SUMMARY: A completion rate is a statement about a defined population, not automatically about everyone affected by a service. Before interpreting the percentage, ask what qualifies for entry into the count and what remains outside it.
Start With the Unit
An attempt, a request, a person, and a completed case are different units. One person may submit several attempts for one request. A single case may contain several decisions. Combining these units can make a report look precise while obscuring the experience it claims to describe.
In the fictional Empty Denominator, a statement about completed reviews is presented as a statement about the service. The arithmetic need not be wrong for that interpretation to be too broad.
Keep the Reporting Window Consistent
Suppose a synthetic exercise starts with 100 eligible requests received in one week. By the review date, 60 are completed, 25 remain open, ten were withdrawn, and five have unknown outcomes. A completion measure for that original group is 60 out of 100 at that date. That does not establish that the other 40 failed.
By contrast, counting only completed cases and asking what proportion are complete produces 60 out of 60. It answers a different, largely uninformative question. Neither count should quietly include older completions while excluding new arrivals.
Expose Exclusions and Unknowns
Publish the definition, observation date, and exclusion rules alongside the result. Keep pending, withdrawn, duplicate, and unknown states distinguishable. A withdrawal may reflect resolution elsewhere, frustration, or a change of need; its meaning requires evidence beyond the label.
More collection is not automatically better. Use bounded identifiers and aggregate counts where they suffice. Do not gather extra personal information merely to make a dashboard appear complete.
Test the Claim Readers Will Hear
Ask a reviewer to paraphrase the headline, then compare that interpretation with the actual counting rule. The completion-metric audit makes this a synthetic exercise. The accountability guide places measurement beside explanation, ownership, and verified outcomes.