Two hundred and fifty square metres of sales floor. One operative. Twenty minutes. That works out at 4.8 seconds per square metre, and every one of those seconds includes walking in, signing in, getting the machine out of the cupboard, dealing with whatever the last customer of the day left in the fitting room, bagging the waste and setting the alarm on the way out.

Retail close-down cleaning is bought in twenty-minute and thirty-minute units across most of the multiples we work for, and the specification attached to it is usually a document written by somebody who has never held a stopwatch on a shop floor. What follows is what fits.

What does twenty minutes buy?

A defined core, done properly. It does not buy the shop.

Start with vacuuming, because on a carpeted or part-carpeted unit it dominates everything else. A 380mm commercial upright in an open corridor will cover 400 square metres an hour on paper. Put it in a fashion unit with rail bases, plinths, a cash desk run and eight fitting room doorways, and real productivity drops to somewhere between 150 and 250 square metres an hour once you account for turning, cable management and the head coming off the floor at every threshold. On that basis, a full vacuum of a 250 square metre floor is 60 to 100 minutes of time on the floor. It is three to five times the entire visit.

Hard floor is kinder. A 430mm pedestrian scrubber-dryer rated at over a thousand square metres an hour will still only give you 400 to 500 in a fixtured retail environment, and that is a manageable number – a 250 square metre vinyl or LVT floor is a 30 to 35 minute job with edges done properly, or about 15 minutes if you accept that the machine runs the open aisles and a mop does nothing else.

The arithmetic nobody puts in the tender

So the honest twenty-minute specification for a small high street unit reads roughly like this: machine the open floor areas, service the fitting rooms, wipe the cash desk and card terminals, empty and reline the bins, spot-clean entrance glass at hand height, quick check of the staff WC, waste out, alarm set.

That is it. Everything else – fixture dusting, shelf faces, mannequin bases, mirrors above head height, stockroom, back corridor, full-height glass – sits outside the visit and has to be bought as something else.

Why is the shutter setting the limit rather than the shop?

Because the cleaner cannot start until the last member of staff has cashed up, and cannot stay once the alarm needs setting.

In a shopping centre the window is tighter still. Centre management typically allows contractor access for a fixed period after mall close, and if the unit’s own team runs twelve minutes late on a Saturday, those twelve minutes come out of the clean rather than out of the access window. We have units at Westfield White City where the practical difference between a Tuesday and a Saturday visit is about a third of the available time on the floor, on an identical price.

High street units bring the opposite problem, which is that nobody is there at all.

Lone working after nine

An operative alone in a locked shop on Walthamstow High Street at half past nine, with the shutter down and a key and an alarm code, is a lone working scenario that needs a written risk assessment and somebody at the other end of a phone who will notice if the check-in does not happen. Plenty of contractors run this on hope.

Ask how your contractor manages it before you ask what the visit costs. Ask specifically what happens if an operative does not sign out, and who holds the escalation. If the answer is a WhatsApp group, keep asking.

Keyholding is a separate service with a separate liability, and if your cleaning contractor holds keys and codes for forty units, the response arrangement for a 3am alarm activation belongs in the contract in writing. It rarely is.

What comes off the list first, and who notices?

Fixtures. Always fixtures, and the answer to who notices is your area manager, roughly eleven days later, on a store visit.

A shop floor gets judged on three surfaces: the floor itself, the fitting rooms and the glass at eye level. Everything else degrades slowly enough that nobody registers it day to day, until somebody who has not been in for a fortnight walks through and sees grey shelf edges and dusty plinth tops.

The workable answer is zone rotation. Split the fixtures into five zones, do one zone properly each night on top of the core, and every fixture in the shop is touched within a working week. This costs nothing extra. It requires a floor plan with the zones marked on it, kept in the cleaner’s cupboard, and a supervisor who checks which zone was done. Zone four is the one that gets skipped, in every shop, because zone four is always the awkward corner behind the till run where the fixtures are lowest and the light is worst. Almost nobody does it, and it is the single cheapest improvement available to a retail cleaning contract.

The fitting rooms take a quarter of the visit

Eight cubicles, and you are looking at benches, mirrors, hooks, door edges, the floor and whatever has been abandoned in them.

Five minutes is realistic. Four if you are quick and the day has been quiet. On a Saturday in a womenswear unit on Clarence Street in Kingston, we clocked it at nine, because the floor was carpeted with hanger clips and the small hard plastic pins from price tickets, none of which a vacuum lifts reliably and all of which have to be picked by hand before the machine goes through. Pins puncture vacuum bags and they puncture feet.

Fitting rooms also produce almost every complaint that reaches head office, because they are the only part of the shop a customer inspects from eighteen inches away with nothing else to look at. If your visit is twenty minutes and the fitting rooms need nine, you have eleven minutes for the rest of the store, and you should price accordingly.

Where should the money go instead of extra minutes?

Barrier matting, and I would rather sell you fewer hours and see you spend the difference there.

The trade rule of thumb is that six metres of walked-on matting removes the large majority of the soil coming through a door. Most London units have somewhere between 1.2 and 2 metres, usually a decorative logo mat with a rubber back that traps nothing and gets slippery when it saturates. Everything the mat fails to catch goes onto your floor, and you then pay a cleaning contractor to remove it, every night, forever.

Matting beats hours

On a unit off the King’s Road we moved a client from a 1.5 metre entrance mat to a 4.5 metre primary and secondary matting run, and their close-down visit went from thirty minutes to twenty-two with a better-looking floor at the end of it. The matting paid for itself inside a year on the labour reduction alone, and that is before the carpet in the front third of the shop stopped needing extraction twice a year.

The other place to spend is frequency rather than duration. Five nights of thirty-five minutes will beat seven nights of twenty on almost any small unit, and I will argue that with any operations director who wants to have it out. The nightly visit is a habit rather than a requirement. What a shop needs is enough time on the floor to complete a proper cycle, and a Tuesday that gets skipped costs you far less than a Saturday that gets rushed.

What sits outside the scope entirely?

More than most buyers assume, and this is worth having written down before there is an incident rather than after.

Bodily fluid clean-ups and needle finds are separate call-outs with separate equipment and separate disposal routes, and so is gum removal. Any contractor who tells you their operative will deal with a soiled fitting room as part of a twenty-minute visit is telling you something about their training. External glass and fascia cleaning is a different trade with different access equipment. Stockroom and back-of-house work almost never sits in a close-down price, and the stockroom is usually where a mouse problem starts.

Merchandising is the boundary that causes the most arguments. Cleaners do not fold. They do not re-hang, re-size or face up. An operative who tidies a jumper stack is being helpful and creating a precedent, and within a month the store team expects it and the visit overruns. Put the line in the specification and brief both sides on it, because the person who suffers when it blurs is the cleaner, who is judged on a finish time they no longer control.

Waste, and the half hour you are allowed

Retail waste is bulky and it goes out at a time somebody else decides. A growing number of London high streets operate time banded collections under section 47 of the Environmental Protection Act 1990, as amended by the London Local Authorities Act 2007 – Westminster and Kensington and Chelsea among them, with other boroughs consulting.

That means bags can only be presented on the highway inside a specified window, and in Westminster’s case not more than thirty minutes before the allocated collection time. A close-down clean that finishes at 21:40 in a street with an 08:30 band leaves your waste sitting in the back corridor overnight, which is a fire loading question and a pest question at once.

Find the band for the street before you write the visit time. It is on the council’s site and it takes two minutes, and it decides more about how a retail cleaning contract runs than anything in the specification.

Take a tile down. The stained one, nearest the worst grille, and do it before you ring anybody for a quote. Turn it over and look at the back face, then run a clean white cloth across the stain on the front.

