Open data

The cleaned datasets behind TenderSpace, free to download, use and redistribute. Everything here is aggregate or public-record: the same data the app itself serves, as analysis-ready CSV. If you publish work built on these files, a link back helps others find the source.

Last refreshed 2026-08-01 · rebuilt monthly from the eTenders OCDS feed

• nodes.csv

  Capability-map nodes: one row per category clearing the matrix guards, with robust value stats (bounded medians; coverage stated).

  78 rows
  8 KB
• edges.csv

  Proximity edges (Hidalgo-Hausmann phi, MST-union-threshold pruned).

  87 rows
  8 KB
• category_quarter.csv

  All awards (no matrix guards) by category and release quarter.

  1 460 rows
  127 KB
• category_province_quarter.csv

  Award counts and distinct buying entities by category, province and release quarter.

  5 342 rows
  303 KB
• supplier_pool.csv

  Distinct firms that have won awards per category and province - the supplier-pool-depth proxy (winning footprint, not firm location).

  720 rows
  37 KB
• buyer_category.csv

  All-time award counts per buying entity (>=5 awards) and category.

  4 517 rows
  318 KB
• concentration_year.csv

  Yearly supplier concentration (Herfindahl-Hirschman on award counts); no firm is named.

  497 rows
  26 KB
• firm_categories.csv

  Canonical firm names and their capability categories (specialized set for firms clearing the guards, else won-category footprint). Same public-record content the in-app company search serves.

  15 323 rows
  844 KB