Every six months, another self-publishing influencer publishes a tweet thread or a YouTube video about "leaving Amazon" — and every six months, the comments fill up with two camps. The optimists run the numbers on a hypothetical $9.99 ebook and conclude that any author who isn't selling direct is leaving money on the table. The pragmatists point out that Amazon's discoverability engine sells books no D2C store can match.
Both camps are right, and both are missing the same thing: the actual math depends on price, volume, and where the buyer would have shopped if Amazon didn't exist. This post walks through the real per-sale economics in 2026 — KDP at 35% and 70%, Apple Books, Kobo, Google Play, and a direct-to-consumer store on Publica.la — and shows where the threshold lives.
The Headline Numbers
Let's start with a $4.99 ebook, the most common indie price band in 2026. Here's what an author nets per sale on each channel, assuming a typical ~1MB reflowable EPUB:
- Amazon KDP, 35% royalty tier: $1.75 per sale, no delivery fee
- Amazon KDP, 70% royalty tier: $3.34 per sale (royalty: $3.49, minus ~$0.15 delivery)
- Apple Books / Kobo / Google Play: $3.49 per sale (70% royalty, no delivery fee)
- Publica.la (Authors early-access): $3.74 per sale (80% retained, minus $0.25 transaction fee)
Notice anything? At this price point, the Authors early-access program on Publica.la nets more per sale than Amazon's 70% tier — the one most KDP authors qualify for. That's the part most "leave Amazon" arguments skip over.
But the picture changes faster as you move away from the $2.99–$9.99 sweet spot KDP optimizes for.
What Happens Outside the KDP 70% Band
KDP's 70% royalty tier has hard boundaries: $2.99 minimum, $9.99 maximum. Price your ebook outside that window and KDP automatically drops you to the 35% tier. Worse, that 35% applies even on copies sold in your own home market, where reader demand is highest.
Run the same math on a $1.99 short story collection:
- KDP 35%: $0.70 per sale (no delivery fee at this tier)
- Publica.la (Authors early-access): $1.34 per sale (80% retained, minus $0.25)
Now D2C is netting nearly double per sale. The same gap opens up above $9.99 — common for nonfiction, technical manuals, illustrated cookbooks, and box sets. A $14.99 box set on KDP nets $5.25 (35%); on Publica.la it nets $11.74. More than double.
This is the first place where KDP's narrative quietly breaks down. The "70% royalty" headline only holds for a narrow price band, and most authors who price intentionally — for low-friction discovery, for box sets, for premium nonfiction — sit outside it.
The Volume Threshold
Per-sale math only tells half the story. The harder question is volume: even if D2C nets more per sale, can you sell as many copies on your own storefront as Amazon does on yours?
The honest answer is no — at least not in year one. Amazon's discovery engine, ranking system, sponsored ads marketplace, and Kindle Unlimited subscriber base are an aggregate machine no D2C store competes with on raw discovery. If 100% of your sales are coming from Amazon search and you flip to D2C-only overnight, you will sell less.
The math that actually matters is the blended scenario: keep selling on Amazon, but route the readers you already own — newsletter subscribers, social followers, podcast listeners — to your own store instead of sending them back to Amazon.
The threshold typically lands here: once you have an audience of 1,000–2,000 engaged readers (newsletter subscribers, paying patrons, or active followers), the D2C economics start dominating. Below that, the discovery loss outweighs the per-sale gain. Above it, every additional D2C sale is pure margin lift you wouldn't have captured by sending the reader to Amazon.
The Hidden Costs That Move the Number
The per-sale comparisons above are the gross picture. Three hidden cost categories shift the math further in D2C's favor — and are almost never mentioned in standard KDP-vs-D2C tweets:
1. Currency conversion losses
KDP pays in USD, full stop. If you sell in Mexico, Argentina, Colombia, Chile, or any other non-USD market, your local-currency proceeds get converted at Amazon's wholesale FX rate — typically 1–3% worse than what you'd get on MercadoPago, PayU, or Stripe LatAm. For a Mexican author selling 2,000 copies/year at $4.99, that's another ~$200–$400/year quietly evaporating to FX.
2. Payment timing
KDP pays 60 days after month-end. A sale on January 3 doesn't hit your bank until end of March. For full-time indie authors, that working-capital drag is real. Publica.la pays monthly in local currency once the payment processor settles.
3. Reader data ownership
This is the qualitative cost most authors underweight. Amazon does not give you your readers' email addresses. Every reader who finds your book through Amazon search is a one-time transaction. They might come back to your next launch if Amazon's recommendation engine surfaces it. They might not. The relationship is Amazon's, not yours.
A D2C reader, in contrast, is an email subscriber by default. They become a long-term asset: targetable for the next launch, segmentable by genre, removable from the wrong audience when relevant. Joanna Penn, Mark Dawson, and Brandon Sanderson have spent the last decade emphasizing this asymmetry. They're right, and the LTV difference compounds.
Run the Numbers on Your Own Catalog
You can run your own numbers with our Author Royalty Calculator — plug in your price and monthly sales, and it shows the side-by-side annual revenue across every channel, including the D2C scenario on the Authors early-access program.
What This Means for Your Strategy
The honest synthesis isn't "leave Amazon" or "stay on KDP." It's this:
- Below 500 sales/month or 1,000 engaged readers: Stay on KDP. Build the audience first. D2C economics don't compensate for discovery loss yet.
- Above 1,000 engaged readers, KDP 70% band: Run dual distribution. Keep selling on KDP, but build a D2C store and route owned-audience traffic there. The 80% retention on the early-access tier beats KDP's 70% even before counting FX, payment timing and data ownership.
- Pricing outside the $2.99–$9.99 band, or selling in non-USD markets: D2C wins on day one. KDP's 35% tier and FX losses are punitive.
- Selling box sets, illustrated nonfiction, or audiobook bundles: D2C is the only sane choice. KDP's 35% tier on a $14.99 box set is genuinely brutal.
The "Amazon KDP vs D2C" debate has been framed as binary for a decade. It isn't. The right answer for most indie authors in 2026 is both — Amazon for discovery, D2C for the audience you already own. The question is which D2C platform makes that dual distribution operationally sane, instead of forcing you to stitch together Shopify + BookFunnel + ConvertKit + Stripe + ACX and pray.
That's the gap Publica.la is built to close. If you're thinking about adding a D2C channel to your existing KDP setup, request access to the Authors early-access program — we'll walk through the specifics of your catalog and audience together.
