KDP Sales Estimator

Print (paperback and hardcover)

A zero-inclusive negative-binomial spline fitted to 2008 Amazon.com observations [R3]. With x = ln(rank) and hinge(x, k) = max(0, x − k), the weekly mean is exp(10.083 − 0.782x − 0.378·hinge(x, 11.78) − 0.058·hinge(x, 12.46) − 0.463·hinge(x, 13.02)); the daily pace is that divided by 7. Supported ranks: #100 to #5,000,000 — a product boundary, not a claim that the curve is current there. Both bindings share this one curve; no paperback-versus-hardcover adjustment was established.

Kindle ebook (paid)

An author's 2013 heuristic: daily paid copies ≈ 100,000 ÷ Kindle rank [R4], used only from #100 to #100,000 and never described as an Amazon formula. Kindle Unlimited borrowing and reading make a rank harder to read; this tool does not convert pages read into copies.

30-day pace, scenarios and what is not known

“30 days” always means the modelled daily pace × 30 — a run rate, not a calendar month or a forecast. The sensitivity scenarios (×0.5 to ×2 normally, ×0.2 to ×5 for print above #100,000, ×0.1 to ×10 for unknown, new-release or promotion context) are Zerdly assumptions that show how much an answer depends on the model's scale. They are not confidence limits, and actual sales can fall outside them.

Not established by the research: how accurate either curve is today, Amazon's ranking weights, how many sales a one-day spike needs, a Kindle Unlimited to paid-copy conversion, or any other country's scale. Rank reflects recent and past activity relative to other books, so it can move without your sales changing [R1].

Worked examples (model output)

1. Paperback, overall BSR #25,000, steady rank: the 2008 print proxy gives approximately 37 copies per 30 days (approximately 1.2 a day), with a sensitivity scenario of approximately 19 to approximately 75.
2. A target of 50 copies per 30 days is 1.7 a day. For a paperback (steady) the model-implied BSR is about #17,000 (exact #17,192), with scale scenarios about #2,200 and about #130,000. For a paid Kindle ebook it is about #60,000; the lower-scale scenario is about #30,000 and the higher-scale scenario is outside the supported range (low-activity end).
3. 50 actual paid copies at AUD 6.00 royalty each is AUD 300.00 per 30 days, before tax and other business costs — plain arithmetic that validates no BSR model.

Sources (reviewed 2026-10-10)

Dates describe the source or its data, not a present-day calibration. Formulas are reimplemented; nothing is copied.

• [R1] Amazon KDP Help — Sales Ranking (current guidance) — What a BSR is and is not; separate ranks per format; free and paid Kindle lists differ.
• [R3] Brynjolfsson, Hu and Smith — A Longer Tail? Estimating the Shape of Amazon's Sales Distribution Curve in 2008 (WISE 2009) (2008 data) — Print proxy: equation 3 and Table 4, reimplemented (raw data not supplied).
• [R4] Edward W. Robertson — A Quick Way to Calculate Amazon Sales from Rank (February 2013) — Kindle proxy: daily ≈ 100,000 ÷ rank, an author heuristic.
• [R2] Brynjolfsson, Smith and Hu — Consumer Surplus in the Digital Economy (August 2002 working paper) — Historical context only (an earlier print relationship).
• [R5] Deschatres and Sornette — The Dynamics of Book Sales (2004 submission) — Context: curved relationships and changing activity regimes.
• [R7] Amazon KDP Help — Dashboard (current guidance) — Orders and KENP pages read are reported separately; your KDP reports are the better record.