Methodology

Where the data comes from, and where it doesn't.

Kindred Grants is built on public IRS filings. It's a research tool, not a purchased donor list — and it's only as good as what nonprofits actually report on their tax returns. This page lays out how it's assembled, and the honest limits.

Sources

  • IRS Form 990, 990-EZ, and 990-PF e-filings — the annual tax filings that most tax-exempt nonprofits and private foundations are required to submit. Grantmakers report the organizations they funded on Schedule I (990) or Part XV (990-PF).
  • IRS Exempt Organizations Business Master File — the authoritative list of tax-exempt organizations, including their EIN, name, address, and NTEE (National Taxonomy of Exempt Entities) cause-area classification.

How grantors are matched to recipients

A single recipient can appear a hundred different ways across filings — different abbreviations, punctuation, address of the local chapter instead of the national HQ, and so on. We resolve identities in tiers, and each resulting grant carries a confidence level so you can judge how much to trust it:

  • highMatched by EIN and the reported address confirms.
  • mediumMatched by EIN or exact name, but the address didn't confirm.
  • address_onlyRecovered via a shared address — name and ID didn't match cleanly. Useful signal, but treat with more caution.
  • fuzzyApproximate name match in the same locality. The weakest tier; we surface it for completeness rather than rely on it.

How ranking works

For each grantmaker we compute a simple fit score from your inputs:fit = 0.6 × (share of grants in your cause) + 0.4 × (share of grants in your ZIP-3 area)Grantmakers with no giving in your cause area are dropped by default, and results are sorted by fit, breaking ties by total giving. You can layer on filters (minimum grant count, minimum total in the cause, high-confidence only) before ranking.

Known limitations

  • Organizations that file the simplified 990-EZ won't appear as grantmakers — that form doesn't require reporting grants.
  • Some recipient names couldn't be matched with any confidence and are excluded rather than guessed at. A blank is safer than a wrong lead.
  • Giving patterns reflect the filings loaded, not necessarily current-year priorities. Foundation strategy changes; always verify recent activity before approaching.
  • Cause-area concentration is a signal of fit, not a guarantee of interest. A foundation that funds heavily in Education may still restrict giving to invited proposals.