Why Authors Quit Kindle Unlimited: The Real Economics

In September 2025, Amazon paid Kindle Unlimited authors $0.004521 per page read. For a 300-page novel read cover to cover, the author earns roughly $1.36.

A 99-cent ebook sale, by comparison, pays the author about $0.35. A $2.99 ebook sale pays about $2.10. The math, at first glance, says KU wins on volume: get enough completed reads and you out-earn direct sales.

Except the math does not stop there.

There is an exclusivity tax. There is a scam economy. There is a payout rate that has been quietly drifting down for years. And there is a growing list of authors (sci-fi pros, romance bestsellers, midlist careers) who have run the numbers and decided KU is no longer worth what it costs.

This is what they are seeing.

---

The page-read rate, defined

The Kindle Edition Normalized Page Count (KENPC) was Amazon's solution to a problem of their own making in 2015. They had originally paid KU authors per borrow, regardless of how much of the book the reader actually finished. So every author wrote shorter, faster books to maximize borrow count.

The page-count system was supposed to fix it.

Here is how it works. Amazon allocates a monthly Global Fund. The total pages read across every book in KU during that month (a number Amazon calculates internally) divides the fund. That gives the rate: pool divided by demand.

The result has been a slow downward drift. In the program's early years, the rate hovered closer to $0.005 per page. By 2024, the rate had settled around $0.0045. By September 2025, the rate was $0.004521, with monthly fluctuations between $0.0041 and $0.0046 throughout the year.

The fund grows. The total reads grow faster. Authors get less.

Amazon's official position: this is what a fair market looks like. The author position, increasingly: this is a slow-bleed compensation model that hides its decline behind a wall of opaque internal numbers.

The exclusivity tax

To collect KU page-read revenue, an author must enroll in KDP Select, Amazon's exclusivity program. That means a 90-day commitment, automatically renewing, during which the ebook cannot appear on Apple Books, Kobo, Google Play, Barnes & Noble, the author's own website, library platforms like OverDrive, or competing subscription services.

In exchange, the author gets three things: KU page-read revenue, Kindle Countdown Deals, and KDP's free-book promotions.

For an established romance author with a binge-reading audience, this is sometimes a fair trade. KU readers in romance, urban fantasy, paranormal romance, and LitRPG are voracious. A five-book series at 300 pages each can pull $5-8 in page-read revenue per binge-reader per month. That math, at scale, can outpace direct sales.

For an author writing in any other genre (literary fiction, business, narrative non-fiction, slow-burn thrillers), the exclusivity is rarely worth it. The KU audience does not read those genres at binge volume. The author loses access to roughly 30% of global ebook revenue (Apple Books and Kobo) for marginal page-read income.

That is the basic shape of the choice. Genre determines whether the math works.

The KU economics are calibrated for one specific reader behavior (high-volume romance binge reading) and structurally hostile to almost everything else. The same dynamic shapes web-fiction platforms, where reader-velocity defines what gets written.

The scam economy

Because Amazon counts pages by the location the reader has navigated to, not by pages actually read, the KU system has been gameable since launch.

The first wave was crude. Click farms with thousands of fake Amazon accounts opening KU books and skipping to the last page, generating a full-book "read" in seconds. Cryptographer Bruce Schneier wrote about the original wave in 2016.

The second wave was book stuffing. Romance author Chance Carter became the public face of this in 2018, running KU titles that ran to 3,000 pages: the actual romance novella padded out with bonus content, repeated chapters, full-text catalogs of other books, and other space-filler. Each stuffed book earned approximately $13.50 per "read." The Inc. exposé revealed Carter and others were pulling in mid-six figures monthly before Amazon acted.

The third wave is the AI book farm. Current as of 2025-2026. Bad actors generate hundreds of AI-written titles, enroll them in KU, run paid advertising to surface them on category charts, and bank the page-read revenue from idle scroll-throughs.

Some operations spend $20,000 a month on ads to make $30,000 in KU payouts. Amazon's content disclosure rule from 2023 (authors must report AI-generated content at upload) is widely ignored without enforcement.

The legitimate author pays for all this twice. The Global Fund is fixed in any given month, so every fake "read" subtracts from the legitimate per-page rate. And the legitimate book gets buried under AI-generated chart-stuffers in the same categories. The detection problem in AI-generated romance is documented elsewhere; the economic problem is here.

The exits, named

Authors quit Kindle Unlimited for different reasons but the patterns rhyme.

British crime/thriller author Rachel Amphlett writes that exclusivity is a "business risk" she will not take. Amazon can change KU terms at any time. An author with 100% of their revenue inside Amazon has no leverage. She has been wide since her launch and has never enrolled.

Indie sci-fi author Glynn Stewart pulled two of his series from KU in 2023, citing declining per-read economics and growing library and wide opportunities. His public reasoning: at his scale, the exclusivity premium had stopped covering the wide revenue he was leaving on the table.

A growing number of romance authors (the genre where KU economics work best) have been moving long-running backlist out of KU to recover library and international revenue, even while keeping new releases enrolled.

The reasoning: launch a book hot in KU for the first 90-180 days, then transition to wide once the page-read curve flattens. The publishing-identity math behind a writer's career arc often points the same direction.

This hybrid approach is increasingly the default among professional indie authors. KU is a launch tool, not a permanent home.

The hybrid playbook

For an author starting in 2026, the prevailing strategy among full-time indies is structured like this.

A new release enrolls in KDP Select for the first 90 to 180 days. The book benefits from KU visibility (KU readers are higher volume than direct-sale readers in most genres), from Kindle Countdown Deals (price promotion tool only available inside Select), and from KDP's free-book promotion days used strategically to boost rank.

After the launch window, the book leaves KDP Select. It uploads to Apple Books, Kobo, Google Play, Barnes & Noble Press, and the author's own direct site (Payhip, Shopify, or Gumroad). Library distribution opens through Draft2Digital, Findaway Voices, or PublishDrive.

This captures the launch-window upside of KU and the long-tail upside of wide. The math works for most authors with more than a one-book backlist.

The exception: authors writing exclusively in romance subgenres where KU readers dominate (mafia romance, monster romance, dark academia, sports romance, contemporary billionaire) with backlists of eight or more books and active series often stay all-in on KU. The binge economics outweigh the wide opportunity cost. Even there, the BookTok-pipeline math gets re-run every year as KU rates drift down.

What 2026 will decide

The KU equilibrium has been drifting in the same direction for ten years. More total content. More fake content. Lower per-page rates. More pressure on legitimate authors to game the system or quit it.

Amazon could fix this. The technical solutions are not hard. Charge per genuine page progress, not per page navigated. Require AI-disclosure verification at upload, not honor system. Audit suspicious account-level read patterns. Penalize book-stuffing at the term level, not case-by-case after public exposés.

Amazon has not done any of this at scale. The KU business model (flat subscription revenue for Amazon, variable payouts to authors from a fixed pool) is structurally aligned to maximize Amazon's margin and minimize authors' upside. There is no Amazon incentive to fix what fundamentally works for Amazon.

The question every indie author runs through in 2026 is whether the launch-window math still beats the exclusivity cost. For half the indie market the answer is still yes. For the other half (including some of the biggest names in the genre) the answer has flipped.

Every page in Kindle Unlimited pays a little less than half a cent. Once you know that, the decision matrix is yours. The math is the math.