Inspection coverage
Inspection coverage explained: what it measures and how to calculate it
Inspection coverage is the share of the required inspection population that was actually inspected to standard within a defined period, expressed as a percentage. It answers one question: of everything that had to be examined this period, how much of it was. The inspection backlog counts what was missed; coverage expresses the same shortfall as a rate.
How is inspection coverage calculated?
Coverage percent = (inspections completed to standard within the period / inspections required within the period) x 100.
Both halves of that fraction get abused, and in opposite directions. The numerator quietly swells with inspections that were logged but whose record would not survive a review. The denominator quietly shrinks into the number of inspections that were scheduled, rather than the number that were required.
The denominator substitution is the common one and it is close to accidental fraud. If 240 inspections were due and only 190 ever made it onto the plan, reporting 168 completed against 190 scheduled gives 88 percent. Against the 240 actually required, it is 70 percent. Same quarter, same work, and one of those two numbers is the one that reaches a board pack. Coverage has to be measured against the required population, which means it depends on the scheduling process being honest about what it left out.
Inspection coverage explained: a short staffed quarter
An estate of 480 assets on a six month interval generates 960 inspections a year, or 240 in a quarter. A three inspector team lost one member for six weeks and completed 168.
Overall coverage: 168 divided by 240, which is 70 percent. Split by duty class it reads differently. Of 60 statutory examinations due, 51 were completed: 85 percent statutory coverage. Of 180 discretionary condition checks, 117 were completed: 65 percent.
That split is the only version worth putting in front of anyone. It shows the team triaged correctly under pressure, which the blended 70 percent hides. It also shows that 9 statutory examinations still ran past due, which the blended 70 percent hides just as effectively in the other direction.
Coverage, completion and compliance are three different measurements
They get used as if they were one, and they answer separate questions.
- Completion rate asks whether the visits that were scheduled actually happened.
- Coverage asks what share of the required population was reached in the period.
- Compliance asks whether every asset carrying a duty currently holds a valid, in date inspection.
You can hold 95 percent coverage and be non compliant, if the missing 5 percent all carry statutory duties. You can hold 100 percent completion against a schedule that was never big enough to start with. Compliance is binary per asset and unforgiving. Coverage is a throughput measure for managers. Reporting only one of the three is how organisations get surprised by their own data.
What counts as inspected?
A defensible rule: an inspection counts towards coverage when the record identifies the asset, the date, the inspector, what was examined and the outcome, in a form a third party could read a year later without ringing anyone. Anything short of that was a visit, not an inspection.
Apply that rule and coverage usually drops on first measurement, sometimes by several points. The drop is not a decline in performance. It is the first accurate reading you have had, and it means the earlier numbers were the wrong ones. Say so plainly when the chart steps down, or somebody will spend a quarter investigating a problem that was always there.
The assets you miss are never a random sample
Every coverage figure carries an unstated assumption: that the uninspected remainder is broadly like the inspected majority. It almost never is. The units that get missed are the ones that are hard to reach, out on hire, running a shift pattern, at the far site, or held by somebody who does not return calls. Those are also the units with the least day to day oversight and, often, the worst condition.
The National Audit Office made this point about the UK government estate. The Cabinet Office received condition data for only 64 percent of the buildings in scope for State of the Estate reporting in 2022 to 2023, and was concerned that the buildings it held no data for were the ones in the poorest condition (National Audit Office, Maintaining public service facilities, January 2025). Missing coverage skews towards bad news, at estate scale and at fleet scale alike.
Reporting coverage so that it is useful
- Show the denominator. "168 of 240 required" beats "70 percent" with no working shown, and it invites the right question.
- Segment by duty class. Statutory, contractual and discretionary, never blended into one bar.
- Pair it with backlog age. Seventy percent coverage with a 32 day mean overdue age is a bad quarter. The same 70 percent with a six month tail is a different conversation entirely.
- Measure over the cycle, not the calendar. For anything on an interval longer than your reporting period, quarterly coverage is a fragment. Report the rolling figure across a full cycle alongside it.
- Name the exceptions. If nine statutory examinations are outstanding, list the nine assets. A percentage lets people file the problem. Asset numbers do not.
One habit worth building: read coverage against inspection interval changes. The fastest way to improve a coverage number is to lengthen intervals, so a chart that jumps ten points in a quarter usually means the denominator moved rather than the work. Ask which of the two happened before anyone is congratulated.
