Skip to content
Idea Atlas
Sign in

SellerKit

As described:StopNarrowed:Narrow

No as described. Go, small, as a KDP backlist pricing and promotions extension.

self-publishing · ~$20 a month per paying account · last reviewed (next review by 29 Dec 2026) · 29 cited sources

Save
Share

← All research

5. Pre-Mortem: It's September 2028 and SellerKit Is Dead

Three most likely causes of death, ranked by probability. They compound: #1 makes #2 worse, because each outage is a reason to cancel.

Cause #1: Amazon ended it (probability ~45% of failure cases)

What happened. It happened in one of three ways.

  • (a) The page changed. KDP redesigned the Bookshelf in a spring update. The extension's selectors broke for every customer at once. The fix took 5 days because the new page loaded pricing through a different flow. Two author groups posted "don't use it". A third of paying accounts cancelled that month.
  • (b) The feature shipped. KDP added "Edit prices for multiple books" and scheduled price changes. The core job became free overnight. Easy Folders saw the same thing when ChatGPT shipped Projects.
  • (c) The policy moved. Amazon extended its March 2026 Agent Policy language to KDP. Tools that "simulate browser interactions" were named as prohibited. One customer's account was flagged, the post went viral in an author group, and sales stopped.

Early warning signs

  • Any breakage that takes more than 48 hours to fix.
  • KDP announces a Bookshelf redesign, a beta for "multi-title" actions, or anything about third-party tools.
  • An author-group post that says "flagged", "suspended" or "warning" next to any extension name, including a competitor's.
  • The KDP terms page changes its "last updated" date (27 Sept 2024 on 30 Sept 2026). Watch it monthly.

Fixable?

  • (a) Yes, with the breakage system in 07: selector health checks on load, anonymous failure reports, a test account and a 24-hour fix rule.
  • (b) Partly. Move to what Amazon won't build: pricing across Kobo, Apple and Google Play as well as KDP, and promo workflows for wide authors. That's Pivot A.
  • (c) Only by design, and only in advance. Everything user-driven, visible, human-paced and in the user's own tab. No background runs, no servers, no stored logins. If a policy names your pattern anyway, stop selling within a week (the kill switch).

Cause #2: "Reprice once, cancel" (probability ~35%)

What happened. Launch went well. The July 2026 royalty change drove 300 free checks and 90 paying accounts in the first quarter. Most were one-time repricers. They fixed their 35% titles in the first week and cancelled at the next renewal. Month-3 retention was 45%. Promoters, the monthly users, were only a quarter of payers. MRR peaked at $1.8k in month 5 and drifted back to $1k.

Early warning signs

  • Fewer than 25% of payers use the promo scheduler in any month.
  • Month-1 renewal below 60%.
  • Most sign-ups arrive in spikes after Amazon news, not steadily.
  • Fewer than 40% of payers on annual plans by month 12.

Fixable? Yes, mostly. Sell to promoters (rapid-release fiction and presses), not to everyone with a backlist. Push annual plans at launch with a real discount. Make the monthly health report good enough to open. Offer a cheap "event pass" (a one-month plan at a higher price) so one-time repricers pay fairly and don't pollute retention numbers.

Cause #3: The pool was too small for the ladder (probability ~20%)

What happened. The product worked, customers stayed, and churn ran at 4%. But after 18 months, author groups were saturated. Nearly everyone with 20+ titles who pays for tools had heard of it. New sign-ups fell to 15 a month. MRR stalled at $3k. Getting to $10k needed ~500 publishers, and there weren't enough of them who wanted this at $20 a month.

Early warning signs

  • The same 5 communities produce more than 80% of sign-ups at month 9.
  • Affiliates stop producing after the first launch push.
  • Chrome Web Store search brings fewer than 20% of installs.
  • Press plans (the highest price) stay under 5% of accounts.

Fixable? Partly. Raise the price per account (Press plans, several pen names, team seats), add cross-store pricing to reach wide authors, and accept that $3–5k MRR may be the natural size. Run it profitably, or sell it. Merch Wizard sold in 5 hours at a known multiple. A small, clean, profitable extension is a liquid asset.

Honourable mentions (less likely, still real)

  • Chrome Web Store action. A review flags the extension for "single purpose", permissions or a policy change, and the listing is pulled for a week. Keep permissions narrow (KDP domains only) and the single purpose obvious.
  • Payment-provider dependency. ExtensionPay is a small third party that takes 5%. If it fails or changes terms, licensing breaks. Keep a Stripe-direct fallback ready by month 6.
  • A cheaper clone. The code is easy to copy with AI. A free clone appears in a Facebook group. The defence is trust and speed of fixes, not code.
  • A mistake at scale. A bug sets 200 books to $0.99 in every marketplace. Preview, per-batch limits, undo and a "confirm the total count" step are not optional.