The Royalty Math Nobody Actually Runs

The conversation about where to publish adult fiction always seems to skip the part where someone runs the numbers honestly. KDP advocates point to the 70 percent royalty rate. Subscription advocates point to the steady monthly income. Direct-marketplace advocates point to the 80 to 95 percent payouts. Each side is telling a true thing and leaving out everything else.

What's missing is a clean comparison that accounts for the things that don't show up on the platform's marketing page. The returns rate. The dungeoning losses. The AI-flag suspensions. The audience-building cost. The processor fees that get quietly subtracted before the headline percentage is calculated. The realistic conversion rate from "your book is technically for sale" to "money lands in your account." Run all of that and the rankings shift in ways that genuinely surprise people who've been doing this for years.

So let's run it. With actual numbers, with the assumptions stated, with the caveats marked, for a hypothetical taboo author publishing the same book through three different paths. Nobody needs a winner crowned here. What's useful is seeing what the comparison actually looks like once you stop accepting platform-supplied numbers at face value.

The hypothetical

The book is a 60,000-word taboo romance, priced at $4.99 in ebook form. The author has a modest existing readership of about 800 engaged readers across various platforms, accumulated over three years of consistent publishing. The marketing budget for this release is zero, which is the realistic baseline for most working authors. The launch window we're measuring is the first six months from publication, which is when the bulk of any title's revenue tends to land.

Three paths, three sets of numbers.

Path one: KDP exclusive

The simplest setup. Author enrolls in KDP Select, gets the 70 percent royalty rate at the $4.99 price point, accepts the exclusivity requirement that keeps the book off every other platform.

Gross sales over six months, in a realistic dungeoning-adjusted scenario for taboo work, land somewhere between 200 and 400 copies. The dungeoning effect kicks in after week three or four, at which point organic discovery collapses and the only sales coming in are from direct links the author has placed elsewhere. Let's call it 300 copies as the median outcome, which is generous for taboo work but represents an author who's hustling.

Gross revenue: 300 × $4.99 = $1,497. Royalty rate: 70%, minus delivery fee of approximately $0.15 per copy. Net royalty: 300 × ($3.493 - $0.15) = $1,002.

Add Kindle Unlimited page-read revenue, which for a 60,000-word book at 2024-2026 rates averages roughly $0.004 per page. Assume 50 readers borrow the book and read it through, generating about 240 pages of read-through each. KU pages: 50 × 240 = 12,000 pages. KU revenue: 12,000 × $0.004 = $48.

Total six-month KDP revenue: approximately $1,050.

The hidden costs that don't show up on the dashboard. The book is exclusive to Amazon, which means every reader the author has on other platforms is functionally locked out from buying it. The audience the author built on AO3, on Reddit, on their Substack, on their SubscribeStar, can't purchase this title without creating an Amazon account, in many cases against their personal preference for privacy reasons. The exclusivity cost is unmeasurable from inside the system, but the authors who've run side-by-side comparisons consistently find it's significant.

The risk costs that don't show up on the dashboard. Account suspension at any point in the next three years takes the entire backlist with it, including this book and all royalties yet to be paid. The probability of suspension for a taboo author over a three-year window, based on community-reported data, is somewhere between 15 and 25 percent. The mechanism behind that risk, dungeoning included, is worth a separate read, and the payment-processor pressure pattern that drives most of these suspensions sits one layer further out. Adjust your expected lifetime revenue accordingly.

Path two: SubscribeStar primary

The author skips KDP entirely, puts the book up as a downloadable on their SubscribeStar tier system. New subscribers get access to the full backlist including this title. Existing subscribers get it as a new addition to their tier. The mechanics of why the subscription model produces higher per-reader revenue than per-book sales gets the longer treatment in its own piece.

The pricing model is different from per-book sales. The author runs three tiers: $5, $10, and $20 monthly. The new release drives a subscription bump, with maybe 60 new subscribers signing up in the launch window, distributed across the three tiers roughly proportionally to the value offered. The realistic conversion math, based on patterns visible in adult creator communities, looks like 35 new $5 subscribers, 18 new $10 subscribers, and 7 new $20 subscribers.

New monthly recurring revenue: (35 × $5) + (18 × $10) + (7 × $20) = $495.

Of those new subscribers, retention over six months tends to track at roughly 60 to 70 percent. Some cancel after consuming the backlist, some forget they subscribed and stay, some convert to fans who stick around for years. Take a 65 percent six-month retention rate as the baseline.

Six-month subscription revenue from new sign-ups: $495 × 6 × 0.65 = $1,931, approximately, with the actual number depending on the exact cancellation timing.

SubscribeStar takes about 7 percent on average across tiers, plus payment processing of roughly 3 to 5 percent depending on the specific processor in use. Total platform cost: 11 percent.

Net author revenue from new subscribers: $1,931 × 0.89 = $1,719.

Add the existing subscriber base, which is paying regardless of this specific release. The 800 readers from across platforms include some portion who are already subscribed. Assume 80 of them are existing $5 to $10 subscribers, contributing about $600 monthly already. That income continues independent of the new release, so it doesn't factor into the comparison directly.

Total new revenue attributable to the book launch, six-month window: approximately $1,720.

The hidden benefits that don't show up on the dashboard. The new subscribers don't expire at six months. The retention curve flattens out and produces ongoing income for years if the author keeps publishing. A subscriber who sticks around for two years is worth eight to ten times what they were worth in the first month. The compounding effect is significant and real, but it doesn't fit into a six-month comparison cleanly.

