You know the change you intend, but you have no real way to see whether it is happening. A defined outcome with no measure behind it is still just a hope written more clearly.
This work builds the measurement system: it turns each intended change into a small number of honest indicators, decides how and from whom you will gather the evidence including the voices of the people you serve, and sets up a way to collect it that fits inside the work rather than crushing the people doing it or the people receiving it. It is not choosing a database or a software platform; the technology and data plumbing live in operations and systems, and this work will send you there when that is the real need. It is not academic research, and it does not require you to become a scientist. Use this if you have defined the change but have no real way to measure it. If you have not yet defined the change, go to Define What Success Means and Build the Logic Model first. And if collecting your evidence will involve sensitive or personal information or vulnerable people, settle the consent and privacy requirements before you build the instruments.
An outcome is a change; an indicator is the specific, observable sign that the change is happening. For each main outcome, choose a few observable signs that would honestly show it is happening, and reject the ones that are merely easy to count but do not really capture the change. Then check each: if you improved this number without the real change happening, would you be fooling yourself? Drop or fix any indicator that could be gamed or that rewards the wrong thing.
Open the Outcome to Indicator Translator →Every indicator needs a source and a method: who you learn it from, how you gather it, and when. Make those choices honestly, weighing the rigor you want against the effort you can sustain, choosing the most honest method you can actually keep up, and putting the voices of the people you serve at the center rather than the edge as a direct source of evidence about their own lives. Where you choose a lighter method than the ideal, write down what it can and cannot tell you, so you never overclaim later.
If an indicator would require personal, sensitive, or health information, or gathering from children or people in crisis, the consent, privacy, and safeguarding requirements are fixed and sit outside this framework. Meet them, and where you are unsure what they require, bring in the relevant professional before you gather anything. This is a route to a professional, not a stop.
Methods become real only as actual instruments, a survey, a form, an interview guide, a simple record, and an actual routine for using them. Build the actual instruments for each method, as simple as they can be while still capturing the indicator honestly, and test each with a real respondent or user before you rely on it. Then decide who collects what, when, and where it goes, so collection is a defined routine and not a scramble at reporting time.
If building the instruments reveals you need technology you do not have to hold and manage the data, that is operations and systems. Design the instruments and what you need to capture here, then route the technology build there, so the system serves your measurement rather than dictating it. This is a route, not a stop.
A measurement system that overburdens the people delivering the program, or the people receiving it, will be abandoned or resented, and often both. Look at what collection asks of your staff and of the people you serve, and cut anything whose cost in time, intrusion, or dignity is greater than what it will teach you, protecting the service and the relationship first. Then fold the trimmed collection into the natural flow of the program, into intake, sessions, and exit, so it is part of the work rather than an extra task bolted on.
Open the Collection into Delivery Plan →Building a measurement system is the point where data ethics stops being abstract. The moment you gather personal, sensitive, or health information, or ask anything of children or people in crisis, you take on real duties of consent, privacy, and protection that are set by law and professional standards, not by you, and a survey or a data store built carelessly can expose the very people you serve to harm. Where your measurement touches any of this, meet the requirements as fixed points and bring in the relevant professional, a data-protection or safeguarding specialist or an attorney, where you are unsure. This framework helps you design honest measurement; it cannot certify that your data practice is lawful and safe, and that is not optional.
You can point to a handful of indicators that genuinely stand for the changes you intend, not a pile of activity counts. You have simple instruments that a busy staff member can actually use, and a routine that says who gathers what and when. The people you serve are a direct source of evidence about their own lives. And collection fits into the work rather than fighting it.
With the measurement system built, your plan most likely sends you next to Collect and Make Honest Sense of the Evidence. Where building the system surfaced a real need for a database or technology, that routes to operations and systems; where it surfaced a consent or privacy question, that routes to the relevant professional.