The common failure is not analytical. It is producing a correct finding that nobody acts on, repeatedly, until the reporting is ignored.
Why findings stall
They point at something expensive, so they are noted and deferred.
They point at another department's decision.
No owner, so the finding sits in a report.
Too many at once, so none is pursued.
No mechanism to turn a finding into a change, because the programme was scoped as reporting.
The finding is disputed on grounds the analysis cannot settle, usually data quality.
The sequence that works
One finding at a time, the largest.
Verified with the people who do the work, which settles the data quality dispute before it starts and frequently improves the finding.
Costed, roughly, in both directions: what the loss is worth and what the change would cost.
Owned by someone who can make the change, named.
With a date.
Re-measured afterwards, the same way, which is the step that establishes whether the programme is worth continuing.
Reported, including when the change did not work.
Choosing the first one
Visible, so people see that measurement leads somewhere.
Cheap, so it does not need a capital case.
Owned by the operation itself, not by another department.
Uncontroversial, which means not about individual performance.
Wave timing, break staggering, slotting the top hundred items, or a parking position change are the usual candidates, and all are free.
Verifying with the floor
The step that is skipped and that determines whether the finding is real.
Take the finding to the people in that area.
Ask whether it matches what they see.
Ask what the analysis does not know, which is where the data quality problems surface.
Ask what they would change, which frequently produces a better answer than the analysis.
A finding that surprises the people who work there is usually a data problem; one that confirms what they have been saying is usually correct and has been ignored.
Reporting back
What we found, what we changed, what happened.
Including where nothing changed and why, which is the part that preserves credibility.
To the people in the area first, before the wider report.
Regularly, on a schedule, so the programme is a habit rather than a project.
A monitoring programme that has never reported a change it made is a reporting exercise, and everyone involved knows it.
The re-measurement
Same measure, same method, same period length.
After the settling period, not immediately.
Reported honestly, including a null result.
A programme that only reports successful interventions is not measuring; it is selecting, and the selection will be noticed.
What sustains it
A standing slot in the operational review, not a separate meeting.
A named owner with allocated time.
A short list of open findings, with owners and dates, reviewed.
Closure reported, which is what makes the next finding get attention.
The open findings list
The artefact that keeps a programme from becoming a reporting exercise.
One row per finding: what, the loss quantified, the owner, the date, the status.
Short. More than about ten open items means none is being pursued.
Reviewed in the operational meeting, not in a separate forum.
Closed only after re-measurement, not after the change was made.
Overdue items escalated, which is what produces the resourcing.
Published, including the ones that were closed with no effect.
Reporting a null result
The discipline that makes the positive results believable.
We changed X, we re-measured, nothing moved.
Say why you think that is, which is usually that the change addressed a non-constraint.
Say what you will try next.
Do not quietly drop it, which is what usually happens and which is noticed.
A programme that has never reported a failed intervention is selecting its results, and everyone reading it works that out eventually.