Measure the contribution, not only the activity

Start with an endorsed community priority and ask what improvement would look like in members’ experience. Then identify the contribution the Nation’s government or service organization is expected to make. A broad goal such as improved access to learning may involve transportation, scheduling, teaching capacity and family circumstances. No single activity count can represent all of those conditions.

Arcus recommends a small set of connected measures for each material priority: an outcome members value, evidence of the administration’s contribution, and a delivery measure that gives early warning when implementation is off course. Add relevant context rather than expecting one score to do all three jobs.

For example, the number of sessions delivered is a delivery measure. The proportion of intended participants who can attend at a suitable time may indicate access. Whether participants can use what they learned concerns a different outcome and may require a different method of assessment. These are illustrative categories, not prescribed indicators for a Nation’s language, education or cultural programs.

Agree on the interpretation before publishing the indicator. A measure needs a definition, population, time period, source, responsible owner, known limitations and a decision it is intended to support. Without those elements, a dashboard can present precise numbers with ambiguous meaning.

Distinguish four explanations for apparent progress

The service may have improved. Faster response, more reliable delivery or fewer unresolved cases can reflect better processes. Test this explanation against staffing, workflow changes and members’ experience. A plausible mechanism makes the interpretation stronger, although it does not by itself prove causation.

The people counted may have changed. A program can appear more successful if people facing greater barriers stop participating or are excluded from the denominator. Define who is eligible, who was reached and who is missing. Assess differences only where the Nation authorizes the use and the information can be reported without exposing people.

The recording method may have changed. A new case-management system, revised definition or cleaner records can shift the indicator without changing the underlying service. Mark the change in the series. Where comparison is no longer sound, establish a new baseline rather than splice the numbers together as though nothing changed.

External conditions may have changed. Employment, housing costs, transportation disruption or other systems can affect outcomes. The administration should explain what it can influence, what it can only monitor and where partners carry responsibilities. This protects against both overstating success and blaming a department for conditions beyond its control.

A useful review asks which explanation best fits the evidence and what additional information would distinguish them. It does not immediately turn a red indicator into a performance judgment or a green one into a claim of impact.

A worked example: the percentage improves while need grows

Suppose a hypothetical service records 80 requests in one quarter and resolves 60 within its service standard. Its timely-resolution rate is 75%. In the next quarter it records 120 requests and resolves 96 within the standard, producing a rate of 80%.

The rate improved by five percentage points, and the number resolved on time rose by 36. Yet the number not resolved within the standard rose from 20 to 24. The service may be performing better under increased demand while more people still experience delay. Reporting only the percentage would conceal that operational pressure; reporting only the count of delayed requests would conceal the improvement in response.

Before attributing the result to a new process, check whether the request types, start and stop dates, treatment of incomplete requests and recording practice stayed consistent. Also ask whether people unable to access the service are absent from the data. The example is a decision exercise, not a finding about a real Nation.

A proportionate response might be to investigate the longer delays, review capacity for the most complex requests and speak with members through an approved feedback process. The evidence does not automatically justify a new permanent position or establish that the service has failed. It identifies the next questions and possible actions.

Make data governance operational

Before collecting new information, establish who approves the purpose, who can access identifiable records, where information will be held, how long it will be retained and who can authorize sharing. Use the Nation’s own protocols and appropriate expertise to address privacy, consent, collective interests and any relevant agreements.

Separate the information needed for service delivery from the information appropriate for Council reporting and public communication. A public report rarely needs the detail in an individual case record. Small counts, combinations of characteristics and quotations can identify people even when names are removed. Set reporting and review rules before assembling the public dashboard.

Do not claim that a product or consultant is “OCAP-compliant” merely because it uses the acronym or has completed training. FNIGC notes that training is not an endorsement of someone’s work.³ The practical question is whether the actual collection, access, use and sharing arrangements reflect the Nation’s decisions and responsibilities.

Data governance is also an operating responsibility. Assign a steward for definitions, quality issues, authorized access and changes to reporting. A staff departure should not leave the Nation unable to interpret or retrieve its own information.

Cost the measurement system before expanding it

Initial work includes agreeing on definitions, checking existing records, developing collection methods, establishing permissions and training staff. Recurring costs include data entry, cleaning, interpretation, community engagement, reporting and secure administration. Software licences are only one component.

For a hypothetical workload estimate, six indicators requiring two staff-hours each month need twelve hours before interpretation and review. If preparation for the quarterly discussion takes another twelve hours, that adds four hours per month on average. The resulting sixteen-hour monthly estimate remains incomplete unless it includes any community collection and governance work. Use local estimates rather than this illustration to approve resources.

Compare a simple maintained register, an enhancement to an existing system and a new reporting platform. A new platform may improve consistency when multiple departments share information, but it can also add configuration, access management and support obligations. Choose it only after the information model and responsibilities are clear.

Reassess indicators that are expensive to collect but rarely change a decision. Retirement should be deliberate: preserve the history and explain why the measure is being replaced. Changing an inconvenient measure solely to improve the appearance of performance undermines trust.

Start with a small, useful measurement cycle

Select two or three material community priorities for an initial cycle. Confirm the intended outcomes through the Nation’s established processes and map the administration’s contribution. Review existing information before commissioning new collection. Draft the indicator definitions and governance arrangements together.

Run one reporting cycle and ask the people using the report what they can decide from it. Test whether the numbers are understandable, whether missing information is visible and whether the report reflects members’ experience. Record decisions and follow-up responsibilities so the next review can examine whether the response worked.

Expand only when the method is maintained reliably. A credible report can say that evidence is incomplete, a measure is still being developed or a result cannot yet be attributed to an initiative. Those statements support better decisions when accompanied by a clear plan to resolve uncertainty. The objective is learning and accountable action in service of community priorities.

Evidence and source notes

1. Indigenous Services Canada, About the Community Well-Being index. Documents scope, population and interpretation limits; the index does not capture every community priority.
2. First Nations Health Authority, Update on health indicators of First Nations Peoples in BC, August 22, 2024. Describes the Population Health and Wellness Agenda and connection-to-land indicator work. Used as a measurement-design example, not a current health assessment.
3. First Nations Information Governance Centre, The First Nations Principles of OCAP®. OCAP® is a registered trademark of the First Nations Information Governance Centre (FNIGC).

Research reviewed October 18, 2026. Numerical examples are hypothetical. The six-test framework and implementation method are original Arcus analysis; indicator definitions and information governance require the Nation’s own decisions.