How to Research a Foundation Before You Apply

The average grant proposal and selection process consumes around 27 hours of nonprofit time, according to Project Streamline research commissioned from the Center for Effective Philanthropy. For a small organization that is most of a working week, spent on a single funder.

Which makes the most expensive mistake in fundraising obvious: applying to a foundation that was never going to fund you. Not because the proposal was weak, but because the fit was wrong in a way that thirty minutes of research would have revealed — the funder only supports organizations it already knows, or gives exclusively in a county you do not serve, or has not made a new grant in your field in four years.

Good funder research is not about finding more prospects. It is about disqualifying bad ones quickly so your limited proposal-writing capacity goes to the handful that could realistically say yes. This guide is the process.

Why mission statements are the worst source

A foundation's website describes what it intends to fund. Its Form 990-PF records what it actually funded. When those diverge — and they diverge often — the filing is the reliable document.

This is not deception. Priorities shift faster than websites get updated, a board takes an interest in something new, a long relationship absorbs more of the budget than planned. The result is the same: a foundation whose site says it supports youth development across the state, whose filings show that 80% of dollars went to three arts institutions in one city, all of which it has funded for a decade.

Under IRS rules, private foundations must make their last three years of returns available to anyone who asks, and contributor identities are not exempt from disclosure as they are for public charities (IRS). Since mandatory e-filing took effect, the IRS publishes these returns as machine-readable data. You do not need to ask anyone. The record is already public.

The five disqualifying questions

Run these before you write a word. Any one of them can end the conversation, and each takes minutes.

1. Do they fund your cause — in dollars, not words? Look at the actual grant list and total the dollars by area of work. A foundation that made two $5,000 grants in your field and $2m in another does not fund your field.

2. Do they fund your geography? Recipient addresses are itemized. If every grantee is in one metro area and you are not in it, stop. Some funders state a broad region and behave much more narrowly.

3. Do they fund organizations your size? Check grantee budgets. A funder whose grantees all have $5m+ budgets is unlikely to make a first grant to a $200k organization, regardless of what the guidelines say.

4. Do they fund organizations they did not already know? This is the question almost nobody asks and it is the most important. Compare this year's grantee list to the last three. If 90%+ are repeats, the effective new-grantee rate is near zero and an unsolicited application is a lottery ticket.

5. Is your grant size in their range? Find the median grant, not the largest. Asking for $150,000 from a funder whose median grant is $15,000 signals you did not look.

Most prospects fail at least one of these. That is the point — the filter is doing its job.

What to read, in order

Once a funder survives the five questions, go deeper. Work through the filing in this sequence.

Part XV, grants list. The whole thing, not a sample. You are looking for: who they fund, how much, for what purpose, and how the purpose is described. Grants labeled "general operating support" tell you something different from grants labeled for a specific project.

Part XV, application procedures. Whether unsolicited applications are accepted at all. A significant number of private foundations state plainly that they do not — and receive applications anyway.

Part VIII, officers and trustees. Who governs. Cross-reference against your own board and network; a genuine connection is worth more than a strong cover letter.

Part I, revenue and expenses. Total giving and its trend. Is the foundation growing, flat, or contracting? A funder in a contraction year is a poor prospect regardless of fit.

Multiple years. One filing is a snapshot. Three shows direction — new areas opening, old ones closing, relationships beginning and ending.

Building the picture the filings do not show

Reading one funder's filing tells you about that funder. Reading them collectively answers questions that matter more.

Who else funds organizations like yours? The most efficient discovery route is not searching funder directories by keyword. It is finding three or four organizations genuinely comparable to yours — same field, similar size, similar geography — and listing every foundation that funds them. That list is your prospect pipeline, and it is derived from behavior rather than from stated priorities.

Who funds you already, and who funds your funders' other grantees? Foundations that co-fund with your existing supporters are warmer prospects than strangers, because your current funder is implicit social proof.

Has this funder ever backed an organization at your stage? Whether a foundation has a history of first grants to organizations with little institutional funding history is a strong predictor of whether it will make one to you.

Plinth's public dataset supports exactly this pattern: it traces 17,896,418 grants across 205,036 US grantmakers for fiscal years 2017 to 2025, and every nonprofit page lists the funders behind that organization. Searching any nonprofit or funder is free with no account required. Look up three peer organizations, collect their funders, and you have a researched pipeline in an afternoon.

A prospect scoring framework

Once you have a list, rank it rather than working through it alphabetically.

Score, sort, and apply to the top of the list. Five well-matched applications beat twenty scattershot ones, and the arithmetic is stark: at 27 hours each, twenty applications is 540 hours — a third of a full-time year.

A 30-minute qualification routine

Research expands to fill available time, so it helps to have a fixed-length version you can run on any prospect without deliberation.

Minutes 0–5: geography and cause. Open the most recent filing and scan the grant list. Are there grantees in your county? Are there grantees in your field? Two noes and you stop here. This kills the majority of prospects and costs almost nothing.

Minutes 5–12: the grant list properly. Read every line of Part XV. Note the median grant, the largest, and whether purposes are described specifically or generically. Flag any grantee you recognize.

Minutes 12–18: three years of grantee lists. Pull the previous two filings and compare grantee names. Count how many are new each year. This is your new-grantee rate and it is the strongest single predictor of whether an unsolicited application has a chance.

