shorter works
A commitment to niche audiences over broad commercial appeal.
It just seemed like a no-brainer to me. I didn’t have to think about it. He nodded and seemed like he agreed, so I think it was the “right” answer. But after the call ended, I sat there and pondered on it. After all, my clients and social media mutuals aren’t necessarily reflective of the whole market.
So in fine over-thinker fashion, I’ve since spent some time digging into whether the data backs me up or if I said something insane in a job interview. And because all of you are readers, and many writers, I thought you’d be interested in what I found.
The Backdrop: 300,000 Books a Month
All three predictions live inside one enormous fact.
Economists at the National Bureau of Economic Research tracked new e-book releases on Amazon from 2020 through late 2025. Before ChatGPT, Amazon saw roughly 100,000 new e-books a month. By late 2025, it was over 300,000. When the researchers ran AI detection on a random sample of 50,000 titles, they found no AI-generated text in the 2022 releases (obvi) and in more than 60% of the 2025 releases.
Average quality dropped, which surprises exactly nobody. The thing is that the slop fest isn’t displacing human written books. It’s piling up in the bottom tiers of the market beneath the authors who are putting out good work.
Every author now competes for attention against that pile. Readers don’t prefer AI books, but the sheer volume of them makes your book harder to find. Keep that in mind for what follows.
Prediction One: Faster Releases
I was right, but there’s a catch.
The NBER numbers settle the “faster” question on their own. The volume of new books tripled in three years. However, most of that speed belongs to new and low-performing authors using AI to churn out product, and their books are worse on average. If speed alone were the advantage, those authors would be winning. But they aren’t, they’re just hoping that volume will compensate for low per-book sales.
The authors who are actually winning are fast in a different way. They build a deep backlist and keep feeding it.
Freida McFadden is the best example. She’s a physician who specializes in brain injury, and she spent years submitting manuscripts to agents and publishers without getting anywhere. In 2013 she started self-publishing, and she didn’t stop. The Housemaid, the book that made her famous, was her fifteenth novel. In 2025 alone she released three new novels and a novella. Her traditional publisher now reissues her old self-published books between new releases, and a sales executive there told the New York Times the point was to keep her fans “constantly being fed more Freida.”
Will Wight, who self-publishes fantasy, has written twenty-six novels since 2013. Web serial authors like Matt Dinniman (more on him shortly) post chapters on a schedule, which means their readers never go long without something new.
So yes, faster releases win. But the most successful “speedy” authors were the one who had many manuscripts already written by the time they started pumping them out. This also holds true with author Melissa Cave, who wrote all six books in her wildly successful fantasy series BEFORE releasing the first one. It’s paying off in a big way and having that back list ready to release on a schedule fits with where the market is going.
Prediction Two: Shorter Works
I was HALF right
Readers are absolutely consuming books in shorter pieces. Serial fiction is booming on platforms like Royal Road, Wattpad, and of course Substack, and readers on those sites start with a single free chapter instead of a $30 hardcover.
The problem is the second half of my claim. The works themselves are getting longer.
In 2020, Matt Dinniman was paying his bills drawing pet portraits at cat shows. On the side, he posted a free web serial to Royal Road called Dungeon Crawler Carl. It’s about a man and his ex-girlfriend’s cat fighting their way through a murderous alien game show broadcast across the galaxy, and it’s now eight books long with more than six million copies sold. Travis Deverell, who writes as Shirtaloon, started He Who Fights with Monsters on Royal Road because writing about an interdimensional kung fu wizard sounded more fun than the boring academic paper he was supposed to be finishing. The first deluxe hardcover runs 768 pages. Wight’s Cradle series runs twelve books.
None of these readers committed to 768 pages up front. They committed to a chapter, then another, and a year later they were twelve books deep.
The prediction I should have made is shorter entry points into longer works. Because readers, once they find a world they like, want to STAY in that world for as long as possible. They just want to be sure they like it before they drop money on a door-stop of a book.
Prediction Three: Niche Over Broad Appeal
Call me Nostradamus because I nailed this one.
When Dinniman started posting Dungeon Crawler Carl, LitRPG (fantasy built on video game mechanics like stat screens and level-ups) was about as obscure as fiction gets. Even I had barely heard of it, and I have authors in my inbox seven days a week. He called the LitRPG community “kind of a micro niche,” and he wrote for that small audience without sanding down the premise for anybody else.
