Phase 01
Evaluative Foundation
The conceptual infrastructure that makes rigorous evaluation possible: stakeholder map, logic model, theory of change.
Programs & Services
A four-phase evaluation pipeline. Each phase is a link in a chain and a standalone engagement.
Phase 01
Phase 02
Data Architecture
A practical, defensible system for collecting, managing and assuring the quality of evidence.
Phase 03
Analytics & Inference
Collected data turned into defensible evidence, with every finding rated for strength.
Phase 04
Monitoring, Learning & Reporting
Findings converted into systems that sustain learning and accountability over time.
Our program
The process is the program.
Reports and communication support are deliverables, but the program team's active part in pursuing inquiry, scoping analysis and adopting new methods is central to both the ethics of the work and whether it achieves anything.
The implementation support professional coordinates the evaluation, shepherds findings into program changes, and helps build a habit of recursive improvement inside the team. The organization's staff co-create the frameworks, collection methods, pilots and data architecture alongside them. Deeper involvement in the technical work — analysis, research design, reporting — is there for teams that want it and have the time.
The sequence runs in one direction until it reaches recursion and integration, and then it loops: findings become program changes, outcomes feed the next round of evaluation, and the cycle repeats as often as the program needs before graduation.
-
Step 01
Intake
A scoping conversation and intake assessment establish whether the program is a fit, and what already exists to build on.
-
Step 02
Embed
The implementation support professional joins the program team, and learning goals and timelines are agreed together.
-
Step 03
Audit
A structured review of the frameworks, data and systems already in place, so nothing that works gets rebuilt.
-
Step 04
Evaluation
Phases 1 to 3 of the pipeline, co-created with the organization's staff. Phase 2 is the heaviest lift on both sides.
-
Step 05
Communication
What the evaluation found, turned into material for the community, the board and the funders.
-
Step 06
Recursion and integration
Findings become program changes, outcomes feed the next evaluation cycle, and the loop runs as often as it needs to.
-
Step 07
Graduation
After two years, embedded support draws back under a written handoff plan agreed with the program team.
-
Step 08
Onward support
Continue on a discounted monitoring retainer or take a written implementation guide. Our network and literature briefs stay open either way.
Phase 01
Evaluative Foundation
What is this program trying to do, for whom, and why do we believe it will work?
This phase builds the conceptual infrastructure that makes rigorous evaluation possible. Without it, data collection has no direction and findings have nothing to be interpreted against. It is also where the program team starts to think in evaluative terms — the habit of inquiry everything after it depends on.
What Phase 1 produces
- Stakeholder map. Everyone with a stake in the program, distinguishing those whose input shapes the evaluation from those whose outcomes are its subject.
- Program description and context brief. What the program does, for whom, at what scale — set against its evidence base, policy landscape, and peer field.
- Logic model. The program's causal architecture from inputs to long-term outcomes, built with the organization's staff so it reflects implementation reality rather than aspirational design.
- Theory of change. The problem theory, intervention theory, change theory, and equity dimensions, with literature cited for causal claims.
- Evaluability assessment. A structured judgment of whether the program is actually ready to be evaluated at all.
- Evaluation questions and indicator framework. A prioritized set of questions, each matched to measurable constructs, marking what is in and out of scope.
Phase 02
Data Architecture
What data do we need, from whom, and how do we know it is good enough?
This phase turns evaluative questions into a practical, defensible system for collecting, managing and assuring the quality of evidence. It is the heaviest phase for the program team and for ours, because it can mean changing established practices and systems. An organization that enters here with infrastructure already in place, we begin with a data audit instead — a structured review of its instruments, data quality and system architecture against the indicator framework.
What Phase 2 produces
- Data needs assessment. Every indicator mapped to a source and categorized: already collected, collectible with new instruments, derivable from third-party data, or practically infeasible.
- Measurement selection report. For each indicator, the measure chosen, the rationale, the population, and the timing.
- Instrument suite. Designed and adapted instruments, with novel ones pilot tested and reviewed by community representatives before field deployment.
- Collection protocol. Administration, consent and assent, data entry, confidentiality, and how missing data and non-response are handled.
- Data management system. Storage, linkage, backup, security and retention — set up alongside the program team, with the training to run it.
- Data quality plan. Standing completeness, validity, consistency and timeliness checks: who runs them, how often, and what each problem triggers.
Phase 03
Analytics & Inference
What do the data tell us, with what confidence, about whether and how the program works?
This phase turns collected data into defensible evidence. The final analysis runs offsite; the findings come back to the program team through the implementation support professional rather than as a report in an inbox. An organization that enters here with data already collected gets a suitability review first and a feasibility memo specifying which of its questions the existing data can actually answer.
What Phase 3 produces
- Design and analysis specification. Counterfactual strategy, unit of analysis, level of inference, subgroup analyses planned in advance, and threats to validity with mitigations — specified before results are examined.
- Quantitative findings memo. Effect sizes with confidence intervals rather than p-values alone, sensitivity analyses, and subgroup findings, with plain-language interpretation alongside the technical specification.
- Qualitative findings memo. Analytic approach, codebook development, triangulation procedures, and key themes with representative evidence.
- Contribution and attribution analysis. What else was operating, whether the pattern of findings fits the theory of change, and what alternative explanations remain. Required for every non-experimental evaluation.
- Evidence strength summary. One non-technical page on what the evidence shows, at what confidence, with what caveats.
- Integrated findings report. The master document: executive summary, full methods, findings by question, implications for improvement, and limitations.
Phase 04
Monitoring, Learning & Reporting
How do we know the program keeps working, and how do we tell people what we know?
This phase converts findings into systems that sustain organizational learning and external accountability over time. It treats evaluation not as an event but as an embedded practice, and it is deliberately built with few internal dependencies, so monitoring and reporting work can move quickly. In our program, Phase 4 begins at graduation: continue with us on a discounted retainer, or take a written implementation guide and run it independently.
What Phase 4 produces
- Learning agenda. The questions the organization intends to investigate next, sorted by what routine monitoring can answer, what needs a new evaluation, and what requires external research.
- Monitoring and evaluation plan. Which indicators are tracked, how often, from what sources, against what targets — with procedures for updating them as programs evolve.
- Indicator dashboard. Built on the organization's preferred platform — live, automated or periodic, depending on the data architecture from Phase 2 — with views and access controls tailored to program staff, leadership, board and funders.
- Reporting suite. Funder reports, board dashboards, operational reports, public impact summaries, and policy briefs translating evidence into implications for design, policy or funding.
- Dissemination plan. How findings reach peers, funders, policymakers and the communities whose members were the subjects of the evaluation — audiences, messages, channels, timing.
Phase 4 is the last phase, not the end. The learning agenda and monitoring data feed back into Phase 1 whenever a program is refined, expanded or replicated, which makes evaluation a cycle rather than a line ending in a report.