Evidence that stands up after the fact

An audit is rarely tested on the day it is done. It is tested later, often much later, when a regulator asks how you knew a building was safe in a given month, or when something has gone wrong and the question becomes what was known and when. At that point the value of the audit is not the inspection itself but the evidence it left behind. A finding that cannot be located, dated and tied to a place is, for practical purposes, a finding that was never made.

The gap between doing and proving

There is a quiet gap in a lot of compliance work between doing the thing and being able to prove it was done. The work happens. The fire door gets checked, the cladding gets a visual inspection, the asset gets serviced. But the proof ends up scattered: a note in someone's diary, a photo on a phone, an email to a contractor, a tick on a printed sheet that goes into a drawer. Each of those exists. None of them, on its own, reconstructs what happened with the confidence a serious question demands.

The golden thread principle under the Building Safety Act 2022, described by the Institution of Civil Engineers as an accurate, up-to-date digital record carried through the life of a building, is in part a response to exactly this scattering. It treats the record not as paperwork generated after the fact but as the thing the work produces. For higher-risk buildings the expectation is formal, but the underlying habit is good practice for any building. The point is to capture evidence as you go, in a form that is still legible to a stranger in two years.

What makes a finding defensible

A defensible finding is not a longer finding. It is one that answers the questions a sceptical reader will ask without anyone having to remember the context.

• What was inspected, described precisely enough that a different person could find the same item.
• Where it is, tied to a location on a plan or a named asset rather than a vague reference.
• When it was checked, with a date that was recorded at the time and not reconstructed afterwards.
• Who carried it out, and against what standard or checklist.
• What was found, including the absence of a problem, stated plainly.
• What happens next, if anything, with an owner and a date.

The list is unremarkable, which is the point. Defensible evidence is boring by design. The trouble starts when any one of these is left implicit, because the implicit parts are exactly what fade. The inspector remembers the context for a week. The record has to remember it for the life of the building.

Capture changes the quality of the record

Two audits can cover the same building to the same standard and leave behind records of completely different strength, purely because of how they were captured. Consider the same finding recorded two ways.

Neither column reflects more work in the building. They reflect different discipline in recording it. The right-hand column survives a change of staff, a change of managing agent, and the passage of time. The left-hand column depends on someone remembering, and memory is not evidence.

The point-in-time problem

There is a particular failure that catches well-run teams: a record that keeps changing after the audit was signed off. If a finding can be quietly edited weeks later, the audit no longer describes the building as it was on the day it was approved. It describes the building as someone last left it, which is a different and far less useful thing. A regulator asking about a specific month needs the record as it stood that month, not a continuously updated version that has absorbed every later change.

The discipline here is to freeze an audit when it is approved, so the signed-off version is fixed and any later change becomes a new, dated entry rather than an overwrite. We have written about why this matters in why an audit is only as good as its sign-off. Without it, the most carefully captured evidence can still be undermined, because no one can say with certainty what the record showed at the moment it mattered.

Building capture into the work

The way to get evidence that stands up is not to try harder at the writing-up stage. It is to remove the writing-up stage, by capturing the finding, the photograph, the location and the date at the point of inspection, in the same structured record every time. When that is the path of least resistance, the strong record is the one that gets made by default, and the team is not relying on anyone's willingness to do extra admin at the end of a long day.

This is the reasoning behind keeping audit findings, photographs and plans in one connected place rather than three. SAMRISK ties each finding to a location and an asset, holds the photographs with the record, and fixes an audit when it is signed off, so the evidence reflects the building as it was assessed. You can see how the inspection side fits together on the audits page, and how findings turn into tracked work in closing the loop on corrective actions.

Write it for the stranger in two years

The simplest test of an audit record is to imagine the person who will read it without you in the room, long after the day is forgotten. If that stranger can tell what was checked, where, when, by whom and what came next, the evidence will stand up. If they would have to ask you, it will not, and you may not be there to answer. The work in the building is the easy half. The record is what makes it count.