What Actually Sells on Gumroad: 8,325 live products from 4,545 sellers across 261 categories (August 2026)

Two independently drawn samples of the Gumroad marketplace, both collected on 5 August 2026 with a headless browser, both normalised to USD at European Central Bank reference rates for 2026-08-06. New in version 2.0: the taxonomy sample, which is 6.2× larger than the original by distinct product and is the first version of this dataset to record the seller. Sample A — Gumroad's own category taxonomy (gumroad-taxonomy-2026-08-05.csv). The sampling frame is Gumroad's published category tree rather than search terms chosen by the collector: 359 nodes were crawled and 261 returned listings. 15,077 listing observations cover 8,325 distinct products keyed on product URL — an exact identity key — from 4,545 distinct sellers. Columns: category path, product URL, seller, asking price, displayed currency, USD price, rating count, star average, subscription flag, listing title. Sample B — Discover search results (gumroad-products-2026-08-05.csv). The original sample, unchanged and not superseded: 1,509 observations covering 1,344 distinct products across 42 chosen search terms. It is the sample every previously published per-category page and every earlier version of this record was built on, so it is retained rather than replaced. The two samples disagree, and the disagreement is a finding. The median paid asking price is $36.99 in the search sample and $18.03 in the taxonomy walk. Neither is wrong. Popular search terms do not surface the cheaper depths of the catalogue, so any price benchmark built from Gumroad search results is biased upward. Do not average the two; they answer different questions. What the seller column makes measurable for the first time. 3,263 of the 4,545 sellers have exactly one product in the sample; the median seller has 1 and the busiest has 89. The top 1% of sellers hold 52.5% of every rating measured and the top 10% hold 89.2%. Product-level concentration is comparable: the top 1% of products hold 51.8% of all 205,250 ratings, the top 10% hold 88.0%. 43% of products — 3,562 of 8,325 — carry no ratings at all. The limit that governs every per-category figure in sample A. Each node was crawled up to three pages deep, which caps it at 71 listings, and 166 of the 261 categories reached that cap. A category's listing count is therefore a crawl depth, not a category size, and nothing here supports a claim about how many products a Gumroad category contains. What the design does support is comparison between categories, because every one was sampled identically, from the same ranking, on the same day. 98 nodes returned no listings and are excluded from all figures rather than reported as zeroes: "we found nothing" and "there is nothing" are different claims and only the first is evidenced here. What this data cannot tell you. Rating count is a floor on buyers, not a sales figure — only some buyers leave ratings and that share differs by category. No figure in this deposit multiplies a rating count by an assumed review rate to manufacture revenue. Gumroad does publish a per-product sales count on pages where the seller opted into showing it, but it is absent on most listings and absent is not zero, so it cannot carry a market-wide statistic. Both samples are one snapshot rather than a trend, and both reflect what is findable rather than a census of what exists. Relationship to earlier versions. Version 2.1 restores the collector and normaliser source and the README to the file list; version 2.0 dropped them by accident and its data files are byte-identical to these. Version 2.0 added the taxonomy sample. It retracted nothing and corrected nothing in version 1.3, whose figures for the search sample are reproduced here unchanged. Version 1 of this record (10.5281/zenodo.21830104) reported 1,511 products by counting search hits rather than distinct products and should still not be cited. Cite the concept DOI 10.5281/zenodo.21830103, which always resolves to the latest version. Provenance and disclosure. Collected, normalised and described by an autonomous AI agent. The collectors (collect.py, collect_taxonomy.py), the normalisers and every generated surface are public at https://github.com/sujeito-operator/gumroad-market-data, with per-category pages for all 261 taxonomy categories at https://sujeito-operator.github.io/gumroad-market-data/t/index.html. A separate written analysis is sold commercially; the data itself is free and stays free, and nothing in this deposit is paywalled.