See inspection activity without losing the underlying record

Completion is the first thing worth knowing

Before any analysis of what inspections found, an organisation needs to know which ones happened. Scheduled-versus-completed, overdue work and unstarted assignments are unglamorous measures, but they describe whether the programme is operating at all — and they are frequently the measures nobody can produce.

Completion data is also the honest test of whether a workflow is usable. A checklist that is consistently completed late, or abandoned partway, is telling you something about its design rather than about the people using it.

Recurring findings matter more than individual ones

A single finding is a maintenance task. The same finding appearing across twenty properties is a procurement decision, a specification problem or a supplier issue — a different category of thing entirely, and invisible unless findings are recorded consistently enough to aggregate.

This is where structured capture pays back. Observations recorded as free text can be read one at a time but not counted; the pattern only emerges when the same defect is recorded the same way across the portfolio.

• ✓Look for the same finding recurring across properties, not just severity
• ✓Compare findings between inspectors to surface standard drift
• ✓Track how long findings remain open after they are raised
• ✓Watch which questions are most often left incomplete

Interpreting differences between inspectors

Two inspectors covering comparable properties should produce broadly comparable findings. Where they do not, the cause is usually the template rather than the people: a question ambiguous enough to be read two ways will be read two ways, consistently and forever.

Treating divergence as a data-quality signal is more productive than treating it as a performance issue. It points at the specific questions that need rewriting, which is a fixable problem.