That takes ninety seconds and it answers most of the question. If the back of the tile is clean and the staining is only on the room face, you have a deposition problem and there is a reasonable chance of cleaning it. If the discolouration goes all the way through, or the tile has softened, or the reverse is dirtier than the front, the tile is finished and no chemistry sold in this country will bring it back.

What is the brown ring actually made of?

Dust, mostly, but the interesting question is which dust and how it got there.

In a central London office, the fine particulate arriving through the fresh air intake is largely traffic-derived – carbon, brake dust, tyre wear – and it is fine enough and greasy enough to bind to a porous surface. Add skin cells, fibres from carpet and clothing, paper dust from the print area, and in the last few years the glycerol residue from vaping, which is properly sticky and which I now see round extract grilles in washrooms in almost every building we service.

The stain forms where the air changes direction and where the temperature gradient is steepest. Cold supply air leaving a diffuser at 13 or 14 degrees hits a tile face at 22, and particles migrate toward the cooler surface and stick. That is why the pattern follows the throw of the diffuser rather than sitting as a neat square – a halo, wider on the side with the higher velocity.

The smear test

Wipe it with a dry white cloth first, then a damp one. Dry particulate lifts onto a dry cloth and leaves a grey mark. Grease transfers to the damp cloth and leaves a yellow-brown one, and grease means you are looking at kitchen or tea point air migrating through the ceiling void into the return path, which is a completely different repair.

Water staining is the third possibility and it looks nothing like the other two. Condensate tray overflow leaves a tide line with a hard darker edge, usually running away from the grille rather than round it, sometimes with a faint efflorescent bloom. Nobody cleans that. The tile is contaminated through its thickness and the tray is still blocked.

Why is it round the grilles and nowhere else?

Because the filter is not doing its job, or because there is no filter in the airstream at all.

Most London offices are on fan coil units above the ceiling with a basic inlet filter – the grades that used to be called G3 and G4 under EN 779, now expressed as ISO Coarse ratings under ISO 16890, which replaced the old standard in 2018. Those filters are not fine filtration. They are there to stop the coil blocking up, and they let a great deal of fine particulate through by design. Then the practical failures pile on top.

Filters get left out after a service. Filters get fitted in the wrong size and air bypasses round the edge, which produces a very characteristic result: the staining is worse at one corner of the grille because that is where the unfiltered air is coming through. Filters get left in for two years because the FCU is above a tile that needs a scaffold tower to reach. Return air grilles in a ceiling void plenum have no filtration whatsoever, so if your grid is a plenum return, the void itself is the filter and everything in it eventually shows up on the tile faces.

What the M&E contractor should be telling you

Ask for the filter change frequency and the last three dated records. Ask which grilles sit on supply and which on return, because return grilles stain from the room side and supply grilles stain from the plant side, and the fix is different in each case. In a City building with street-level intakes I would expect FCU inlet filters changed at least twice a year, and quarterly on the lower floors of anything fronting a bus route.

We had a floor on Leadenhall Street where the tenant had paid for a full ceiling clean, and it looked superb for about seven months. The filters were on an annual cycle that had slipped to eighteen months, and the intake sat two metres above a taxi rank. The staining came back darker than before, because the surfactant residue from the first clean had given the new particulate something to hold onto. They paid twice and were entitled to be annoyed about it.

Can a mineral fibre tile actually be cleaned?

Sometimes, within limits. By exactly one family of methods.

You cannot wet a mineral fibre tile. The face is a thin painted or vinyl-laminated skin over an open, absorbent body, and water goes through it, marks it and eventually drops it out of the grid. Every ceiling tile ruined by cleaning in this country was ruined by somebody applying too much moisture with the best of intentions.

The chemical sponge and where it stops working

The workhorse is a dry cleaning sponge – vulcanised natural rubber, no water, no chemical – drawn across the face in single overlapping strokes and turned to a fresh surface constantly. It lifts loose dry particulate remarkably well and it is what we use first on every job. Backed up with a HEPA vacuum and a soft dusting brush on the grille itself and the surrounding tegular edge.

Beyond that there are low-moisture ceiling restoration systems, applied as a fine oxidising mist that dries within minutes and lightens organic staining without saturating the tile. They work. They are also oversold. On dry particulate deposition round a diffuser, expect a good result. On grease, expect very little, because you are bleaching a stain rather than removing a film. On anything that has been through a condensation cycle, expect the mineral salts to come back.

Take a tile down and test on it, out of the grid, on a bench. Never test on a tile that is still in the ceiling, because a failed test patch in the middle of a floor plate is far more visible than the original stain, and you have then bought yourself a replacement anyway.

While we are on the subject: never paint acoustic ceiling tiles. It gets suggested on every one of these jobs, usually by somebody who has priced the tiles. Paint bridges the perforations, destroys the acoustic absorption you paid for, and can compromise the surface spread of flame classification the tile was certified to. A painted mineral fibre ceiling is a decorated ceiling that no longer does either of its jobs.

When is replacement the cheaper answer?

More often than our sales figures would like.

A mineral fibre tile is a consumable with a service life, and in an office with an unreliable filter regime that life is around ten to fifteen years. A clean buys you two to four years on a good result. If the grid is at year twelve, if the tiles have edge damage from years of cable pulls, or if more than about a fifth of them are stained, the arithmetic goes to replacement quickly – particularly since standard 600 x 600 tegular tiles in a common pattern are not expensive per square metre. Access and disposal are the real cost, and you pay those on a clean too. Tower hire, out-of-hours working, protection of the floor beneath, waste transfer for the old tiles – none of that changes much whether the tile coming out of the grid is going back in or going in a skip.

Why four new tiles make four hundred look filthy

Here is the trap that catches everyone. Mineral fibre yellows with age and with UV, slowly and evenly, and nobody notices because it happens to the whole ceiling at once. Drop four brilliant white replacements into a twelve-year-old grid and you have advertised the problem from the far end of the floor. The eye reads the new tiles as correct and everything around them as filthy.

The trade solution is rotation. Pull tiles from a store room or a corridor behind a door – anywhere nobody looks – and move them into the visible positions. New tiles go into the low-visibility spots. On a floor near Tottenham Hale we replaced eleven stained tiles round the diffusers by rotating them out of two meeting rooms and a comms room, and the tenant’s total spend was a morning’s labour and six new tiles that nobody will ever see.

What has to happen before anybody touches the grid?

Check the asbestos register. Every time, in any building constructed or refurbished before 2000.

The mineral fibre tiles themselves are very unlikely to contain asbestos. What sits above them is the issue – asbestos insulating board soffits, sprayed coatings on steelwork, lagging on old pipe runs, and the debris from all of it lying on the back of the grid. Under the Control of Asbestos Regulations 2012 the duty holder has to manage that risk and make the information available to anyone who is going to disturb the fabric, and lifting a ceiling tile disturbs the fabric.

The other reason to look up first

While the tile is out, put a torch into the void. Half the time the answer to the staining question is visible from there in about twenty seconds – a disconnected flexible duct blowing into the void instead of the diffuser, or a filter frame sitting empty where somebody never came back.

Take a tile down before the survey, not after it. The surveyor will charge you either way, and you will know what you are being told.

The call came in at 06:40 on a Thursday. Office manager, fourth floor of a building off Fetter Lane, talking fast, standing in about 40mm of black water with a fire crew on the landing behind her. One head had operated at some point in the night. Nobody knew when. The question she asked, twice, was whether we could get a team down there and have the print room clean before the insurer’s people arrived at two.

The answer was no, and telling her that was the single most useful thing anyone said to her that morning.

What happens in the first hour?

