Most Foundations Funded a New Nonprofit Last Year. Not a Few. Most.

One of the most persistent assumptions in nonprofit fundraising is that foundations fund the same organizations year after year, and that breaking in is rare.

The data says otherwise.

What we found

We analyzed 10,613 foundations with $1M or more in annual giving and consecutive IRS 990 and 990-PF filings. For each foundation, we identified which of its grantees in the most recent tax year had not been funded by that foundation in either of the two prior tax years.

The results challenge the "closed shop" narrative:

• 94% of foundations funded at least one organization not previously funded
• Only 6% funded zero new nonprofits
• More than 1 in 4 funded more new organizations than returning ones
• The median foundation directed about a third of its nonprofit grantee slots to new recipients
• Approximately $29 billion in foundation giving went to organizations not previously funded

Stable across every threshold

One of the most striking features of this data is how consistent the pattern is. To test whether the finding is driven by small foundations with only a handful of grantees, we filtered the sample at progressively higher thresholds. The share of grantee slots going to new recipients barely moves:

*Foundation counts reflect those with both new and retained grantees in the analysis year. Of the 10,613 foundations in the full sample, about 1,700 had only new or only retained grantees and are excluded from per-foundation rate calculations.

Even among foundations with 50 or more grantees — the most established, most structured programs — nearly a third of grantee slots go to new organizations every year.

The scale of new funding decisions

In a single tax year, these 10,000+ foundations made roughly 220,000 new funding decisions — grants to organizations they had not previously funded.

That is not a handful of foundations experimenting at the margins. It is a structural feature of how foundation portfolios operate. Foundations do not lock in a fixed set of grantees. They continuously rotate a meaningful share of their portfolio.

But open does not mean easy

High openness does not mean low selectivity. When 10,000+ foundations each make dozens of new funding decisions per year, the total volume of new grant relationships is enormous. No development team can evaluate that manually.

The challenge is not getting in the door. It is knowing which doors to approach.

The foundations most likely to fund a given nonprofit are not random. They follow patterns: thematic alignment, geographic focus, grant size norms, network adjacency to existing grantees. The signal is in the data, but extracting it requires analyzing giving histories across thousands of foundations simultaneously.

The bottleneck is not access. It is signal.

Methodology notes

This analysis uses a multi-year baseline: a grantee is classified as "new" if it was not funded by that foundation in either of the two prior tax years. Using a single-year baseline (year-over-year comparison) produces modestly higher openness rates — about 43% of slots rather than 37% — because some organizations skip a year and return. Both definitions yield the same qualitative conclusion.

We also verified that the pattern is not driven by foundation-level selection effects. Per-foundation medians (computing each foundation's new-grantee rate individually, then taking the median across foundations) produce the same result as aggregate rates, confirming this is not an artifact of a few large foundations dominating the sample.

---

Data source: SciRise Foundation Intelligence dataset. 10,613 foundations with $1M+ in annual giving and consecutive IRS 990/990-PF filings. New recipients are defined as organizations not funded by that foundation in either of the two prior tax years. Excludes DAFs, flow-through and captive foundations, and other non-traditional funders.