The government just said what "enough evidence" looks like (and it's less than you think)
If you run a charity, you've probably had some version of this conversation with a funder: they ask what difference your programme is making, you give them what you have, and it doesn't quite land. Not because you're hiding anything - because you're not sure yourself what "enough" would even look like. So you assume the answer is more: more data, more surveys, more expensive evaluation you can't currently afford.
It's worth knowing that the government recently published what it expects, and it isn't what most charities brace for.
In June 2026, the Social Investment Agency released its Impact Measurement Standards: a shared standard for how much evidence a decision like that actually needs. It's written for government agencies first, but the agency is explicit that it's also meant for "delivery partners" - the organisations, including many charities, that actually deliver the funded work.
Three levels, not one bar
The standard sets out three levels of evidence. They build on each other, and critically, they're not all expected of everyone.
Level 1 is a credible Theory of Change: a clear, evidence-grounded account of what you're doing and why you expect it to lead to the outcome you want. Every programme is expected to meet this level. It doesn't call for a dataset or outside expertise, just a well-reasoned, honest case for your own logic.
Level 2 is evidence that outcomes have actually changed for the people you're working with - collected "once implemented, where practical and proportionate." Not perfect. Practical and proportionate.
Level 3 is evidence that your programme caused the change, not just that it happened alongside it. The standard is direct about this one: it is "not required for every programme." It's reserved for the genuinely high-stakes calls - decisions that are high-value, high-risk, or strategically significant.
More evidence isn't automatically better evidence
The line that matters most is this: higher levels "do not necessarily indicate more appropriate or more useful evidence in all contexts." And further: "No single number, method, or evidence source should determine decisions on its own. Balanced judgement remains essential."
That's the government's own standard telling its own agencies not to chase the biggest possible evaluation by default. The right level depends on the size and stakes of the decision, not on how impressive the evidence looks.
Read against that, most of the reporting anxiety charities carry isn't a Level 3 problem. It's a Level 1 problem: not having a clearly written account of what you do and why you believe it works. That's a fixable, unglamorous gap, and it's the one the standard says everyone is expected to close first.
Kaupapa Māori evaluation
The standard treats Kaupapa Māori evaluation as a legitimate evidence approach in its own right. It expects agencies to build equity considerations into every level, not treat them as a separate exercise. For an organisation already working this way, that's the government's own standard naming practice you may already hold.
Where this leaves you
Before assuming your impact reporting needs a much bigger project than you can currently fund, it's worth checking something smaller first: can you clearly and honestly walk someone through what you're doing and why you expect it to work? If you can, you're already meeting the bar the government itself sets as the baseline. If you can't quite do that yet, that's the actual starting point - not a research budget.
Sources: Social Investment Agency, Impact Measurement Standards (https://www.sia.govt.nz/assets/Publications/Guidance-Docs/Impact-Measurement-Standards.pdf) Guidance for the Impact Measurement Standards (https://www.sia.govt.nz/assets/Publications/Guidance-Docs/Impact-Measurement-Standards-Guidance.pdf)