The contractor presents its quality inspections. Ninety-four per cent, cleanly documented, with photographs and dates, kept without a gap over twelve months. The client leafs through it and afterwards knows exactly as much as before.
And yet nothing about that figure is likely to have been falsified. The problem lies elsewhere, and it is worth naming precisely, because otherwise the wrong conclusion follows. The most common wrong conclusion runs: then let us drop the contractor's own inspections.
What own inspections are good for
They are among the most effective instruments a contractor has for finding problems before anyone else does. Briefing, round planning, complaint handling and supervision on site sit alongside them, and none of these replaces another.
A site supervisor walking the building weekly sees things early: that the time allowed for one round is not enough, that a new member of staff has not mastered the sanitary procedure, that an area has been missing from every round schedule since the refurbishment. That is exactly the information you want, and it only arises where somebody looks regularly.
Abolishing own inspections would therefore be the wrong conclusion. They belong in the contract, and the client should insist that they are carried out.
One refinement is worth making here, because they can in fact evidence something. They demonstrate that an internal control system is operated, at what frequency inspections took place, which deviations the contractor identified itself and how it followed them up. What they cannot deliver is independent evidence that the service owed under the contract was performed as a whole.
Four places where they systematically flatter the picture
The point is not dishonesty. The distortion arises even where everyone involved is acting in good faith.
The selection. Whoever inspects chooses which rooms to enter. Nobody deliberately works against themselves. But everyone knows where things generally look good, and conveniently those rooms lie on the way.
The timing. The inspection happens when the site supervisor is in the building anyway. That is rarely the Friday evening shift and rarely the day two staff are off sick. It captures normal operations, and normal operations are not the problem.
The standard. Anyone dealing with a building daily develops a sense of what is realistically achievable there. That sense is often professionally justified and still diverges from what is contractually owed. Familiarity shifts the standard slowly and in one direction only.
The consequence. A finding that goes into your own report creates a need to explain yourself to your own branch office. A finding cleared up briefly in conversation instead creates none. Both lead to the defect being remedied. Only one leads to it appearing in the statistics.
The limit that remains even with the best intentions
Beyond these four points there is a structural limit, and it explains why even good own inspections are not enough on their own.
A building walk-through assesses results. The organisation producing those results simply does not appear in a results-based form.
That the round layout does not work in the time available. That the hours costed are too tight for the services agreed. That turnover is so high that uninstructed staff are working continuously. That periodic services have been pushed back for a year because there are not enough people. Those are the findings that interest a client, because they explain why defects recur, and they rarely appear in the inspection report.
That is not down to the attitude of the site supervisor. Well-run companies expressly want such observations. It is down to the instrument and to the incentives: a form asking about rooms and service types has no field for the costing, and a note about hours being too tight ends up, in practice, in a conversation with the branch office rather than in the official report.
Capturing that level requires other instruments alongside: a process check examining workflow, materials and staffing levels, and an internal audit asking whether the quality system itself is working. The contractor can carry out both, and both are something other than the walk-through.
What raises their value
Three things turn own inspections into a usable instrument without making them evidence.
The same standard as the client's inspection. Same room groups, same service types, same assessment levels. The same levels alone are not enough, though; the population, the selection rule, the sample size, the timing of the inspection and the definition of what counts as a defect have to be comparable too. As long as the two sides measure differently, they are talking about different things and arguing about the result.
The most effective single element here costs half a day a year: two inspectors assess the same rooms independently and compare only afterwards. Where they diverge, the standard gets settled against an actual floor area rather than in a discussion about terminology.
A traceable selection. Where the rooms inspected are drawn by a fixed rule rather than at discretion, most of the distortion disappears. The rule does not have to be complicated.
Submission as of right, not at will. A contractor who presents results only when they are good is not running quality control but a marketing exercise. The obligation to submit all results belongs in the contract.
The difference is the real finding
Where the contractor's own inspection and the client's use the same standard, something emerges that neither procedure delivers on its own: a measurable divergence between two views of the same thing.
A contractor whose own inspection sits consistently close to the client's is seeing what the inspector sees. That is the precondition for everything else, but it is not yet evidence of a functioning quality system. Whether it functions shows only in whether deviations are investigated for their cause, measures are implemented and recurrences become rarer. If both sit at 76, you have a quality problem, though a workable one, because both sides are talking from the same starting position.
Where the contractor's own inspection sits consistently and markedly above the client's, that is initially an indication of differences in selection, standard or calibration, and not yet a statement about cleanliness. Establish the cause before drawing anything from it. If it is confirmed that the site supervisor does not recognise defects a third party does, the contractor's own steering is not working, and that can be met with training and calibration rather than deductions.
Rarer and equally revealing is the reverse case: the contractor's own inspection comes out stricter. This occurs with contractors holding high internal standards and speaks in their favour. It may also stem from differing criteria, a different sample or simply from a lack of calibration, which is why this finding too is a reason to analyse and not a proof of good faith.
The divergence becomes meaningful only over several rounds in any case. A single comparison may come down to a bad day or a strict inspector. How large a deviation has to be, and how many comparable pairs of inspections are needed before one speaks of a pattern, should be settled before you start. Draw the line only once the figures are on the table and you will draw it where it happens to suit.
When no answer comes
One point in closing, which regularly draws objections in meetings and which we nonetheless hold to be right.
In a performance inspection several parties are usually surveyed: the client, the users of the areas, the contractor itself. Not everyone responds. And how you handle missing answers decides more about the result than it first appears.
Where the client or the users do not respond, that must not count against the contractor. It has no influence on their response rate. Their share therefore drops out of the calculation and the remainder is scaled up accordingly. That is not entirely unproblematic either, because those who do not respond may judge differently from those who do. The response rate therefore belongs in the report, and with it a note on how robust the figure still is.
With the contractor the case is different, and here a distinction is required that we impose on ourselves. Refusing to provide your own information breaches a contractually agreed duty to cooperate. That is a finding, and a serious one. It is not, however, a finding about the cleanliness of the building.
Two things should therefore sit side by side rather than being calculated into each other. The quality score is derived from the data actually collected and states the basis it rests on. The failure to cooperate is reported as a finding in its own right, with whatever consequence the contract provides for it. Press both into one figure and you end up measuring cleanliness, documentation and willingness to cooperate at the same time, and nobody can afterwards say what pushed the figure down.
What matters here still stands: silence must not pay. A procedure in which non-participation improves your own result rewards refusal. The answer lies in making cooperation visible and attaching a consequence to it, not in calculating it into the quality score. We have seen more than once that after the first round in which this point became visible, a contractor filled in the form. Usually with usable information, occasionally with observations about problems the client had caused and nobody knew about.
What follows for the contract
Three sentences are enough to turn own inspections from a point of contention into an instrument.
The contractor carries out its own quality inspections at a set frequency and with a set sample size. Critical findings are reported immediately; otherwise a standardised report is submitted at an agreed interval, with the full data made available on request.
The contractor's and the client's inspections use the same methodology, which is annexed to the contract.
The basis for deductions and contractual consequences is exclusively the client's inspection.
The third sentence is the most important, because it dissolves the conflict that would otherwise arise. As long as the contractor's own inspection has no financial consequences, it can be honest. Once it has them, it will not remain so.