Five program metrics worth tracking from day one

Program dashboards tend to fill up with numbers that are easy to collect rather than numbers that change decisions. Total applications is the classic: it goes up, everyone nods, and nobody does anything differently.

These five are the ones worth instrumenting before your first cohort, because each one has an obvious action attached when it moves.

1. Completion rate, by question

Not how many people applied — how many who started, finished, and where the rest stopped. Drop-off clusters at specific questions: the one that needs a document, the one that is ambiguous, the one that asks for something people would rather not disclose.

What it changes: which question you rewrite or remove before the next round. A single badly-worded eligibility question can cost a fifth of your pipeline, and it is invisible in the application total.

2. Time to first response

The gap between a submission arriving and a human acknowledging it — and separately, the gap to a decision. Track the median and the worst case, because the worst case is what gets talked about.

What it changes: staffing and sequencing during review weeks. It is also the single best predictor of whether people reapply or recommend the program.

3. Participation depth, not attendance

Attendance says someone showed up. Depth says whether the program is working: submissions handed in, sessions attended out of those offered, milestones reached. Measure it per participant, not as a total, so you can see the distribution.

What it changes: who gets a check-in. A cohort average of 80% hides the fact that most people are at 95% and six are at 20% — and those six are the ones who will quietly drop out.

4. Reviewer load and spread

How many items each reviewer handled, how long they took, and how their scores compare to the mean. The last part is uncomfortable and genuinely useful: reviewers drift, and a consistently harsh or generous reviewer distorts outcomes in a way nobody notices.

What it changes: assignment in the next round, and whether your criteria need rewriting. If two reviewers disagree systematically, the rubric is ambiguous, not the applicants.

5. One outcome you promised

Every program makes a claim: participants get funded, hired, certified, placed, renewed. Pick the one you actually promise and track it per cohort, even if the number is small and takes a year to arrive.

What it changes: everything about how you talk to funders and applicants. It is also the only metric on this list that tells you whether the program should continue in its current shape.

Two rules for the dashboard itself

• Every metric needs an owner and an action. If nobody can say what they would do differently when it moves, delete it. It is costing attention and lending false precision.
• Prefer distributions to averages. Program cohorts are small enough that averages hide the individuals — and the individuals are the program.

What to leave off

Page views, form opens, email open rates, and total applications across all time. They are real numbers that measure interest rather than the program, they rise on their own as you grow, and they create a pleasant feeling of progress that survives a bad cohort.

Start with completion, response time, depth, reviewer spread and one outcome. Five numbers, each with something you would actually do about it. You can add more once those are boring.

← All articles