What is an impact dashboard?
An impact dashboard presents evidence about the changes a program or portfolio aims to support. It brings together outcome measures, trends, relevant comparisons and information about the coverage and quality of the evidence. Teams use it to review progress, investigate questions and communicate findings.
The word “impact” does not make every displayed result proof of impact. A dashboard can show a change in skills or employment, for example, without establishing that the program caused that change. The evaluation design and supporting evidence determine which conclusions are justified.
A useful dashboard makes that distinction visible. Readers should be able to tell what was delivered, what was observed, who is represented and what remains uncertain. They should also see how the findings will inform the next decision.
Separate delivery, outcomes and evidence coverage
Scroll horizontally to see all columns →
| Layer | Example measure | Question it answers |
|---|---|---|
| Delivery | Sessions delivered or participants attending | What work took place and who took part? |
| Outcome | Demonstrated skill change or reported use of a skill | What changed or was achieved? |
| Coverage | Usable follow-up responses among those eligible | How much of the relevant group does the evidence represent? |
| Interpretation | Reviewed findings, contextual evidence and limitations | What can reasonably be concluded? |
| Action | Improvement owner, next step and review date | What will the team do with the finding? |
Delivery measures still matter. A change in outcomes can be difficult to interpret without knowing whether the activity took place as planned. Keep them alongside outcome evidence, with clear labels, rather than presenting attendance as an outcome or discarding it altogether.
Impact dashboard examples
Training and workforce programs
Show participation, assessment results, application of learning and relevant employment outcomes. Include the time since participation, matched-record coverage where change is calculated, and the basis of each rate. A placement measure needs an explicit eligible group and a defined follow-up window.
Youth development and education
Present progress against suitable measures and the context needed to interpret it. Student, parent and practitioner perspectives may answer different questions. Keep those sources distinguishable rather than treating several observations about one person as several independent participants.
Grant and investment portfolios
Bring together comparable measures across organizations while preserving differences in their work. A shared outcome category does not automatically justify adding every result into one total. Keep measurement methods, periods and reporting coverage available for review.
Membership and networks
Connect member activity, participation and reported benefits over time. Separate the experience of a member organization from the experiences of the individuals it represents. Local chapters can contribute different evidence while maintaining an agreed core for necessary comparisons.
For operational delivery, use the program dashboard guide. For the wider leadership view, including organizational resources and fundraising, see nonprofit dashboards.
A worked example: looking behind an outcome rate
Consider a fictional training program operating at two sites. Each site has 50 participants eligible for follow-up. Site A receives 40 usable responses, of which 28 report applying the skill. Site B receives 20 usable responses, of which 16 report applying it.
Scroll horizontally to see all columns →
| Measure | Site A | Site B | Combined |
|---|---|---|---|
| Eligible for follow-up | 50 | 50 | 100 |
| Usable responses | 40 | 20 | 60 |
| Follow-up coverage | 80% | 40% | 60% |
| Reported application | 28 | 16 | 44 |
| Application among respondents | 70% | 80% | 73.3% |
Site B has a higher reported application rate but lower coverage. The dashboard should not automatically rank it as the stronger site. Outcomes for the nonrespondents are unknown, and the sites may differ in participant circumstances or opportunities to use the skill.
The combined respondent rate is 44 divided by 60, or 73.3%. It is not the simple average of the two percentages. Nor does it show that 73.3% of all eligible participants applied the skill.
Suppose some comments at Site B mention difficulty attending follow-up sessions. That finding can guide an investigation into participation and collection arrangements. It does not demonstrate what happened to every missing respondent, or prove why the outcome rates differ.
A practical next step is to review the follow-up process at Site B, check whether respondents differ from the eligible group using appropriate available context, and agree whether more evidence is needed. Record who owns that work and when the team will review it.
How to build an impact dashboard
1. Define the decision and intended change
Start with the decision the evidence must support. Is the team considering a program change, reviewing progress across sites or preparing a funder update? Then specify the outcome, population and time period. Broad goals such as “improve lives” need a more concrete question before they can become useful measures.
2. Choose measures and appropriate evidence
Decide what would count as relevant evidence of the change. An assessment, service record, survey or interview can each contribute, depending on the question. State whether a measure is observed, self-reported or assessed by someone else. More indicators are not a substitute for choosing measures that fit the intended outcome.
3. Plan collection and record relationships
Map which people or organizations contribute, what they provide and when. Use a stable identifier where authorized matching over time is needed. Group-level or anonymous evidence can also be useful; do not collect identity merely to make a chart clickable.
Keep source references, dates and reporting periods with documents and responses. A narrative from last year may provide context, but it should not silently become evidence for the current reporting cycle.
