Skip to main content

Coverage & reconciliation

Every number, and the rule behind it

The site publishes several counts because they answer different questions. Dataset 851 is the benchmark census (as of 2026-10-30); the report catalogue is smaller, and headline totals exclude rows the source flags as stale. Same data, different filters — not different truths.

ScopeCount30-day revenue

Revenue Dataset (benchmark census)

The 851-product census behind /benchmarks; not a report count.

851—

Report specs collected

Reference report layouts we hold locally, published or not.

742—

Published report pages

status = published.

738—

Report cards with a screenshot (LISTABLE)

Published + has a screenshot. This is what /library lists.

738$14.62M

Excluded: revenue flagged stale

The source marks the revenue figure stale; the page stays, the number leaves the totals.

90$2.56M

Counted population (headline totals)

LISTABLE minus stale rows. Homepage totals, rankings and this page use it.

648$12.06M

Cards with non-zero MRR

MRR > 0 across every stored row (not restricted to the counted set).

597—

Cards with a 30-day revenue figure

revenue_last30d is not null.

716—

By business model

  • Subscription88 · $4.69M
  • (unlabelled)534 · $2.33M
  • Marketplace Fee9 · $1.41M
  • Managed Service2 · $975,043
  • Hybrid2 · $904,675
  • Transaction Fee1 · $681,454
  • Usage-based4 · $653,774
  • Freemium8 · $417,783

Needs review

56 cards where 30-day revenue and MRR differ by 10× or more. The figures can both be right (a one-off charge, an annual prepayment), but they should not be quoted as a run-rate without checking.

Growth methodology

  • The growth rankings use the source-reported 30-day change: current 30-day MRR vs the previous 30 days. They are not a computed 7-day window, and a daily 7-day series is not published yet.
  • Only cards whose base MRR clears a minimum threshold are ranked, so a tiny base cannot produce a 10,000% row.
  • One-off charges are not excluded: a single large transaction can lift a marketplace or service row for a month. Rows where 30-day revenue and MRR disagree by 10× are flagged "Needs review" for that reason.

See also: Method · Same-model rankings · Data changelog