Ask around your organisation when the radiators were last cleaned. Not whether, but when. The contract states an interval, nobody has the date in their head, and the service record shows a tick without a name and without a date.
The same question works for the roof lights, for the blinds, for the deep clean of the sanitary areas and for the light fittings. In most buildings at least one of these draws no answer at all.
Why it is these services that get dropped
When the daily routine cleaning does not happen, everyone entering the building notices the next morning. The bin is full, the floor is dull, somebody reports it.
A missed annual service has no visible outcome. Nobody walks into a room and observes that the radiators should have been cleaned fourteen months ago. The condition deteriorates slowly and evenly, and slow, even deterioration goes unnoticed because there is no point of comparison.
A set of circumstances pulls in the same direction. Periodic services are labour-intensive and need preparation. They frequently require access to areas that are occupied during normal operations, so they get pushed into holiday periods and shutdowns, where every building then falls due at once. They are rarely priced separately and are instead absorbed into the price of the routine cleaning. And they are performed by staff who have other duties.
From this follows an incentive that operates independently of any individual firm: whoever does not perform a periodic service saves immediately and carries no discernible risk. That is not a charge against the industry. It is a property of the contract form, and it can be fixed.
What counts as one
The term is vaguer than it looks, and there is no generally accepted definition. Some organisations mean everything that does not occur daily, others only the annual work. In this article we use it as a working term for schedulable services outside the daily and customary weekly routine, which therefore need a scheduling and evidencing process of their own.
The assignments below are examples from buildings we know and not a recommendation. What interval is right follows from use, soiling, materials, manufacturers' specifications, accessibility, the hygiene plan and the quality level sought.
Falling due monthly to quarterly are typically: radiators including the gaps between fins, skirting boards, door leaves and frames across their full surface, light switches and surrounds, internal glazing in doors and partitions, the inside of waste bins, lift door tracks, handrails in full.
Half-yearly to annually: deep cleaning of sanitary areas including tile grout, blinds and roller shades, light fittings and shades, surfaces above reach height, ventilation grilles, basement areas and ancillary rooms, façade glazing depending on the arrangement. Plus the periodic floor treatment, which goes by different names depending on the floor system: intensive clean, restoration of the polish layer, deep clean with a new build-up, or a manufacturer-specific recoating. On factory-finished floors, a conventional deep clean with stripping may expressly not be intended.
Glass cleaning occupies a special position here, because its result is so visible that users notice of their own accord. That does not make it immune; it too can be missed or only partly delivered. It merely shifts the likelihood of somebody noticing. Everything above eye level or behind a furniture front behaves the other way round.
Why the usual monitoring does not capture them
A quality inspection examines the condition on the day of the inspection. It is designed around routine cleaning, because the result of that is visible on the same day.
For a periodic service this no longer works once time has passed. If the radiators are dusty again four weeks after cleaning, that says nothing about whether the work was done. And if they are clean, that may equally be down to low usage as to the service having been performed.
It does not follow that visual inspection is useless. It simply has to happen at the right moment, namely promptly after the work is completed. Then it shows whether the scope is right, whether areas were skipped, what the result looks like and whether anything was damaged in the process. Later it answers none of those questions.
Monitoring periodic services therefore consists of two parts running in sequence: the check on timing and scope, and the prompt check on the result. Do only one of them and you have either a date without a result or a result you cannot attribute.
The record and its weakness
The usual arrangement is a service record or a task card that gets ticked off. The method has three weaknesses, and all three can be remedied.
It gets filled in afterwards. Completing a list at the end of the quarter produces a document that evidences little. Proving that in an individual case is hardly possible, and indications such as handwriting or ink are no use for the purpose. What matters is how the record is constructed: a paper sheet without prompt confirmation, without an unambiguous reference to the area and without traceable amendment notes can be added to at any time, which is why its evidential value is limited from the outset.
It lacks a reference to the area. "Radiators cleaned" without stating the area leaves open whether the whole building is meant or the ground floor.
And it lacks confirmation from the other side. A record kept by one party alone is a self-declaration. Having every small item countersigned, however, generates more administration than insight. What works is a graduated approach: a joint acceptance for the large items, a client spot check within a period for objection for the medium ones, and a digital completion notice with time stamp and area reference for the small ones.
A look at material consumption is a useful supplement. Deep cleaners, polish, grout cleaner and descaler are needed for periodic work and not for daily cleaning, and order quantities are hard to fabricate after the event. As a plausibility check that is serviceable; as proof it is not. Material may come from existing stock, be purchased across several buildings, be sourced through a subcontractor or be misallocated, and on some floor systems it is not envisaged at all. Conversely, an order does not prove that the product was used here.
What the contract has to say
The most effective intervention costs nothing and is nonetheless rarely made: at the start of the contract, the interval becomes a schedule.
Instead of "quarterly" there is then a window for each quarter, aligned with the occupancy of the building. Instead of "annually" there is a window with lead time and a latest completion date. A window is more workable than a fixed date, because weather, holidays, building works and access problems move things, and a plan that does not allow for it is abandoned at the first conflict.
That changes the nature of the monitoring. It no longer asks whether a service was performed at some point, but whether an agreed window was met and what happened within that window in which area. That can be answered.
Three additions make the plan robust.
Notice with a deadline. The contractor reports the date a week in advance. That allows the client to be present if it wishes and to adjust the occupancy. Above all it creates a trail independent of the service record.
Acceptance for the large items. For deep cleans and sanitary deep cleans, a short joint walk-through on completion is worthwhile, with a note. For smaller items the advance notice is enough.
A rule for postponement. Which reasons justify a postponement, by when it must be reported, which replacement window applies, and what happens if that one passes as well? Without such a rule every missed item ends in a discussion about whether it can still be made up.
One point is frequently overlooked here: a service that was missed and one that was performed inadequately are two different matters in professional terms and have to be documented separately. In one case nothing happened, in the other something insufficient did. Which claims and contractual consequences attach to each is a legal assessment, and we do not make it. What counts in practical terms is that the distinction remains possible in the documentation at all, because anyone booking both as a defect cannot make it later.
The procurement side
Anyone tendering afresh has a lever that a running contract no longer offers.
Periodic services should be set out separately in the specification with area, interval and method, rather than running along inside the routine cleaning position. That does not make comparing tenders harder, it makes it easier. Because only then does it become visible which tenderer has costed these services seriously and which has let them disappear into a tight overall price.
A tenderer pricing a periodic item conspicuously low may have good reasons: efficient equipment, a different method, favourable conditions. They may also have miscalculated or misread the item. That is exactly what the clarification of abnormally low tenders under Section 60 VgV is for, and it is the right route. The question can only be put, however, where the item is set out separately. Where it vanishes into the overall price, there is nothing for a clarification to fasten on.
On the order of magnitude: what share periodic services make up of the annual price depends heavily on the type of building. We are not aware of any reliable industry figures, and so we quote none. What our own costings show is this: the amount is high enough that its loss over a multi-year contract adds up to a material sum, and low enough not to stand out in the overall price. It is precisely that combination that makes it a blind spot.
Three questions for the next review meeting
When was the last deep clean, in which areas, and who was present?
Is there a schedule with dates for the periodic services, or does the contract only state frequencies?
Ask to see the order quantities for deep cleaner and polish over the past two years. An evasive answer proves nothing, but it is a good reason to look more closely.