A number that helps someone get help
Imagine a service report that makes a missed household visible, accepts the resident’s account and starts a funded response. Measurement can become a shared instrument of care: an early warning, a memory of promises and a way to learn whether life has improved.
A useful measure can be an act of attention
Imagine a public water team noticing that one neighbourhood reports a longer collection journey each week. An accessible report could help residents explain what has changed, help an operator locate the fault and help a community choose an intervention. Later, it could show whether people actually recovered that time. This is a possible use of measurement, not a claim that more data always produces better services. The appealing prospect is that knowledge becomes shared: people do not have to prove the same problem from the beginning whenever a different office answers the telephone. A good record can preserve the institution’s memory of a person’s need.
Understand what the number stands for
Manheim and Garrabrant distinguish several ways a proxy can fail as pressure is placed on it. Their analysis is a reason to investigate the relationship between a measure and its purpose. Consider our hypothetical team counting completed service tickets. That might help manage work, while telling residents little about whether water is available. A quick administrative closure and a successful repair could both add one to the total. The remedy begins with a clear description: what is counted, whom it covers and what conclusion it can support. A modest measure with an honest boundary can be more useful than a universal claim nobody can inspect.
Make important work easier to recognise
SourceCred’s contribution graph shows how recorded work and relationships can support attribution; its adjustable weights make design choices explicit. DisCO’s manifesto brings paid work, commons production and care into the conversation about value. Together they invite a practical experiment in our imagined service team: recognise diagnosis, maintenance, documentation and support, alongside the final repair. The choice of categories is still a human judgement. Some care should be acknowledged through time, staffing or shared responsibility rather than turned into a competitive score. The opportunity is to ask people what sustains good work and give that work a place in the institution’s attention.
Combine instruments with people’s accounts
DECODE’s pilots explored sharing citizen-generated and sensor data with privacy protections and control over its use. That provides one source of inspiration for a shared evidence process. In our example, a sensor might report pressure while a resident describes an inaccessible collection point. These observations answer different questions and can both matter. Reports should identify what is known, what is disputed and where coverage is missing. Residents should be able to contribute without exposing unnecessary personal information. A missing instrument need not become a missing person, and uncertainty can be an invitation to investigate rather than a reason to close the case.
Inside the Commission
The Basin 06 continuity figure accurately describes scheduled delivery intervals for registered connections. Its apparent success becomes misleading when readers treat it as a statement about everyone's access to water. Three hundred thousand residents never enter the denominator. The Commission can maintain an excellent result while their situation worsens. This is also a problem of institutional jurisdiction: no provider owns the omitted population's shortfall, so nobody sponsors the measure that would expose it. The missing number reflects a missing responsibility.
Give new evidence somewhere to go
Meadows’ discussion of information, rules and goals helps explain why another dashboard may be insufficient. Our proposed service process would give a verified omission a destination: a responsible team, a response deadline and access to a remedy budget. A resident could ask for review if the evidence category itself excludes their experience. The team could publish the reason for declining a request, and another body could reconsider it. These are suggested institutional arrangements, not features established by the essay. They connect seeing with doing. A worse headline after including omitted residents may then mark the beginning of a more capable service.
Use a learning cycle people can join
Begin with one familiar indicator and ask three people affected by the service what it misses. Publish an additional view, explain any uncertainty and choose one improvement to test. Return to the same people after the work and ask whether the promised benefit arrived. This suggested exercise is small enough for a community project and serious enough to change its priorities. In the laboratory, change the population boundary and the duty to assess and respond to a resident’s report. Distinguish that duty from evidence that a service has actually improved. The aim is an institution that can become better informed and more useful over time, with residents participating in its understanding of success.
What could your institution learn if an omitted person could change both the evidence and the response?
A fuller picture. A route to respond.
Compare service continuity with population coverage. Give a resident’s report an assessment and response duty, and distinguish that obligation from a completed remedy.
Open this experiment