A wet pipe sprinkler system holds water that has been sitting in steel for years. It comes out black, oily, full of iron oxide and whatever microbial population has established itself in a dead leg since the last flow test. A single K80 head at typical operating pressure delivers something in the order of 55 to 80 litres a minute. If it ran for twelve minutes before anyone got to the stop valve, you have roughly 800 litres of that on your floor plate, and a proportion of it is already two floors down.

So the first hour belongs to safety and isolation, and nothing else. Electrical isolation of the affected area comes before anybody walks in – a print room is full of three-phase kit and floor boxes. Then the water goes down through service risers and cable penetrations, which means the floors below need checking immediately, and the tenants on those floors need telling.

Photograph it first. Before the fire crew leaves, before anybody lifts a ream of paper, before you have made a single decision about what is salvageable, get a full set of images and a video walk-through with the timestamp visible.

Who is allowed to close the sprinkler valve

Not you. The stop valve stays open until the fire and rescue service or your appointed sprinkler contractor confirms the incident is over, because a closed valve in a building where a fire might still be developing is a catastrophic decision. Heads operate individually, and one going off does not mean a fire existed. Impact damage and a corroded bulb seat are both common causes, and so is a stack of paper trolleys shoved up into the spray zone.

Once the system is shut down and drained, the building has no sprinkler protection until the head is replaced. That is a fire risk assessment event. BS EN 12845 requires spare heads to be held on site, and the replacement is a same-day job, not a next-week job. Your insurer will ask when protection was restored, and the gap between discharge and restoration will end up in the file.

What must not be moved before the loss adjuster arrives?

Almost everything you will want to move. This is the tension at the centre of every water damage claim, and nobody explains it well at the point it matters.

Your policy contains a condition requiring you to mitigate – to take reasonable steps to prevent further damage. It also, in practice, requires you to preserve the damaged property for inspection. Those two duties pull in opposite directions from about hour two onwards, and the way you resolve them is documentation. Anything you must move to prevent further damage, you record exhaustively before you move it, and you keep it in the building.

The skip is the enemy

I have watched a client at a business park near Ealing Broadway lose the better part of a £40,000 contents claim because a well-meaning facilities team had a skip on site by ten in the morning and 60 boxes of ruined letterhead and a soaked A0 plotter in it by lunchtime. The adjuster arrived to an empty room and a stack of photographs taken on three different phones with no timestamps and no reference marks.

Move damaged stock into a corner of the same floor, under sheeting, on pallets or blocks. Label it. Do not dispose of anything without written agreement from the adjuster, and get that agreement by email rather than on a phone call. If something has to go for genuine health reasons – saturated food waste, a fridge that was in the run-off – photograph it first, then log the reason with the time.

What has to be cleaned immediately regardless of who has authorised it?

Water extraction. That is the one thing nobody will criticise you for starting before the adjuster arrives, because standing water at hour six is doing damage that standing water at hour two was not.

Sprinkler discharge should be treated as category 2 water from the moment it lands, and in a warm office it degrades toward category 3 within around 48 hours as the bacterial load in the pipe water multiplies in the carpet backing. Every hour of delay moves the affected material from cleanable to disposable. Carpet tiles that come up and go into a controlled drying stack on day one are often salvageable; the same tiles left down until Monday are waste, and the adjuster will pay for waste at a lower figure than they would have paid to dry them.

Get the water off. Get air movers and dehumidifiers in. Lift the tiles at the perimeter and check whether water has tracked under the raised floor, because in a print room it almost always has – the floor void is where the power runs, and the power runs where the machines are.

Toner and cold water

Print rooms have a specific problem that generic restoration crews get wrong.

Toner is a thermoplastic powder engineered to fuse under heat. Waste toner bottles crack and cartridges leak, and the discharge then carries the pigment across the floor in suspension. If anybody attacks that with hot water or a steam machine, the toner fuses into the carpet fibre and the vinyl grain and stops being a cleaning problem at all. Cold water extraction only, at low agitation, over as many passes as it takes.

The paper is the other one, and it is a bigger problem than it looks. An 80gsm sheet wicks water upward through a stack far beyond the visible waterline – half a metre of apparently dry paper above a wet pallet base is usually wet paper. It swells, it cockles, it doubles in weight, and at anything above 20 degrees it starts growing mould inside 48 hours. Nobody dries bulk paper stock economically. Write it off early, count it accurately, then get it out of the room so it stops loading the atmosphere your dehumidifiers are working against.

The digital press and the MFDs stay switched off. Nobody powers on a wet machine to “see if it still works”, because that is how a corrosion claim becomes an electrical incident, and because the manufacturer’s engineer needs to see it in the state it was found. Chloride and iron in sprinkler water start attacking circuit boards immediately, and an electronics restoration specialist has a working window of roughly 48 to 72 hours before the corrosion is permanent.

Which trades do you need, and in what order?

Sprinkler contractor and electrician first, on the day. Then a specialist water damage restoration firm with extraction units, air movers, dehumidifiers and moisture meters. Then the electronics restorer for the machines. Then, some days later, a commercial cleaning contractor for the reinstatement clean – deodorising, hard floor treatment, ceiling grid, everything the drying process left behind.

That order matters because every firm on the list will happily start earlier than they should. A restoration crew that begins lifting carpet before the electrician has isolated the floor boxes is a liability, and a cleaning contractor who turns up on day two with a rotary machine will destroy evidence and toner in the same pass.

The job we turn down

We do not do the drying, and neither should most commercial cleaning contractors, whatever they tell you on the phone. A wet vacuum and three fan heaters is not a drying programme. Structural drying needs calculated air changes and a moisture log that stands up when the adjuster’s surveyor queries a fourteen-day equipment hire invoice. We have been asked to take that work on and have said no, including on a job at a client’s unit near Union Street where the alternative was the client waiting another day for a proper firm. They waited. It was the right call.

What we do is the clean afterwards, and the clean afterwards only works if the drying was done properly. A print room that has been dried to 12 per cent moisture content in the screed can be cleaned and reoccupied. One dried to 18 per cent will smell of damp in six weeks and nobody will be able to work out why.

What does the loss adjuster want to see when they walk in?

A room that is still recognisably the scene, with the water off it and the drying running.

They want a timeline: discharge time if known, discovery time, isolation time, the hour the restoration firm arrived. They want the photographs, dated. Photograph it first is the whole of the advice in this article, and the number of claims that turn on it is out of all proportion to how much effort it takes. They want the damaged stock still on site and countable. They want to know who authorised what, and they want to see that you did not sit and wait for permission before mitigating.

The moisture log

Start it on day one, in writing, whoever is doing the drying. Relative humidity and moisture content readings at fixed marked points across the floor plate, recorded daily against ambient temperature, with the operative’s name on each entry.

Adjusters challenge drying invoices more than any other line in a water claim. A log with dated readings settles that argument in about four minutes. No log, and you are negotiating from memory against somebody who does this every day.

The office off Fetter Lane reopened the print room eleven days later. Their letterhead stock was in a skip by then, photographed and counted, and paid at full value.

Yes.

With three conditions attached, one of which will cost you money and none of which appear in the industry white papers. I say this as somebody whose company still runs a great many night contracts and would keep every one of them if the client asked. The night model made sense when a London office was full from half eight to seven and empty the rest of the time. Most of the buildings we service no longer look anything like that, and a cleaning shift that starts after the last person leaves is now servicing a building that emptied out at four.

What does daytime cleaning actually change?

The shape of the labour, mostly. A day contract puts a resident team on the floor plate through the working day, running a rolling task cycle rather than a single compressed sweep. Somebody is on site at eleven when a coffee goes over the meeting room carpet, and at two when the kitchen sink fills up with cereal bowls, and at four when the third-floor washroom has taken 180 uses since the last check.

