Engagements

Find out what your evidence can claim — or build the evidence you’re missing.

Two ways in, each scoped to one short first step: an independent read of what your current evidence can honestly claim, or help building the evidence you don’t yet have.

Prova Method · assess & gradeProva Field · find & build

Prova Method

Find out what your evidence can claim

Programs make claims about what they achieve. Prova Method reads the evidence behind those claims and tells you how much each one honestly supports, and where it runs out.

Find out what your evidence can claim

Independent assessment. Prova Method applies its four-grade claim-evidence rubric. The fee is fixed in advance, and conflicts are disclosed and managed.

01Evaluability Assessment

Know what one program can claim

You want to know what one program’s evidence can honestly claim today.

What it is

A short, bounded read of the claims an organization makes, the records it keeps, and the distance between the two.

You leave with

a claims-to-evidence map, every major claim placed within Prova’s four-grade claim-evidence rubric; a prioritized measurement plan realistic for your budget; and funder-ready language for what can be claimed honestly today and where you are working to strengthen it.

What it does not do

It is a readiness assessment, not an evaluation. It works from existing records, with no data collection, no fieldwork, and no findings about program impact.

Leads to

a fuller study when a decision warrants one, or ongoing measurement.

02Portfolio Evidence Review

Compare many grantees on one standard

You fund many grantees and cannot compare them on a common standard.

What it is

A diagnostic that applies Prova’s four-grade claim-evidence rubric across a portfolio of grantees on one shared standard — grading the evidence behind each grantee’s public claims, and where useful assessing each grantee’s readiness to evidence them. Portfolio size, sampling, confidentiality, and comparability are set during scoping.

You leave with

a defensible, graded map of where each grantee’s evidence stands on one standard; the aggregate pattern across the portfolio, including the share whose claims downgrade; and a shared reporting standard the portfolio can carry forward.

What it does not do

It grades the evidence, never the worth of the program. It does not fix grantees, run new studies, or tell you which to cut. A diagnosis, not a prescription. Per-grantee gradings are held in confidence.

Leads to

a deeper appraisal where a claim matters most, or a standing read on call.

Get the evaluation right

Also independent assessment. Prova Method does not bid on the evaluation work it reviews.

03Risk Pre-Mortem

Stress-test an evaluation before committing

You are about to commission a major evaluation.

What it is

An independent stress-test of the planned design before the money is committed, run against the failure modes that sink large studies.

You leave with

a risk map of the proposed design, graded by likelihood and consequence, with specific strengthening recommendations, delivered before the commitment.

What it does not do

It is a second opinion on the design, not a verdict on whether to proceed, not a competing design Prova would run, and not a review of the evaluator’s competence.

Leads to

a stronger evaluation, commissioned wherever you choose.

04Evidence Spend Review

See what years of evaluation produced

You have spent years on evaluation and want to know what it produced.

What it is

A retrospective audit of three to five years of evaluation spend, building the ledger of what was commissioned, what it cost, where its claims sit in Prova’s rubric, and what changed in practice.

You leave with

a ledger of evaluation spending matched to the evidence it produced, and an honest account you can take to the board.

What it does not do

It reports what was delivered without characterizing past procurement decisions as failures, and it does not conduct or manage an evaluation.

Leads to

an evidence strategy, or a standing read on call.

Also available, scoped on request

Evidence DeskYou want an independent read on evidence questions as they arise.

Want to know how a Prova grade can be trusted — how the firm that judges and the firm that builds are kept structurally apart?

How Prova stays independent →

Prova Field

Build or find the evidence you’re missing

When the evidence you need does not exist yet, Prova Field helps you find it in the data you already have, or build it from the ground up. You own everything it produces.

Build or find the evidence

A separate practice, in a plain advisory capacity. It helps you find or build the evidence, and never also grades it.

01Counterfactual Finder

Find causal evidence without a new study

You want causal evidence but cannot run a new study.

What it is

A structured search of a running program’s operational data for variation that may support a quasi-experimental design, such as an eligibility cutoff, a staggered rollout, or a capacity limit.

You leave with

the candidate comparisons in your data, each with its identifying assumptions and validation tests named. Sometimes no credible comparison exists, and that is a useful finding in its own right.

What it does not do

It identifies the designs. It is not a commitment to run any of them, and you are free to take the specifications to any evaluator.

Leads to

a full causal study built on a design, or ongoing measurement.

02Measurement System Builder

Make grantee reports add up

Grantee reports do not add up to a portfolio view.

What it is

A scoped design for shared outcomes, reporting fields, and aggregation logic. Implementation requirements depend on the portfolio’s existing systems, data, and governance.

You leave with

a shared outcome framework; reporting templates designed for what grantees can realistically produce; and aggregation logic with a board-ready view.

What it does not do

It is the portfolio-level architecture. It does not include individual grantee evaluation, capacity-building, or ongoing reporting operations.

Leads to

a living measurement system that keeps the system running.

Also available, scoped on request

Living Measurement SystemA program is live and worth following as it runs.

Something else

One defined evidence question, scoped before the work begins, on the same method and standard, on one side of the wall. It is scoped to a single defined question, not an open-ended retainer. A single engagement assesses or builds, and not both for the same evidence.

A specimen

What the method looks like when the answer is no.

This is the method worked in the open: a historical case with no present-day owner, the kind of question the Evidence Desk answers. The verdict is the one the published evidence forces, including when that verdict is no.

Evidence Desk · worked example

“Does a Scared Straight–style program reduce youth reoffending?”

Refuseone of three: certify · downgrade · refuse

Systematic reviews of randomized trials find these programs do not reduce reoffending, and several find participants go on to reoffend more than comparable youth left alone. The claim cannot be certified; on this evidence, it is refused.

An illustration on published evidence — not commissioned, endorsed, or reviewed by any organization. It grades the evidence, never a program’s worth, and uses a historical case so no living organization is made the example.

The fit-check call

Unsure which fits?

Bring the question you are trying to answer. The first conversation is a fit check, taken by the founder, with no obligation.

Start a conversation See what your evidence can claim · no charge →