Path three: Direct marketplace

The author publishes the book on Maliven, a direct-to-reader marketplace that runs on BTCPay rather than traditional payment processors. The book is priced at the same $4.99, with the author keeping a higher percentage of each sale because the platform doesn't have to absorb Stripe or Visa fees.

The platform takes 15 percent on each sale, leaving the author with 85 percent of the gross before any small Bitcoin network fees. The realistic sales for a six-month window depend almost entirely on the author's audience-driving capability, because the platform's organic discovery is smaller than Amazon's by orders of magnitude.

For the 800-reader baseline audience, with active promotion across AO3, Reddit, Substack, and any other personal channels, a realistic conversion rate is somewhere between 8 and 15 percent. Take 11 percent as the median.

Gross sales: 800 × 0.11 = 88 copies. Gross revenue: 88 × $4.99 = $439. Net author revenue: $439 × 0.85 = $373.

That's the launch window from existing audience alone. The platform's own traffic adds some additional sales on top, driven by category browsing and search. For taboo work specifically, the marketplace surfaces titles in ways the mainstream platforms don't, so the contribution can be meaningful but unpredictable. Add a conservative 25 percent on top for platform-driven discovery.

Total six-month direct marketplace revenue: approximately $470.

The hidden benefits that don't show up on the dashboard. The author retains the email list and reader relationship without platform mediation. The book never gets dungeoned, never gets pulled, never causes an account suspension. The audience the author already had can purchase the book without creating accounts on a platform they don't want to use. Multi-platform parallel publishing is allowed, so the same book can run on SubscribeStar, Maliven, and the author's personal site simultaneously without exclusivity conflicts.

The comparison, with the honest caveats

Six-month launch window for the same book, same author, same audience:

KDP exclusive: $1,050, with 15 to 25 percent suspension risk. SubscribeStar primary: $1,720, with smaller suspension risk and compounding ongoing revenue. Direct marketplace alone: $470, with effectively zero platform risk.

The temptation is to read these as a ranked list and pick one. The right read is that they're not mutually exclusive, except for the KDP option which forces exclusivity. The author who runs SubscribeStar and a direct marketplace in parallel, with the book also for sale through smaller retailers like the Smashwords store and similar non-exclusive paths (with the asterisks noted in our retailer-filtering breakdown), or through self-hosted options like Payhip, generates revenue from all of those streams simultaneously. The SubscribeStar number doesn't go down because the book is also for sale elsewhere. The direct marketplace number doesn't go down because the book is also available on a subscription tier.

The combined six-month revenue for the parallel-publishing path is approximately $2,190, which exceeds the KDP-exclusive number by a meaningful margin while also producing the compounding subscription audience and the unblockable direct sales infrastructure.

The KDP path is not the obvious winner the marketing implies. For taboo work specifically, it isn't even competitive once the realistic sales numbers and suspension risks are factored in. The platforms that look smaller, slower, less polished are doing better math than the platform that looks like the default.

The longer time horizon, which changes everything

A six-month launch window is a useful slice for comparison, but it's a poor representation of how these revenue paths actually behave over an author's career. The KDP path produces a sharp peak and a long tail of declining sales as the dungeoning effect accumulates. The subscription path produces a slow ramp that compounds for years. The direct marketplace path tracks the author's audience growth across all platforms, with the marketplace becoming a higher-converting tail surface as the audience matures.

Project the same hypothetical book three years out and the picture changes considerably. KDP exclusive revenue in year three tends to fall to under 20 percent of the launch year, because the dungeon doesn't lift and the algorithm has fully written the book off. Subscription revenue from the readers acquired in the launch window, retained at typical adult-creator rates of about 25 to 35 percent over three years, continues to produce monthly income that often exceeds the original launch month. Direct marketplace revenue grows roughly in proportion to the author's expanding audience, which compounds because the marketplace doesn't take it away from them.

Three-year revenue from the same single book, by path: KDP exclusive at approximately $1,200 to $1,400, with suspension risk knocking 15 to 25 percent of that off in expected value. SubscribeStar primary at approximately $5,000 to $8,000, depending on retention specifics. Direct marketplace alone at approximately $900 to $1,500. Parallel publishing path at approximately $7,000 to $11,000.

The longer the time horizon, the worse KDP looks in comparison. And the worse it looks, the harder it gets to justify the exclusivity requirement that prevents the author from running any of the other paths in parallel.

The numbers nobody publishes

The reason this analysis doesn't appear in the dozens of "where should I publish my erotica" articles that circulate every year is that nobody publishes the data needed to run it cleanly. Amazon doesn't publish dungeoning rates. SubscribeStar doesn't publish retention curves. Direct marketplaces don't publish per-author conversion numbers. The figures in this article are pulled from a combination of community-reported data in adult writing forums, the Alliance of Independent Authors' 2025 piece on erotica restrictions, conversations with working authors who track their own metrics, and the kind of pattern observation that anyone running a platform in this space picks up over time.

The figures are approximations. The methodology is informal. The point of running them isn't to provide a definitive forecast for any specific author's situation. The point is to show that the comparison most authors make in their heads, where KDP is the default and everything else is a sacrifice, doesn't survive contact with the actual numbers. The default has been wrong for several years. The replacement options have been viable for longer than most authors realize (here's the full map of where adult fiction actually sells today). The migration that experienced taboo writers have been doing quietly since 2023 is starting to be visible enough that newer authors can finally see what their predecessors figured out by accident.

The math nobody runs, when run, doesn't say what most authors expect.