Under a night contract, all of that waits. The spill sets, the kitchen looks appalling for six hours in front of people who use the building, and the washroom runs out of hand towels at half two with nobody in the building who can refill it. What you buy at night is a building that looks perfect at eight in the morning and degrades unattended until nine at night. What you buy in daylight is a building that never gets past about 80 per cent of perfect, and never drops below it either.

The twilight shift is the worst of both worlds

Half the market’s answer to this is a four-till-eight twilight shift, and I think it should be abolished. It captures none of the responsiveness of day cleaning, because the working day is over by the time the team arrives on the floors that matter. It captures none of the efficiency of night cleaning, because the building is still half occupied and the team is working around people. And it is the hardest shift in the industry to recruit for, since it wrecks the evening without paying a night premium. Twilight exists because it lets a facilities manager say they have moved away from nights without any part of the working day changing.

Does it save money, and where does the saving come from?

Some of it is real. A lot of what gets claimed is not.

The London Living Wage went to £14.80 an hour for implementation by May, against a statutory National Living Wage of £12.71, and any decent London cleaning contractor is paying the former. Night work typically carries an uplift of 10 to 20 per cent on top of that, plus a higher recruitment cost per head, because night retention in this industry is dreadful and every leaver costs you three weeks of a supervisor’s time. Move the shift into daylight and the premium goes.

Now the part our sales team would rather I left out. A day team needs more hours than a night team to deliver the same specification. Interruptions are real – a cubicle is occupied, a meeting room is booked, a bank of desks has people at it, so the operative moves on and comes back. On the transitions we have run, the honest figure is somewhere between 10 and 20 per cent additional hours for an equivalent scope. That eats most of the shift premium saving. If a contractor tells you the labour line falls by a quarter, ask them to show you the hours.

The energy figure everybody quotes

You will see 20 per cent energy savings claimed in every daytime cleaning brochure printed since about 2009. The number is invented, or at least nobody has ever shown me the study behind it.

The genuine energy and services saving depends entirely on your lease and your building. In a single-let building where you control the plant, closing the building down at seven rather than eleven saves lighting on every floor and the air handling run-on behind it. In a multi-let tower it can be bigger, because out-of-hours HVAC is charged back by the managing agent by the hour per floor, and out-of-hours access often carries a security charge on top. A tenant we work with at Chiswick Business Park was paying for four hours of out-of-hours conditioning across two floors, five nights a week, purely so the cleaners could work in a comfortable building. That charge alone was worth more than the entire shift premium.

Ask the agent for the out-of-hours plant charge before you ask anyone for a cleaning quote. That single number decides most of these cases.

What stops working when you move the whole spec into daylight?

Plenty, and this is where most transitions go wrong, because the model gets sold as a straight swap.

Noise first. A standard commercial upright runs at somewhere around 72 to 75 dB(A), which is unusable next to a desk where someone is on a call. Low-noise machines sit around 58 to 62 dB(A) and are entirely workable, but they cost two or three times as much, and a contractor who wins a day contract on price will bring the old fleet and hope. That is the single most common cause of a day cleaning trial collapsing in month two. Put the machine specification in the tender document with decibel figures attached, and ask to hear one running in front of people at a desk before you sign anything.

Then washroom access. To service a washroom properly during occupancy you have to close it, which means enough washroom capacity to take one out of service for twenty minutes. In a 1980s building on a tight core with one gents per floor, you do not have that, and the day team ends up doing quick checks all day and nothing thorough. Deeper washroom work then has to go somewhere else.

Confidential floors go the other way, in our experience. Legal and HR floors are usually the ones flagged as unsuitable for daytime work, and they are the ones that benefit most. A partner floor cleaned at eleven at night, unsupervised, with nobody from the firm in the building, is a security arrangement most risk teams would refuse if they were designing it from scratch today. Daylight work puts the client’s own staff in the room on floors where documents sit on desks.

The jobs that stay at night

Nobody strips and reseals a vinyl floor at two in the afternoon. Nobody hot-water extracts a carpet that needs six hours of drying time, or works off a low-level access tower under a live ceiling, or descales a washroom with an acidic product while the washroom is in use.

So the honest model is split. Day team for the daily and weekly cycle, plus a periodic crew that attends out of hours on a fixed calendar – monthly, quarterly, annually, depending on the task. Any tender that promises to move 100 per cent of your cleaning into daylight is either misunderstanding the periodics or planning to quietly stop doing them. Ask specifically how the floor maintenance programme is delivered. If the answer is vague, the answer is that it isn’t.

Which London offices suit it, and which don’t?

Single-let buildings with their own front door and a three-day occupancy pattern are the clearest yes we see. The Monday and Friday gaps give a day team room to do the heavier work with almost nobody around, and the building never sits unattended.

Multi-let floors are more complicated, since you can move your own demise to daylight and the common parts will still be cleaned to the landlord’s schedule at whatever hour the managing agent has always used. You end up with two regimes in one building, which works, but stops the reception areas benefiting.

Where I would say no

Trading floors and 24-hour operations. There is no window, the occupancy never drops, and the desks are never clear. A dealing floor in the City gets cleaned in the two-hour gap it has, and that gap is at night.

Laboratories and clinical spaces have their own access protocols that usually rule it out. And any building where the client’s real objective is a lower invoice rather than a better-serviced building should stay where it is, because a badly resourced day contract is visible to every employee, every day.

How do you make the switch without a mutiny?

Talk to the incumbent team first. TUPE will carry them across if you change contractor, and their hours and premiums come with them, so the saving does not arrive on day one. Some of them work nights because it fits a second job or a childcare arrangement, and you will lose people who cannot make the new pattern. That is a real cost and it should be in the business case.

Then tell your own staff, with names. The single largest cultural gain from daytime cleaning is that the cleaner becomes a colleague with a name badge rather than an invisible presence who moves things overnight. Retention improves sharply once people are working in front of people who say good morning to them.

Move one floor, not the building

Six to eight weeks of overlap, one floor at a time. Pick the floor with the fewest phone-heavy roles, run it for a month, and let the rest of the building watch. Every objection you will hear in consultation – the noise, the interruption, the awkwardness of someone emptying a bin at your feet – either shows up on that floor in the first fortnight or turns out to have been imaginary. Measure it with complaints logged rather than with a survey, because surveys about cleaning attract people who were already annoyed about something else.

Hold the periodic schedule out of the trial entirely. If you change the daily model and the floor maintenance programme in the same eight weeks, you will never know which one caused what. We started a transition at a client on Buckingham Palace Road with the fifth floor only, because the fifth floor had the fewest phone-heavy roles. The eighth floor asked for it in week three.

Half past seven on a Tuesday, third floor, doors open ninety seconds. The washroom looks immaculate – chrome dry, mirrors clear, floor still faintly damp along the skirting, bin liner squared off at the corners. And it smells of ammonia and something sweet laid over the top of it, which is the worst version, because the sweet part is the air freshener trying and failing. By ten o’clock it will have gone. By the following morning it will be back. Somebody will email the cleaning company, the cleaning company will send a supervisor, the supervisor will find nothing wrong with the clean, and the cycle repeats until either the tenant changes contractor or someone finally works out that the smell has almost nothing to do with what happened at nine the previous evening.

Where is the smell actually coming from?

The floor. It is nearly always the floor.

Urine is mostly water and urea when it leaves the body, and within a few hours bacteria break the urea down into ammonia and a set of uric acid salts. Those salts are the problem. They are crystalline, they bind to porous surfaces, and they do not dissolve in water – which means a mop and a general-purpose neutral detergent will slide straight over them and leave them exactly where they are. Add warm water and you rehydrate the salts just enough to get them volatilising again a few hours later. Every night clean, on some floors, is a small reactivation event.