4. Define calculations and comparisons
Create a data dictionary covering units, denominators, eligibility, exclusions, missing values and versions. If you calculate pre/post change, use an appropriate matched group and show its coverage. If you compare sites, examine whether instruments, populations and periods are sufficiently compatible.
5. Design the display around interpretation
Show the main result with its base count, period and coverage. Put context and limitations close to the number rather than hiding them in a distant appendix. Use trends for comparable observations over time and tables where a precise definition or action is more useful than a chart.
6. Test and use the view
Recalculate selected figures, check joins and review a sample of source evidence. Test missing data, duplicate submissions and late corrections. Then use the dashboard in a real discussion: does it help the team decide what to do, or does it simply produce more questions nobody owns?
Compare different programs without forcing one survey
Federated networks and multi-site programs often need local flexibility. Agree on the few common fields and definitions required for the comparisons you intend to make. Local instruments can include additional questions and different workflows around that core.
Stable contact or registration information can be collected once where practical. Update changing context, such as location or employment, when it matters. Preserve the historical context used in earlier reporting so a later change does not rewrite the meaning of past results.
Use mappings only when the underlying meaning is compatible. “People reached,” “visits” and “organizations supported” should not be combined as though they were one unit. For portfolios with different outcomes, a grouped view of evidence and progress may be more honest than one headline total.
Bring explanations into view without overstating them
Comments, interviews and documents can show experiences that a numerical measure misses. Present the question asked, whose perspective is represented and how the material was reviewed. Include important contradictory evidence rather than choosing only quotations that support the preferred story.
Where ratings and comments were collected together, their connection can help interpretation. Where interviews involved a separate purposive sample, explain that relationship. Do not present interview themes as percentages of the full program population unless the design supports that calculation.
AI can assist organization and review, but a generated theme is not an established cause. Keep the source available, review consequential interpretations and distinguish suggested analysis from approved reporting. The mixed-methods analysis guide explains how different kinds of evidence can be brought together.
Keep the dashboard usable and governed
Set access by purpose. Reviewers may need source detail while leadership or public audiences receive appropriate summaries. Small groups, distinctive comments and combinations of filters can reveal more than a headline count suggests.
Match the refresh schedule to the evidence and the decision. Immediate updates may help some operational questions; a reviewed quarterly snapshot may fit a reporting commitment. Display update dates and approval status. Faster data is not automatically more reliable evidence.
Keep a dated snapshot and the definitions used for issued reports. Document corrections and instrument changes so reviewers can understand why an earlier figure differs from the latest view.
Evaluate the workflow beneath the charts
Spreadsheets and BI tools can support useful impact dashboards. The right choice depends on the sources, analysis requirements, maintenance capacity and access controls your team needs. Do not assume a charting tool cannot display qualitative context or connect records when appropriately configured.
Sopact's approach focuses on recurring collection, connected evidence, reviewed analysis and governance. Test how those capabilities support the specific outcome workflow: local forms, common definitions, repeated reporting, corrections and approved summaries. Verify the actual integrations and analysis needed in a pilot.
Consider the ongoing effort to prepare and maintain the evidence, not just the initial visual design. The team should be able to explain a result and update its process without rebuilding the entire reporting chain each time.
Watch: outputs and outcomes
This companion introduces the distinction behind many dashboard measures. Use it to review whether a proposed indicator describes activity, an outcome or evidence of change.
Use the dashboard as the basis for reporting
An impact report adds the narrative around the displayed evidence: what happened, what can be concluded, what remains uncertain and what will change. Keep the report and dashboard aligned on their reporting period and approved definitions.
Continue with the impact reporting guide, read How to Write an Impact Report, or browse report examples for ways to present the findings.
Frequently asked questions
What should an impact dashboard show?
Show relevant outcome evidence, delivery context, reporting periods, coverage and limitations. Add meaningful comparisons and a clear next action where the dashboard supports program improvement.
Can an impact dashboard prove causation?
Not by itself. It can present findings from an evaluation, but the underlying design and evidence determine whether a causal claim is justified. A trend or participant explanation alone does not establish attribution.
Should every measure link to a person's record?
No. Use appropriate, authorized record linking where the question requires it. Anonymous and aggregate evidence can be valuable. Source traceability can lead to a batch, document or calculation without revealing a person.
Can different sites keep different surveys?
Yes. Agree on the small common core needed for comparison and document local variations. Separate measures that cannot be mapped to compatible definitions, units or periods.
How is an impact dashboard different from a report?
A dashboard organizes selected evidence for review, often with filters or trends. A report develops the interpretation and decisions in a narrative. Both should identify their period, sources, definitions and limitations.
Choose your next step
King Center
Related practice: using program feedback to support review. The fictional dashboard example is not a result from this customer story.
Read the customer story →