Who it’s for · Programs and nonprofits
Your existing records are the first place to test what your evidence can support.
You know the work matters. What you have not had is an affordable way to find out how much, and for whom. Prova is being built to start from the records you already keep, find the strongest claim they honestly support, and tell you plainly where the evidence is thin and what it would take to strengthen it. The work stays confidential, and the findings are yours.
The bind
Bespoke impact evaluation is often beyond a smaller program’s budget.
Many smaller programs cannot justify the cost or operational burden of a bespoke impact evaluation. Routine reporting therefore often records outputs, such as people served and sessions held, while outcomes and causal attribution remain harder to establish. The gap between what you say and what you can show stays invisible until a funder, a board member, or a journalist asks.
That gap is not evidence of weak diligence. It reflects the cost, data, design, governance, and expertise that stronger claims require. Prova is being built to test how far part of that burden can fall.
What Prova does
An honest read, from what you already record, that respects your team and the people you serve.
Prova starts from the records you already keep. The additional work depends on their coverage, quality, permissions, and fit to the question. It establishes the strongest claim your evidence honestly supports, and tells you plainly where it is thin and what it would take to strengthen it. A modest claim, honestly made, is a real result you can stand behind. If a result disappoints, that is knowledge you can act on rather than a verdict on your worth.
Prova Method applies Prova’s four-grade claim-evidence rubric to the claim, never to the people behind it. It does not run the programs it assesses, and no fee turns on a flattering answer, so the read is honest. What it finds is yours, held in confidence, and goes no further without your word. AI does the labor-intensive reading and synthesis; Prova is being built to establish how far that capability lowers the total cost for smaller programs. The judgment, and the care owed to the people in the records, stays human.
How to start
Start with the claim you most need to stand behind.
A bounded first step can assess what your current evidence supports, where it is thin, and what stronger measurement would require. Timing and burden depend on the records, permissions, governance, and question. Any ongoing measurement function is separately scoped after that first engagement.
Start a conversation
Bring the claim you are least sure of.
If a funder is asking for evidence you are not sure you have, or you would rather know than wonder, that is the conversation to have. The first one is a fit check, with no obligation.
Common questions
The questions a careful reader asks first.
Is cheaper evidence weaker evidence?
A lower price does not require a lower standard. Frontier AI models can now perform much of the reading, extraction, and synthesis in bounded tasks. Prova is being built to establish how far that capability lowers the total cost of reaching a given grade in its claim-evidence rubric without shifting work or risk elsewhere. Design labels do not guarantee credibility. A modest claim remains modest, and a thin record remains thin.
AI can be confidently wrong. Why trust it here?
It can, and that is a central design problem. AI does the labor-intensive reading and synthesis; people decide whether a finding holds, what it means here, and where the evidence runs out. Prova is being built so claims retain provenance and uncertainty. Its reliability across real engagements is still being tested.