They accumulate in grout lines, in the porous edge of a cut vinyl tile, in the joint where the tiling meets the skirting, and behind the pedestal of a WC pan where nobody has ever put a cloth. On a busy Wednesday a 200-person floor might see 400 uses of a two-bowl urinal, and a proportion of every one of those is going onto the floor rather than into the bowl. Ten years of that in a building with unsealed grout produces a floor that smells regardless of how well it has been mopped.

The half-metre splash zone

Draw an imaginary semicircle half a metre out from the front of every urinal and every WC pan. That is the zone. Get down on your knees in it with a torch and look at the grout colour against the grout two metres away, and you will usually see the difference before you smell it – a yellow-brown tidemark following the line of the appliances.

That zone needs an acidic descaler or a properly dwelling enzymatic product, applied, left to work for its stated contact time, agitated with a deck brush and extracted rather than pushed around. Ten minutes of dwell time is not optional. Most night teams are given four minutes for the entire washroom.

Why does it come back overnight if the washroom was clean at ten?

Two things happen between the cleaner leaving and the first person arriving, and both of them are building services rather than cleaning.

The first is ventilation. Sanitary accommodation in an office should have extract running at around 6 litres per second per WC, and in most London buildings that extract is wired into the BMS alongside the lighting and the air handling. Out of hours, everything drops back. So from about eight in the evening until the plant ramps up around seven, you have a sealed warm box with no air changes and a floor slowly off-gassing ammonia into it. The smell is not being produced overnight so much as concentrated overnight. Open the door at 7.30 and you get twelve hours of it in one go. By ten, when the extract has been running for three hours, it has cleared – which is why the facilities manager who investigates at eleven o’clock never finds anything.

The second is flushing, and this is the one almost nobody checks.

Your urinal controller stops flushing at seven

The Water Supply (Water Fittings) Regulations 1999 cap automatic urinal flushing at 7.5 litres per hour per bowl where a cistern serves two or more urinals, and require a device that stops the cistern filling when the building is empty. Sensible regulation – urinals used to flush all night for nobody. The controllers that deliver it work off a PIR sensor, a hydraulic pressure drop from the taps, or a time switch.

Here is what goes wrong. A good controller runs a hygienic flush after a set period of no detected use, typically every twelve hours, precisely so the bowl and trap do not sit loaded with concentrated urine all night. A cheap one does not. A time-switch installation set to office hours in 2016 and never touched since does not. So the last flush of the day happens at about half past six, and then a two-bowl urinal with the day’s residue in the trap sits at 22 degrees until Wednesday morning. Thirteen hours of anaerobic bacterial activity in a warm sealed room, and the cleaner who came through at nine had no way of affecting any of it.

We found exactly this in a serviced office off Great Portland Street. Two floors, identical fit-out, identical cleaning spec, one floor fine and one floor foul every morning. The difference was a failed solenoid on the fourth-floor controller that had been flushing on a fixed daytime schedule for two years. A plumber replaced it for about £180 and the problem vanished. We had been on site for eleven weeks by then, being blamed.

While we are here: waterless urinals should never have been fitted in general-access office washrooms, and I will argue that with anyone. The sealant cartridge principle is sound in a controlled environment with a disciplined maintenance regime. In a building where a relief cleaner will eventually pour bleach down one because it smells, the oil seal breaks, the chlorine hits the urine, and you get a far worse smell plus a trap full of uric scale in a pipe run that was never designed to be flushed. The water saving is real. It is not worth what follows.

Is the cleaning method feeding the problem?

Sometimes, yes, and this is the part we can be held responsible for.

What the mop is moving around

A flat mop or a string mop taken from the WC floor to the corridor and back into the same bucket does one useful thing and one destructive thing. It lifts visible soil, and it redistributes uric salts in a thin film across a wider area, including out through the door into the lobby carpet. If your washroom smells at the threshold rather than at the appliances, you are looking at a mop routine.

Colour-coded systems exist to prevent this and are on every contractor’s method statement, including ours. What matters is whether the sanitary mop head goes into the laundry every night or gets rinsed in the same slop sink and hung damp until Thursday. Ask to see the mop store. Get down on your knees and look under the shelf.

The other common error is chemical. Neutral detergent every night, forever, with no acidic descale in the rotation, on the reasoning that acid is aggressive and the tiling is expensive. Twelve months of that and the grout is loaded. The correct answer is a scheduled acidic treatment of the splash zone on a fixed cycle – weekly in a heavy-use core, monthly in a quiet one – with neutral cleaning in between.

When is it the building rather than the cleaner?

When the smell survives a properly executed descale. That is the test, and it is a cheap test to run.

The pan that has been leaking underneath itself

Failed sealant at the base of a WC pan is the classic. Urine wicks under the pan into the screed, and the screed holds it indefinitely. No surface cleaning reaches it. The fix is to lift the pan, kill the contamination, reseal and reset – a plumbing job costing a few hundred pounds that will outperform any cleaning intervention you could buy.

Then there are voids. Duct panels behind the pans, IPS panelling with a gap at the floor junction, and the space under a raised access floor where a leaking spigot has been dripping for a year. We had a case in an office on Old York Road in Wandsworth where the smell was coming through a floor box eight metres from the washroom, tracking along the void from a cracked WC connector. Nothing on the surface would ever have touched it.

Older buildings around Stratford and the E15 fringe throw up another one: cast iron soil stacks with degraded joints in the riser cupboard, venting into the washroom every time the stack takes a slug of water from the floors above. That smell arrives in waves rather than sitting constant. If it comes and goes on a rhythm, stop looking at the floor.

What should you stop paying for?

Two things, and we sell both of them.

The fragrance unit works against you

Scent diffusers first. The battery-powered aerosol units and the plug-in fragrance systems on a monthly rental are the single most effective way to delay diagnosis of a washroom odour problem, because they lower the complaint volume without changing anything. We sell them. We would rather you cancelled them until the underlying cause is fixed, because the fragrance and the ammonia together produce that sweet-sour note everyone recognises and nobody can name.

Stop paying for repeat deep cleans, too. One properly specified acidic descale of the splash zone will tell you whether cleaning is the answer. If the smell is back within a fortnight, buying a second one is money set on fire. We have turned down that second job more than once and told the client to get a plumber onto the flush controller instead.

Get down on your knees at half past seven, before the extract has had a chance to lie to you.

“You’re here five nights. We’re here three days. So the invoice comes down by two fifths, doesn’t it?” That was a facilities manager in Farringdon, on the phone, in January, sounding entirely reasonable. He was about half right, which is the awkward part. Some of what we charge that building for tracks his headcount almost exactly, and should fall the moment his people stop coming in. The rest tracks the calendar and would cost the same if nobody ever walked through the door again. A flat percentage cut takes the same slice out of both. The second slice is the one that comes back at you eight months later, usually as a smell from a kitchen floor gully that nobody can locate.

What does three days in the office actually do to occupancy?

Less than the arithmetic promises. Hybrid working compressed the week rather than flattening it. Tuesday to Thursday carry almost everything, Monday limps along, and Friday in most of the buildings we cover has become a quiet day with a skeleton crew and a lot of empty chairs.

We look after a media business on Curtain Road in Shoreditch, 340 staff, desk sensors on every bank. Their average weekly occupancy last quarter came out at 58 per cent. Their Wednesday peak was 81 per cent. Friday averaged 19. So the busiest day in that “three-day” office runs at roughly four fifths of what a full-time office ran at in 2019, and the washrooms on that day have no idea what the weekly average was.

Why Wednesday sets your specification

A cleaning specification is a capacity plan wearing a different hat. The soap dispenser on the third floor either lasts through Wednesday afternoon or it does not. The bin under the coffee machine either holds Wednesday’s cups or it overflows onto the floor by half three. Neither of those things is negotiated down by the fact that the same floor was empty on Friday.