Minutes 18–23: governance and procedure. Read Part VIII for trustees — cross-reference against your board and staff networks. Read the application procedures section for whether unsolicited proposals are accepted at all.

Minutes 23–30: the peer check. Look up two organizations similar to yours and see whether this funder appears among their backers. If it funds three of your peers and not you, that is a strong prospect and a specific opening line. If it funds none of them, ask why before assuming you are the exception.

Thirty minutes, and you have either disqualified the funder or have a researched, specific basis for approach. Against 27 hours of proposal writing, the ratio is not close.

What research cannot tell you

Being clear about limits keeps you from over-reading data.

Filings lag 12 to 24 months. A foundation's most recent public filing may describe a strategy it has already moved on from. For current priorities, the website and a phone call matter.

Open opportunities are not in tax returns. RFPs, new programs, and deadlines exist outside the filing record entirely. This is what curated directories are for — see Candid Foundation Directory Alternatives.

Grant purpose fields are inconsistent. Some funders describe grants precisely; others write "general support" for everything.

Individual and DAF giving is invisible. A donor-advised fund grant appears as coming from the sponsoring organization, so the person who directed it is not identifiable.

Relationships are not in the data. The single best predictor of a first grant is often a personal connection, and no dataset captures that.

Making contact after the research

Research changes the conversation, and this is where the work pays off.

Do not open with your organization. Open with something specific and true about theirs: "I noticed you have supported three organizations working on housing stability in the east of the county over the past four years — we work in the same area on the prevention side, and I wondered whether that is still a live priority."

That sentence demonstrates you did the work, respects their time, and asks a real question they can answer in one line. It is also honest — you did notice.

Three rules for the approach:

• Never fake familiarity. "I have long admired your work" from someone who has clearly never read a filing is transparent.
• Ask one answerable question. Program officers respond to questions, not to introductions.
• Accept a no efficiently. A fast no is a gift. It saves you 27 hours.

Where software fits

If you are on the receiving end of this process — a foundation rather than a grantseeker — the mirror image is worth noting. Applicants who research well are asking whether your public behavior matches your stated priorities. Foundations that describe grant purposes clearly in their filings, publish real guidelines, and are honest about whether they accept unsolicited applications get better-matched applications and waste less of everyone's time.

Tools like Plinth support that on the funder side, with applications that screen eligibility in real time so applicants find out quickly whether they qualify, rather than after forty hours of writing. That is a kindness to grantseekers and a saving for program staff. For the underlying discipline, see What Your Form 990-PF Reveals.

Keeping a research file that stays useful

Funder research decays. A prospect you qualified eighteen months ago may have changed program areas, and a foundation you disqualified may have opened a new one. Treating research as a permanent record rather than a one-time task is what turns it into a pipeline.

Keep, for each prospect: the fiscal years you reviewed, the median grant size, the new-grantee rate you calculated, the date you checked, whether they accept unsolicited applications, and any connection you identified. Five fields and a date. The date is the important one, because it tells you when the assessment expired.

Two habits make this compound:

Re-check on filing cycles, not calendar quarters. A foundation's data changes when it files, not continuously. Re-reviewing a prospect six months after their most recent filing is wasted effort; re-reviewing when a new one lands is not.

Record the noes with reasons. A disqualified prospect with "geography — funds only Cook County, checked Mar 2026" written next to it is a decision you never have to make again. A disqualified prospect with nothing written next to it will be researched again next year by someone else on your team.

Organizations that maintain this find the marginal cost of research collapses after the first year, because most of the work is already done and only the changed prospects need attention.

Frequently asked questions

Where can I read a foundation's Form 990-PF for free?

Private foundations must provide their last three years of returns on request, and the IRS publishes e-filed returns. Free access points include ProPublica's Nonprofit Explorer, Grantmakers.io, and Plinth's funder pages, which require no account.

How recent is the data in a 990-PF?

Typically 12 to 24 months behind the fiscal year described, because of filing deadlines, extensions, and publication schedules. Use filings for pattern and behavior; use the website and direct contact for current priorities.

How do I find the median grant size?

List all grants in Part XV and take the middle value. The median is far more useful than the average, which a single large grant can distort badly.

What is a good new-grantee rate to look for?

Compare grantee lists across three years. If more than about 20% of grantees each year are new, the pipeline is genuinely open. Under 5% means an unsolicited application is unlikely to succeed.

Should I apply if the foundation says it does not accept unsolicited proposals?

Generally no. It is a real statement, and applying anyway consumes your time and their goodwill. The better route is an introduction through a shared connection.

How many funders should I research before applying anywhere?

Enough to rank: typically 20 to 40 prospects scored down to a shortlist of five to eight. The research is fast; the writing is slow, so the filter should sit before the writing.

Is it worth paying for a grant database?

It depends on the question. For deadlines, RFPs, and application guidelines, a curated directory earns its fee. For giving history and funder behavior, free public filing data is the same underlying source.

What is the single fastest disqualifier?

Geography. Recipient addresses are itemized in every filing, so you can rule out a funder in about two minutes.

Recommended next pages

• Candid Foundation Directory Alternatives — choosing research tools
• Co-Funder Networks — finding funders through peer organizations
• What Your Form 990-PF Reveals — the filing in detail
• Grant Application Best Practices — after qualification, the writing
• The 5% Payout Rule Explained — how much a foundation must give

Last updated: August 2026