Who actually sells on Gumroad

Four things the seller column says that a product table cannot

• Demand is more concentrated across sellers than most marketplace commentary assumes. The top 1% (45 sellers) hold 53.1% of all 201,554 ratings measured; the top 10% hold 89.2%; the remaining 4,079 sellers share 10.8% between them.
• It is not a catalogue effect. The rank correlation between how many products a seller has and how much demand they attract is 0.286 — real, but weak. The median seller inside the top 1% has 2 products, and 12 of the 45 have exactly one. Listing more is not the mechanism.
• One-product sellers are the marketplace. 3,251 of 4,532 sellers — 71.7% — have a single product here, and as a class they hold 30.0% of all ratings. The median seller has 1 product; the mean is 1.83, which is what a long tail does to an average.
• 38% of sellers have never been rated at all — 1,710 of 4,532, across their entire measured catalogue. Not a single listing with a single rating. That is the outcome you are competing against, and it is the modal one.

How demand is distributed across sellers

Sellers ranked by total ratings, split into ten equal groups of about 453. Each row is that decile's share of all 201,554 ratings measured.

Cumulatively: top 1% 53.1%, top 5% 79.9%, top 10% 89.2%, top 25% 97.1%, top half 99.7%. The bottom half of sellers hold 0.3% of measured demand between 2,266 of them.

The 300 sellers with the most measured demand

Ranked by total ratings across every listing found. Sellers with 5 or more measured products have a page of their own; the rest are listed here and are in the CSV in full. A product count is what this crawl found, not a catalogue — see the method note below.

Method, and what this cannot tell you

Sellers are derived from 8,311 distinct products keyed on product URL, collected by walking Gumroad's own 353-node category tree on 5 August 2026. Each node was taken up to three pages deep, capping it at 44 listings, and 187 of the 255 non-empty nodes hit that cap. So every product count here is a lower bound on that seller's catalogue, biased down for exactly the sellers whose listings rank deepest. Concentration figures are less affected than counts — a seller missing from the crawl is missing their ratings too — but they are still measured on what was found, not on the whole marketplace.

Ratings are a proxy for relative demand, not a sales figure. Only some buyers rate, that share differs by category, and no revenue estimate anywhere on this site is derived from them. 98 crawled nodes returned no listings and are excluded rather than reported as zeroes.

These are public commercial listings, published under public Gumroad storefront names and linked back to the storefront they came from. Nothing here is inferred about any person. If you run one of these storefronts and would rather not appear, open an issue on the repository and it comes out of the next build.

The data — free, no signup

One row per seller for all 4,532: data/gumroad-sellers.csv — storefront, product count, categories, rating total, demand rank, share of all ratings, median and range of asking prices, free and subscription counts, and their most-rated listing. Every figure on this page is emitted as sellers-summary.json. Derived by normalize_sellers.py from gumroad-taxonomy.csv, so it can never disagree with the product-level release. Data CC BY 4.0, code MIT. Use it for anything.