You can reduce how many times we attend. You cannot reduce what a peak day demands from us on the occasions we do attend – and if you try, what you get is a cheaper contract that visibly fails one day a week, which is the one day everyone is there to see it.

Which parts of the contract genuinely scale down, and which don’t?

Two groups, and one awkward outlier.

The occupancy-driven group falls almost in line with footfall. Consumables – hand towels, tissue, soap, sanitiser, sacks – track use directly. General waste volume tracks use. Washroom replenishment rounds track use. Carpet soiling is a footfall function too: soil comes in on shoes, and fewer shoes means slower loading of the traffic lanes, which pushes your extraction cycle out by months rather than weeks. Touchpoint work scales with the number of hands touching things. All of that should come down, and if your contractor resists it, ask them why.

The time-driven group does not move at all. Dust settles on an empty desk at the same rate it settles on an occupied one, and rather more visibly, because nobody has disturbed it. Internal glass gets marked by cleaning equipment, sunlight and building movement as much as by fingers. Hard floor maintenance is a wear-and-chemical cycle, and vinyl in a Croydon office block will still need its annual strip and seal whether the desks above it were full or not. Drainage, pest prevention, high-level dusting, external window cleaning, washroom deep cleans – all calendar work. Cutting them because the office feels emptier is the single most common false economy I see in tender documents.

Then the kitchen, which misbehaves upward. Office days have become social days. People who come in on a Wednesday come in to see each other, and they do that around the coffee machine and the fridge. Per head, we are clearing more cups, wiping more spills and pulling more forgotten lunches out of fridges than we were when everybody was in five days a week. A client of ours off Beadon Road in Hammersmith cut to three cleaning nights and spent the next quarter complaining about the fridge. The fridge was not the problem. Thursday night to Tuesday morning is 96 hours, and a chicken salad does not care about your occupancy dashboard.

Consumables and waste – the honest saving

This is where you should push hardest, and it is money out of our pocket, so take it as offered.

Plenty of contracts still bill consumables as a fixed monthly line agreed when the building ran at full occupancy. If yours does, you are paying for volumes you stopped using in 2020. Move to consumption-based billing with a monthly stock reconciliation. On a 300-desk floor that shift alone has been worth four figures a month to buildings we work in, and it costs us the margin we used to make on the difference.

While we are being honest: daily desk sanitising was mostly theatre, and we sold a great deal of it in 2021. Surface transmission was never the main event for respiratory illness. Keep the touchpoint work – door furniture, lift buttons, tap heads, handrails – and let the desk wipe go, particularly in a hot-desking layout where the cleaning matters more but the schedule that delivers it is the wrong one anyway. Wiping a hot desk at 8pm does nothing for the person who sat down at it at 2pm.

Who is actually paying for the reduction?

Worth working out before you start negotiating, because in about half the buildings we service the tenant has no authority to cut anything.

If you occupy a floor in a multi-let building, the cleaning of your demise is yours and the cleaning of the common parts is the landlord’s, recovered through the service charge. You can move your own floor to three nights tomorrow. The lift lobbies, the ground-floor reception, the shared washrooms on your level and the goods route all sit outside your gift, and the managing agent will not reduce them because one tenant is in on Wednesdays. Nor should they, since the other tenants may be running a different pattern entirely.

We see this play out in the tower blocks around Canary Wharf constantly. A tenant cuts their demised cleaning by 40 per cent, their service charge does not move, and the saving on the total occupancy cost turns out to be around 8 per cent. The conversation they needed was with the managing agent, about whether the common parts specification still reflects how the building is used. Slower argument, better money.

The sublet floor nobody cleans

The other trap is the floor you have taken out of use. Furniture still in place, lights off, door locked, pulled off the cleaning route to save money. Six months later it needs a builders-standard clean before you can show it to a subtenant, and that quote will exceed everything you saved. Keep a monthly attendance on mothballed space.

What happens to a building that sits empty four days a week?

It develops problems that occupancy used to hide.

Water is the serious one. HSG274 Part 2 puts weekly flushing of little-used outlets in the routine monitoring table, and ACOP L8 expects it written into your scheme of control. In a five-day office, taps got used and the flushing regime was largely academic for anything except the disused shower room. In a three-day office with half the desks dark, whole runs of pipework go quiet – the far washroom on the sixth floor, the cleaner’s sink in a store nobody opens. Somebody has to run those outlets. In practice it is often our team, because we are the only people in the building with a route sheet and a reason to walk it. Cut us to two nights and lose the flushing, and you have created a gap in a written scheme that a Health and Safety Executive inspector will find in about four minutes.

Traps dry out as well. Floor gullies in kitchens, unused shower drains, the odd hopper behind a riser door – water evaporates from the U-bend across a four-day gap, and in a warm August you get sewer gas coming back up through the floor. We had it at a client on Cheapside two summers ago. Three weeks of investigation, a drainage contractor, an air quality consultant, and the answer was a dry gully in a kitchen the tenant had stopped using.

The bin that nobody emptied on Thursday

Food waste over a long weekend is the other one. Under a Monday-to-Friday clean, nothing organic sat in a building for more than about 60 hours. Under a Tuesday-Wednesday-Thursday clean, a Thursday afternoon takeaway container in a desk bin sits until the following Tuesday, and by then you have flies. If you cut nights, cut them at the front of the week. Clean Thursday night no matter what.

Should you cut nights or cut scope?

Nights. Almost always nights.

Scope cuts feel painless because nobody notices the missing task for a year, and then everybody notices at once – the carpet lanes set, and the washroom limescale turns into a resurfacing job rather than a descaling one. Frequency cuts show up straight away, and that is a mercy. Three thorough nights aligned to Tuesday, Wednesday and Thursday, with the full periodic programme intact, will hold a building better than five thin ones with the periodics stripped out.

The contestable bit: most London offices at three days should be cleaning in daylight, and the reason they don’t is embarrassment. Facilities managers tell me staff find it awkward to have someone vacuuming near them. What they mean is that they find it awkward. Daylight cleaning removes the unsocial hours premium, cuts the building’s out-of-hours lighting and HVAC load, and puts a supervisor on site while there is someone to talk to. It also means the washrooms get serviced during the Wednesday peak rather than nine hours after it, which is the entire point. A three-day office has the natural gaps to make it work – Monday and Friday are effectively free.

What to ask before you sign the variation

Ask for the reduction broken down by task and by hours, never as a percentage. Anyone who agrees to 40 per cent inside a phone call was either overcharging you before or is planning to cut labour hours quietly and hope. Ask who is flushing the outlets, and get the name of the schedule it lives on.

The gully in the ground-floor kitchen at Curtain Road gets a jug of water tipped down it every Thursday night. Takes eleven seconds. It is item nineteen on the sheet.

The important end of a commercial vacuum is the end the air comes out of.

Everyone buys on suction. Facilities managers ask how powerful it is, cleaners talk about pickup, the spec sheet leads with wattage in a big bold font. Meanwhile the thing that decides whether a machine improves an office or quietly fouls it happens at the exhaust, where the air the vacuum has just drawn in comes back out into the room – carrying, or not carrying, the fine dust it was meant to trap.

Get that end wrong and you’ve bought an expensive way to move dust off the carpet and into the air, where people breathe it.

What is a commercial vacuum actually for in a London office?

Picking up the visible stuff is the easy ninety per cent. Crumbs, grit, biro lids, the shredded corner of a Pret bag under a desk – any machine with a motor manages that much. The job that actually earns the money, and the one cheap machines quietly flunk, is the fine particulate. The trade calls it the fines: the sub-visible dust, the skin cells and fibre, the pollen, the traffic-borne PM2.5 that drifts in off the Euston Road and settles into every carpet tile in the building.

