The difference between a status report and a data-driven status report is not the presence of numbers. It is the relationship between the numbers and the conclusion. Numbers that appear after the conclusion are decoration. Numbers that produce the conclusion are reporting.
At Intuit, I built compliance dashboards for TurboTax and Credit Karma that ran on actual issue data: volume, age, severity, closure rate. Leadership could see whether the program was improving without asking me to interpret it. That is the target. A report that requires the author to explain what the numbers mean is a report the author controls. A report the reader can interpret independently is one the author has built correctly.
The metrics that belong in a program report are the ones that lead the outcome, not the ones that describe it after the fact. Burndown rate, dependency closure rate, open risk count with aging, decision latency. Those tell you where the program will be in four weeks. Completed tasks and milestones met tell you where it was.
The hardest part of data-driven reporting is committing to the metric before you know what it will say. Teams that define their success metrics after the work is done select metrics that make the work look good. Define the metrics at the start, report them consistently, and resist the urge to swap them when they are unfavorable.
Good program data is not about surveillance. It is about giving everyone the same view of reality so the conversations about what to do are about the program, not about competing interpretations of how the program is going.