No acquisitions editor in Manhattan was ever going to buy that pitch. Upon seeing LitRPG, their internal dialogue would have gone something like this: “Because what do you mean you have character stats on the page? Like a video game or something? What the hell is this? Are these notes you left in by accident? I feel like I’m having a stroke!”
Full disclosure, this is exactly what I though the first time I laid eyes on LitRPG.
The readers, on the other hand, loved it. And they paid money for it. Penguin Random House picked the series up only after the audience was already there. Peacock is developing a TV show, and a tabletop RPG based on the books raised $4.5 million from fans.
Will Wight built his following the same way, through Reddit’s fantasy communities instead of book tours. A couple of years into his career, a fan asked why he didn’t just take his sales numbers to a big publisher like Tor. He said a publisher simply didn’t offer him enough to be worth it, since he was already making enough to write full-time. He went on to hit the New York Times bestseller list anyway.
The broad-appeal genres are feeling the squeeze. Romantasy still sells, but agents are openly using the word “saturation,” and editors are asking for something fresh, which is industry shorthand for “please, no more fae princes.” Readers are drifting into darker sub-niches: villain romances, morally ugly heroines, and horror-romance hybrids someone has named “femgore,” which I like more often than I’d like to admit.
McFadden is the honest exception here. Domestic psychological thrillers are about as mainstream as fiction gets. Her success doesn’t prove the niche prediction. It proves something else.
The Most Important Thing
The thing that my interviewer and I were most stoked about is that all these trends point to one thing: Authors building their own audiences.
Look at the authors I used as examples here. The industry ignored Dinniman until readers had already found him. The same is true of Andy Weir and Hugh Howey before him. Wight had enough going for him that when a publisher came calling, he passed. McFadden spent nine years without an agent while agents and publishers turned her down, and her publisher is now selling her fans the exact same books those agents rejected.
What this means is that you don’t need to tailor your writing for anyone except who you want to be your reader. If you write, ask who your book is for and be specific about it. “Everyone” now includes 300,000 new releases a month, and you’ll lose that fight. Give readers a cheap or free way in, keep giving them more, and build a readership you own. Never forget Andy Weir published the Martian serially and for free before he got picked up by a publisher.
About 30% of indie authors already sell direct through their own storefronts, newsletters, and Kickstarters, and another 30% plan to start. The reason is that readers like to read; that’s it. They don’t require that you be signed to a publisher. They just need to know your book exists and that it’s something they’ll enjoy.
If you read, the front table at the bookstore is the least interesting part of the store. I made a similar argument in Nobody Made You Read This, and it still holds. The authors worth finding are on Royal Road and on Substack. It’s highly likely that the best recommendation you’ll get this year will come from the one friend who won’t shut up about some book you’ve never heard of.
Spend your money on the weird shit that Barnes & Noble won’t stock yet.
My Final Score
Two and a half out of three. Faster releases are indeed winning for authors who plan ahead and focus on quality. Shorter works turned out to be shorter entry points, and niche audiences are winning outright.
If I’m asked for a follow-up in a second interview (fingers crossed), I’ll add the fourth prediction I should have led with: publishing is going toward the authors the big 5 aren’t looking for. We’re all so sick of being sold to, of being hyped up for things that fall flate. Authenticity is the name of the game in all media, and we crave it relentlessly. Indie authors already know that, because finding readers who crave their exact flavor of weird is their entire business model.
And I think it’s about to be everybody’s else’s business model too.
Your point about A.I. flooding the market with low-quality books dovetails with an experiment I recently conducted on my live editing show. Generating readable prose is getting cheaper and faster, but producing a compelling story still takes authentic editorial judgment.
I think that's the point authors chasing rapid release are missing. Five hastily generated novels with structural problems won't necessarily outperform one carefully developed book. More inventory means little if readers have no reason to return.
Re: niche audiences, the real competitive advantage is giving a target readership stories they can't get anywhere else. As A.I. publishing grows, do you think readers will be more discerning about which authors they follow?
I said it years ago: you're gonna see publishers heading towards indies who have big followings. They want readers, so they'll snatch up people who have them.