The fines are the fraction small enough to hang in the air when disturbed and small enough to travel deep into a lung. A vacuum with poor filtration takes those fines, stirs them up off the floor and puffs them back out the far end finer and more airborne than they were sitting still. So you’ve taken settled dust, harmless on the carpet, and turned it into respirable dust at head height in an open-plan office. In a sealed City tower where the windows are decorative and the air just recirculates, that dust has nowhere to go but round the floor plate all afternoon.

I once got called into a media agency off St John Street in Clerkenwell where half the floor was blaming the air-conditioning for their itchy eyes and endless sniffles. The air-con was fine. Their contractor was running a knackered upright with a torn bag every afternoon at three, refilling the room with everything it lifted. New machine, problem gone in a fortnight. Nobody ever did apologise to the air-con.

The dust you can’t see is the dust that matters

London hands a vacuum more of the fines than most cities do. There’s always a floor being stripped out somewhere nearby, and refits shed a fog of gypsum and brick dust that finds carpet three storeys down through the smallest gap in the risers. Add the diesel particulate the traffic has laid down over decades, still lurking in the fabric of every older building, and the fine-dust load in a Farringdon office runs higher than whatever the filter’s manufacturer assumed when they printed the claim on the box.

What does “HEPA” actually guarantee – and what does it not?

Here’s where the marketing turns slippery, because “HEPA” is four letters that a great many machines wear without ever earning them.

Real HEPA is a defined thing. Under the European standard EN 1822, a genuine HEPA filter carries a grade – H13 traps at least 99.95 per cent of particles at the hardest possible size to catch, H14 at least 99.995 per cent. That hardest size, around three-tenths of a micron, is the most penetrating particle size, and it’s a genuine quirk of physics: particles both larger and smaller than that are easier to capture, so a filter that holds the line right there holds it everywhere else too. That’s the number with meaning behind it. “EN 1822, grade H13, 99.95 per cent” is a claim you can hang a health argument on.

“HEPA-type.” “HEPA-like.” “HEPA-style filtration.” Not one of those is graded, tested, or promising you a single thing. They’re words chosen to stand next to the real one and borrow its coat, and they appear on a startling number of machines sold into London offices. A filter badged “HEPA-type” might catch a fair share of the fines, or it might catch very little – the phrase commits to nothing, which is exactly why it was reached for.

A HEPA element bolted to the back of the machine isn’t the end of it either. A properly built vacuum filters in stages: the bag or bin catches the bulk, a pre-motor filter takes the next tier down, the HEPA element mops up the finest fraction last. Skip the earlier stages and the HEPA filter chokes within a week, the suction collapses, and some poor soul pulls the filter out to “let it breathe” – which is the single most common way a HEPA vacuum stops being one.

Why a HEPA filter in a leaky machine is theatre

Now the part almost nobody checks, and the one I’ll cheerfully die on a hill about. A HEPA filter only does anything if all the air is forced through it.

That sounds too obvious to say, and it gets ignored daily. If the body of the machine isn’t sealed – a gap around the filter housing, a lid that no longer seats properly on a tired seal – then the air behaves like air and takes the easy road, which is around the filter rather than through it. You can fit the finest H14 element made in Germany, and if the machine leaks, the fines stream out through the seams unfiltered while that immaculate filter sits there looking like a clear conscience. This is the whole distance between “sealed HEPA system” and “fitted with a HEPA filter,” and it’s a distance most buyers never think to ask about. A machine waving a HEPA filter with no sealed body behind it is theatre. Costly theatre, in a nice livery.

Why does a vacuum’s noise level decide when you can use it?

For decades London offices were cleaned overnight, so the noise a vacuum made was nobody’s problem. That has changed, and quickly. Buildings would rather not light and heat a whole empty tower at eleven at night for a cleaning crew when energy costs what it now costs, insurers prefer fewer lone workers after dark, and daytime cleaning has become ordinary across the City and Canary Wharf. Which puts the vacuum in the room while people are trying to work.

An older commercial upright, or a domestic machine pressed into service, can push past eighty decibels. Loud enough that running it down an occupied Holborn office at two in the afternoon makes the cleaner a nuisance through no fault of their own – and loud enough, across a full shift, to start mattering for that cleaner’s own hearing. A well-engineered quiet machine sits around sixty. That twenty-decibel gap is the difference between a vacuum you can run past someone mid-call and one that clears the meeting room. The Quiet Mark, awarded by the Noise Abatement Society, exists to point at the machines that took the trouble.

The daytime cleaning problem

There are two people in this, and the client usually remembers only one. There’s the office worker wanting to hear themselves think. And there’s the operative behind the machine for hours, whose ears the Control of Noise at Work Regulations 2005 are actually written to guard – the lower action value lands at eighty dB(A), which plenty of older vacuums sail straight past. Volume isn’t even the whole of it. A machine at sixty-two decibels with a low, smooth note is far easier to sit near than one at sixty-two with a thin electric whine, because what the ear objects to is the pitch as much as the level. The better manufacturers tune the note the machine makes. The number on the meter is only half of what you’re paying for.

Is a more powerful motor actually a better vacuum?

No. And the belief that it is has shifted an enormous number of bad machines.

Wattage measures what the motor swallows, not what the machine cleans. A 1,400-watt vacuum draws more electricity than a 900-watt one; whether it actually moves more air through the nozzle depends on the airflow and the shape of the cleaning head far more than on the motor’s appetite. When the EU capped corded vacuum motors at 900 watts in 2017, a good part of the industry forecast feeble machines, and the serious manufacturers simply engineered the same cleaning out of less power – they’d never been leaning on raw wattage to begin with.

A vacuum that roars is usually leaking air somewhere past its motor and turning the waste into noise; the roar is wasted air, the sound of suction going missing. The buyer who stands in the demo, hears the thing howl and murmurs “now that sounds powerful” has been trained by forty years of bad design to read a fault as a virtue.

What airflow and water lift actually tell you

Want to compare two machines honestly? Ignore the watts and ask for two numbers. Airflow, in litres per second, is how much air the machine actually moves – that’s what lifts the fines up out of the pile and carries them to the bag. Water lift, sometimes called sealed suction, is how hard it can pull – that’s what hauls grit out of the depth of a heavy contract carpet. A machine strong on both while sipping its 900 watts will quietly beat the two-kilowatt monster your predecessor signed off, and beat it without emptying the room.

So which machine belongs in which building?

Depends on the floor, and anyone who gives you one answer for every site is selling something. Detailed work around a dense desk layout in a Soho creative office wants a cylinder or tub machine you can steer into corners. A big open Canary Wharf floor plate wants a cordless backpack, for the speed and, just as much, for losing the trailing cable that turns into a trip claim the moment someone’s carrying coffee. Long corridor runs are upright territory. Whatever the format, the filtration and the sealing standard have to hold the same across all of it – the shape of the machine is only ever about what suits the room.

The Henry question

Somebody always asks about Henry. And I’ve nothing against the little red tub – Numatic build them properly down in Chard, in Somerset, and with the right HepaFlo bags and the commercial spec they filter well and run for years. The mistake is assuming the smiley face means the machine is sorted for the job. A Henry off the shelf on a domestic filter, run eight hours a day across a City floor, is doing work it was never specified for, and pushing a fair share of the fines back out as it goes. Buy the machine for the building. Not for the face.

Six months into a contract, a facilities manager rings me. Not a client of mine – a friend in the trade, asking me to cast an eye over it – because the cleaners she’d inherited “weren’t doing the kitchen properly.” So I ask to see the contract. Two pages. The relevant clause reads, in full: “The Contractor shall keep the premises in a clean and tidy condition.”

