How KDPTop Research makes its numbers

Where the data in every report comes from, how each metric is calculated, what the limits are, and how reports are updated and corrected.

Data sources

Amazon Best Sellers and New Releases lists

KDPTop captures every Amazon.com Books and Kindle eBooks Best Sellers list and New Releases list once a day (Top 100 per list), plus Kindle Top 100 Free. Full daily coverage started on 2026-10-27.

Book pages and Keepa

Title, author, price, reviews, rating, publisher, publication date, page count and trim size come from the list cards, the book pages and Keepa's sales-rank history.

Search demand

Monthly Amazon search volume, purchases and month-over-month growth for book search terms come from a third-party keyword tool (SellerSprite). They are estimates and lag by about a month.

Book classification

Each book is tagged with a product type (coloring book, journal, cookbook…), an age band, holidays and audiences by a language model reading its title, description and categories, with rules as a fallback.

Definitions

Unique Best Sellers books
Distinct books (ASINs) on any Best Sellers list on a given day. A book on ten lists counts once.
Breakout book
A book published in the report window that first reached any Best Sellers list in the same window. Days to chart = first Best Sellers day minus publication date.
Independently published (indie)
Published through KDP or CreateSpace, read from the publisher field ("Independently published", empty publisher) or, when that is missing, from the author. Shares are over books whose publisher is known.
Opportunity score
KDPTop's 0-100 score per category for how much room it leaves a new indie book: the review barrier, the share of low-review books, list turnover, the share of new releases and price room. Categories need at least 20 fully parsed books; scores on smaller samples are pulled toward 50.
Niche
A cross-category product market: product type × holiday × audience × age band (for example Christmas coloring books for kids 6-8), over the books on current Best Sellers and New Releases lists.
Niche score
0-100 from demand (estimated sales and search volume, 30%), review barrier (20%), new-release inflow (15%), price room (15%), indie share (10%) and breakouts (10%), plus timing for seasonal niches; multiplied down when many books are not something KDP can print, and shrunk toward the average on small samples.
Estimated sales
Daily and monthly unit sales estimated from Amazon sales rank (BSR) with a rank-to-sales curve per store, calibrated against books with known sales. A single book can be off by ±30% or more; use estimates to compare books and markets, not as anyone's income.
Review barrier
How many reviews the books already on a list have: the median, and the share with fewer than 10, 50 or 100 reviews. A low barrier means books with few reviews still rank.
Launch window
For seasonal books: publish two to four months before the peak month, so the book has a rank and reviews when demand arrives. The peak month comes from Google Trends for the niche's search term, else the holiday.
Median and quartiles
Prices, reviews and pages are medians (the middle book), so a few outliers do not move them; P25 and P75 are the quarter and three-quarter points.

Limits

  • Lists show rank, not sales. Sales figures are estimates from rank.
  • Books KDPTop has not seen on a list are not counted; long-tail books that never chart are missing.
  • Adult content categories are excluded from every report.
  • Reports with less than a full period of data say how many days they cover.

Versions and corrections

Reports are versioned. When a report is refreshed with newer data, the new version replaces the page at the same URL, its version history lists the date and what changed, and every version stays on GitHub as a tag. Errors are corrected the same way and noted in the history. Report a problem: [email protected].

Text and data are released under CC BY 4.0: quote freely with "Source: KDPTop (kdptop.com)" and a link. Data and history on GitHub →