Evidence-based grant making: what it means and how to do it

Most grant decisions are made on incomplete evidence. Almost everyone who works in grant-making knows this, and most of them are not comfortable with it. They worry that they are missing something important, that they are being swayed by a well-written annual report or a compelling meeting with the CEO, that the charity they decided not to fund might turn out to have been doing exactly what they were looking for.

This is not a failure of rigour or good intent. It is a structural feature of the sector. Grant-makers have limited time. Charities, particularly smaller ones, produce uneven documentation. The evidence that would really answer the question, longitudinal outcome data, independent evaluation, long-run comparison groups, is expensive to produce and rare to find. It exists for some interventions in some cause areas. It does not exist for most of the organisations that most foundations are considering funding.

Evidence-based grant making, understood properly, is not about waiting for perfect evidence before deciding. It is about using the evidence that exists systematically, grading it honestly, and being clear about what you know and do not know when you make your decision.

What "evidence-based" actually means here

The phrase has its origins in medicine, where it carries a specific technical meaning tied to randomised controlled trials, systematic reviews, and meta-analyses. In grant-making, it is used more loosely, and that is probably appropriate. The question of whether a community theatre programme improves mental health outcomes in young people is not well-served by a clinical trial methodology.

For grant-making purposes, evidence-based means that the decision rests on documented, traceable reasoning rather than impression. It means the evidence you used is recorded, the quality of that evidence is acknowledged, and the reasoning that led from the evidence to the decision can be reconstructed afterwards. It does not mean the evidence is certain. It means its limitations are visible.

This distinction matters because it changes what you are aiming for. You are not trying to eliminate uncertainty. You are trying to be honest about it.

Grading evidence quality

Not all evidence is equally useful, and treating it as though it were produces distorted assessments. A charity that has commissioned an independent evaluation of its programme is telling you something different from one that has published a selection of beneficiary testimonials. Both types of evidence appear in annual reports. They are not equivalent.

A simple grading approach works well for most foundations. Strong evidence is independently verified, uses a methodology that is explained and reasonable, and measures changes in outcomes rather than just counts of activity. Moderate evidence is plausibly self-reported, distinguishes between what the organisation did and what changed as a result, and acknowledges limitations. Weak evidence counts activity without connecting it to outcomes, or makes causal claims without any basis for them.

Missing evidence, a charity that reports nothing about its impact beyond financial summary, is itself a data point. It might mean the organisation has not had the resources to invest in evaluation. It might mean the organisation is not thinking about outcomes at all. Knowing which is part of what a good assessment establishes.

Where the evidence actually comes from

For most UK charities, the evidence available to a grant-maker without a direct relationship comes from a small set of sources.

The annual report and accounts, filed with the Charity Commission, are the most consistent source. The trustees' report is required to describe the charity's activities and achievements, and more recently the Charity Commission has emphasised the expectation that charities report on public benefit. The quality of this reporting varies significantly. Some trustees' reports are substantive and honest. Others describe activities at such a level of generality as to be uninformative.

The charity's own website and any published impact reports provide additional self-reported information. These should be read critically. An organisation that publishes only what went well is telling you something by omission. An organisation that acknowledges what the evidence does not show, or where a programme did not achieve what was hoped, is demonstrating a kind of intellectual honesty that correlates with better practice more broadly.

For some cause areas, academic research on the intervention type is available and relevant. A grant-maker funding employment support programmes can read what the research literature says about which approaches tend to produce sustained employment outcomes and which tend to produce short-term ones. This does not tell you whether a specific charity is effective, but it tells you what to look for and what questions to ask.

360Giving publishes open grant data from UK funders, which lets you see which other foundations have supported an organisation and at what level. A charity that has been consistently funded by well-regarded foundations over several years has passed a series of informal assessments that are worth knowing about, while not substituting for your own.

The problem with anecdote

Annual reports are full of case studies. A named beneficiary describes how a programme changed their life. A photograph of a smiling group at a workshop. A quote from a community leader. These are not evidence of impact in any rigorous sense, and it is a mistake to treat them as though they are.

But they are not useless, and dismissing them entirely is also a mistake. A case study tells you something about how the organisation thinks about its beneficiaries, whether it respects their dignity in how they are described, whether it uses their stories instrumentally or treats them as participants in something. A charity that publishes beneficiary stories with their subjects' full agency and voice tends to relate to beneficiaries differently from one that presents them as objects of its good work.

The appropriate response to anecdote is to note it and file it in the right column: this tells me something about culture and values, not about population-level outcomes. Both matter. They are just different things.

Building a process that does not take forever

The biggest practical barrier to evidence-based grant making is time. Trustees and grants staff are not usually paid by the hour for due diligence. A process that takes eight hours per charity is, in most contexts, a process that will not be used consistently.

The solution is standardisation. Decide in advance which sources you will consult for every charity on a shortlist: the Charity Commission register, the most recent two sets of accounts, the charity's website, and a direct conversation. Decide which evidence quality grades you will apply and what each grade means. Build these into your assessment template so that the time spent is on reading and thinking, not on deciding what to look at.

What requires deeper attention is the thing that looks unusual: the drop in income that needs explaining, the auditor's qualified opinion, the claim about impact that does not match the methodology described. Standardisation means you find these things reliably, rather than discovering them by chance.

Deciding under uncertainty

Some funding decisions will always be made with incomplete evidence. The choice is not between deciding with evidence and deciding without it. It is between being explicit about the evidence you have, its quality and its gaps, and being implicit about it.

A trustee who says "the evidence of impact is moderate, self-reported, and does not distinguish outcomes from activities, but the governance is strong and the financial position is stable, and we believe the intervention approach is plausible given what we know about the cause area" has made an evidence-based decision. A trustee who says "we liked the annual report" has made the same decision, possibly with the same quality of evidence, but without the reasoning visible.

The purpose of making the reasoning visible is not to prove you were right. It is to give yourself and your fellow trustees something to learn from when you find out, a year or two later, whether the decision led where you hoped it would.

cleargiving.io grades the evidence quality for each criterion in an assessment automatically, surfacing what is strong, what is thin, and what is missing, so trustees go into every discussion knowing exactly what their decision is based on. You can read more about how the evidence grading works, look at our guide to evaluating impact, or request access to see it working on a real charity.