That’s the whole scope. That’s everything.

There was no argument for her to win, because there was nothing on paper to argue about.

What actually goes wrong when the scope is vague?

Nearly every cleaning dispute I’ve watched unfold in a London office traces back to the same root. The client thought they were buying one thing. The contractor thought they were selling another. Nobody wrote down which, so both were free to be quietly, reasonably, completely wrong about each other. The client pictures the fridge emptied and wiped out every Friday and the glass partitions kept clear of finger marks. The contractor priced a bin-empty, a vacuum and a wipe-down of the desks, five nights a week, and not a thing beyond it. Both parties are behaving in good faith. Both are heading for a fall.

Vagueness always resolves in somebody’s favour, and it is almost never the client’s. When “clean and tidy” is the whole of the standard, the contractor is the one who gets to decide what clean and tidy means, and they will decide it at precisely the level they costed. Push them on it and they’ll point at the clause, fairly enough, and explain that what you’re describing is a variation they’d be glad to quote for. So now you pay twice. Once for the vague version you signed, and again for the version you had in your head all along.

That’s the row. Six months in, every single time.

The “as required” trap

The two most expensive words in a cleaning spec are “as required.”

They sound like coverage. They are the opposite of it. “High-level dusting as required” means the tops of the partitions get done when the contractor judges it required, which turns out to be roughly never, because the person judging is the person who’d have to pay someone to climb the steps. “Carpets cleaned periodically,” with no period actually named, is a task that lives permanently in next month. If a frequency isn’t a number – weekly, monthly, twice a year – then it isn’t a commitment. It’s a word standing where a commitment should be, keeping the space warm.

What should a proper scope of works actually spell out?

A scope worth the paper it’s printed on does three things without flinching: it names every task, ties each task to a specific area, and pins each one to a stated frequency. Miss any of the three and the gap becomes the thing you argue about eighteen months later.

Start with the frequency split, because it’s the backbone. Daily tasks are the ones people picture: bins out, desks wiped, kitchen worktops cleared and cleaned, hard floors mopped and carpets vacuumed, washrooms scrubbed and restocked, the glass entrance doors spot-cleaned of the day’s handprints. Weekly tasks are the ones that quietly slip: skirting boards, the full run of internal glass, telephone and desk sanitising, the fridge if you’re lucky enough to have it named. Then the periodic column, which is where the real money and the real trouble both live – high-level dusting, carpet extraction, internal window cleaning, hard-floor buffing, the deep kitchen clean.

Each task needs an area written against it. A proper spec carries a floor-by-floor, room-by-room breakdown: reception, the third-floor kitchenette, the two accessible WCs on each level, the client-facing meeting rooms that matter far more to the impression you make than the store cupboard nobody ever sees. “Clean the office” is not an instruction. “Vacuum and spot-clean the second-floor breakout area nightly, extract the carpet quarterly” is.

And each task needs a standard you can actually check it against. This is where a decent spec leans on something like the British Institute of Cleaning Science appearance standards rather than a pile of adjectives. “Sparkling” is not a measurement. “No visible debris, no streaking on glass, dispensers filled” is one. When the standard is checkable, the monthly walk-round takes twenty minutes and settles matters. When it’s a mood, the walk-round is just two people disagreeing about the word “clean” all over again.

The dull clauses matter as much as the glamorous ones, and they’re exactly the ones a thin contract skips. Access hours – is this a night clean wrapped up before the building opens, or the daytime presence that’s become normal since everyone drifted back to the office three days a week? Keys, alarm codes, the security escort for the comms room. Whether the two hours on site mean two cleaners for an hour or one cleaner stretched across a whole floor they can’t realistically cover. Put the hours in. Put the headcount in.

The periodic tasks are where the money hides

Everyone fixes on the nightly list, because that’s the visible part, the bit you can see wasn’t done. The disputes almost all live one column across.

Deep cleans, carpet extraction, high-level work, external glass – these are expensive and infrequent, and they’re the first things a keen quote goes silent about. A contractor bidding low knows exactly what they’ve left out. They’ve priced the nightly service to win the tender and left the periodic tasks either unmentioned or buried under “as required,” so that when the carpets in your Fitzrovia media office go grey and flat a year and a half in, the deep clean turns up as a fresh invoice rather than a thing you were sure you’d already bought. Name the intervals. Twice-yearly carpet extraction, quarterly high-level dusting, monthly hard-floor maintenance – whatever the site actually needs, but a number, on the page, agreed by both signatures.

So who’s buying the toilet roll?

Here’s the small clause that starts more rows than any other, pound for pound.

Consumables. The loo roll, the hand soap, the blue roll in the kitchen, the bin liners, the hand towels, the little plug-in air fresheners nobody will admit to caring about right up until they run out. Somebody supplies all of it and somebody pays for all of it, and in a vague contract that somebody stays gloriously undefined until the Monday morning the dispensers are empty and reception is fielding the complaints.

The bin-liner question

Decide it in writing. Either the contractor supplies and bills consumables at cost plus a margin, or the client buys their own stock and the cleaners simply fit and refill. Both work. What doesn’t work is discovering, three weeks into a contract in a busy Soho serviced office, that nobody ever agreed which, and the toilet roll has become a small diplomatic incident with an email chain of its own.

Why do the cheapest quotes come with the thinnest scopes?

Because the thinness is the strategy.

I’ll say it plainly, since somebody should: the lowest quote on a London cleaning tender is very often the one to bin first. Not always. But the gap between the cheap bid and the sensible one is rarely down to efficiency. The cheap bid buys its headline number by narrowing the scope – quietly, deliberately – so the money finds its way back later through variations and periodic “extras” a fuller spec would simply have folded into the monthly figure. On the contractor’s side, a vague scope is usually a choice. It’s a pricing tactic, and a fairly clever one, and you can’t really blame someone for playing a game the tender set up.

The managing agent’s copy-paste spec

The other reliable source of thin scopes is the managing agent running the building. A managed floor in the City or over at More London tends to arrive with a cleaning spec the agent has pasted straight from the last building, and the one before that, tailored to none of them. It’ll list tasks for a marble reception you haven’t got and say nothing at all about the two kitchens you have. Read it as though it were written for your floor specifically, because it wasn’t, and every gap in it is yours to inherit.

What happens when someone asks for something that isn’t in the contract?

Something always does. A leaving do in the boardroom that leaves forty prosecco glasses and a carpet full of trodden crisps. Builder’s dust settling over every surface for a fortnight after the refit on the floor above. A burst pipe up in the ceiling void that comes through on a Sunday and turns Monday into a salvage operation. None of it belongs in the nightly scope, and none of it should.

A contract worth signing says so, and says what happens next. It carries a variation mechanism – a clause setting out how out-of-scope work is requested and priced before anyone lifts a mop, ideally with an agreed hourly and day rate already sitting in the schedule so there’s no haggling in the middle of a crisis. Without one, every unplanned job becomes a fresh negotiation held at the worst possible moment, when the thing plainly needs doing and you’ve no room left to bargain.

Changing hands, and the TUPE surprise

Then one day the contract changes hands, and the scope starts mattering in a way people rarely see coming. When a cleaning contract in the UK moves from one provider to the next, TUPE usually applies – the existing cleaning staff transfer across to the incoming contractor on their existing terms and conditions. The new provider inherits the people, but they price and deliver against the scope. A fuzzy scope, then, doesn’t stop causing trouble when your current cleaners walk. It gets handed on, unresolved, to the next lot, who are left guessing at what “clean and tidy” was ever supposed to mean while absorbing a wage bill they didn’t set.

Which is roughly where my friend’s facilities manager found herself. Two pages, one useless sentence, and a kitchen nobody could prove was in or out.