01
Outputs, not outcomes
We deliberately measure outputs - concrete, countable units of delivery - rather than long-term outcomes or impact. A meal provided. A tutoring session completed. A night of crisis support kept available.
This is a deliberate choice. Claiming downstream outcomes - lives changed, cycles broken - requires a level of certainty that even the best-resourced evaluators approach carefully. We'd rather be honest about what can be known than reassuring about what can't.
Why this matters to you
When Ritmo tells you your contribution helped cover one tutoring session, that's exactly what it means - nothing more, nothing less. You'll always know what you're contributing to, and we'll never stretch the truth to make it sound bigger.
02
How we calculate cost
Our goal: traceable numbers from primary sources. For every output, we aim to trace the cost figure back to a primary source - an annual report, a public filing, an independent evaluation. That's the standard we hold ourselves to, and where the data allows it, that's exactly what you'll see.
In practice, nonprofit data varies. Some organisations publish cost-per-unit figures explicitly. Others require us to calculate from separate figures in the same document. In some cases, we're working from website claims that haven't been corroborated by financial documentation. We don't hide this - it's what the reliability labels are for.
03
Output reliability scores
We collect data from thousands of nonprofits. Some have been independently evaluated by specialist researchers. Others report clearly in their annual filings. Others publish claims on their websites that are harder to verify.
Every cost figure on Ritmo carries one of four output reliability labels. Here's what each one means:
The cost figure and output methodology have been reviewed and published by a specialist impact evaluator. These evaluators don't just look at financials; they examine what was actually delivered and how the cost was calculated. This is the highest standard of evidence we recognise.
The cost figure comes directly from the nonprofit's own official reporting - such as tax filing documentation or annual reports - where both output counts and program expenses are explicitly stated. We can calculate the cost per unit directly, without needing to estimate or blend figures.
The cost figure is derived from official sources - like combining program expenses from a tax filing with output counts from a program report. The nonprofit is still the source, and the data is real. But we've had to do some estimation to arrive at a per-unit figure.
This figure gives a general sense of scale. The cost couldn't be fully derived from the organisation's own documentation alone - some assumptions were needed to arrive at a number. Treat it as directional, not precise.
What grades do not mean
These grades reflect the quality of available data, not a judgment on the nonprofit. Most nonprofits don't publish their cost-per-unit data in a form that's easy to verify - that's a sector-wide challenge, not a failure by individual organisations. A lower output reliability label means we have less documentation to work with, not that the work is less valuable.
04
Our sources
All cost figures trace back to primary sources - documents the nonprofit itself is accountable for, or published assessments from recognised evaluators.
Nonprofit tax filing documentation
Public tax filings that include program expenses and, in some cases, output counts.
Annual and program reports
Nonprofit-published reports that describe delivery activity, program reach, and sometimes explicit cost-per-unit figures.
Independent evaluations
Published analyses by third-party impact evaluators that have reviewed program methodology and cost.
Nonprofit websites
Program pages and published impact claims.
What we don't do
- We don't claim downstream impact - lives changed, cycles broken - unless it comes from a rigorous independent evaluation.
- We don't publish a figure without at least one official source we can document and trace.
- We don't treat a nonprofit's size or reputation as a substitute for data quality. A well-known organisation's website claim still gets an Indicative output reliability label if it lacks documentation.