When the Roof Leaked: Building Quality In Instead of Inspecting It Later
A First Nations government we worked with had just finished a new community centre. The ribbon-cutting was a good day. Three weeks later, the first heavy rain found a gap in the roof flashing, and water ran down an interior wall onto a brand-new floor. The fix was not expensive in itself, but the disruption, the closures, and the loss of trust were. The painful part was that an inspector had walked that roof and signed off. The defect existed before anyone checked for it; the inspection simply failed to catch it.
This is the trap behind "we will inspect it at the end." Inspection sorts good from bad after the work is done. It does not create quality, and it cannot recover the time and money already spent producing a defect. By late 2021, with crews still thin from the pandemic and materials arriving late and out of sequence, every avoidable redo hurt more than usual. The lesson was not that the team needed more inspectors. It was that quality had to be planned into the work, not bolted on at the finish line.
What planning quality in actually means
Building quality in means deciding, before work starts, what "good" looks like and how each step will protect it. It shifts the question from "did it pass?" to "is the work set up so it cannot easily go wrong?" For the community centre, that meant a few concrete changes on the next phase of the build.
Define acceptance up front. For each major work package they wrote down the acceptance criteria — the flashing detail, the slope, the test — and shared it with the trades before the work, not during the sign-off.
Verify at the source. The crew doing the work checked their own output against the criteria as they went, rather than waiting for a separate inspector days later.
Make defects visible early. A short daily walk and a simple photo log surfaced problems while they were still cheap to fix, not after the wall was closed up.
Treat handoffs as risk points. Where one trade's work fed another's, they confirmed the upstream work was right before the next crew built on top of it.
Why this is cheaper, not slower
Teams resist this because it feels like extra effort up front. In practice it is the opposite. The cost of a defect grows the further it travels: a flashing error caught on the roof is a ten-minute fix; the same error found after the interior is finished is a closure, a remediation crew, and an unhappy client. Inspection at the end finds the expensive defects; planning quality in prevents the defect from being built at all.
Inspection is detection — it tells you something is already wrong.
Prevention is design — it stops the error from happening in the first place.
Clear acceptance criteria turn vague expectations into something a crew can hit on purpose.
Catching issues at the source removes the multiplier that turns a small miss into a major redo.
On the next phase, the team did not add a single inspector. They added a one-page acceptance sheet per work package and a fifteen-minute daily check. The roof on the new wing has not leaked. More importantly, when the auditors and funders came through, the records showed not just that the work passed, but how it was kept right throughout — which is exactly what a capital project under public scrutiny needs.
If your team is still chasing defects after the fact instead of planning quality into the work, XNM's program & project delivery advisory can help you build it in from the start.