Blog Thoughts

For most of my career water was the one thing on the job I never thought about. You turned the tap and it ran. The bill was an afterthought. That era is closing. As I write this, Thames Water is pleading with fifteen million Londoners to put the hosepipe down, the counties around us are already under formal bans, and the Environment Agency has the South East marked as seriously water stressed – not for a bad summer, but for good. Anyone running a cleaning operation in this city needs to understand what that means for how we work, because it’s already changing.

How bad is London’s water situation, really?

Start with the weather, because it’s the part everyone sees. Spring 2025 was the driest in more than a century, 2026 has broken heat records that stood for decades, and demand for mains water has jumped by half across London and the Thames Valley in the hot spells. Thames Water hasn’t imposed a formal ban as I write, but it’s a whisker away, and South East Water in Kent and Southern Water on the coast already have theirs in force, with Anglian’s covering the east. Thames has already pushed an extra billion litres into its network just to keep pace, and that only buys time.

Weather is only the trigger, though. The deeper problem is structural, and it doesn’t disappear the week it finally rains.

Why “seriously water stressed” is more than a bad summer

The South East is one of the most water-stressed regions in the country – drier per head than plenty of places you’d think of as arid, with more people packed onto less available water. The Environment Agency’s own planning says that under low river flows there is essentially no spare water left to take from the ground or the rivers across most of the region, and that England will need up to five billion extra litres a day by the mid-2050s just to stand still. That’s the equivalent of filling Wembley Stadium four and a half times over, every single day. London sits at the sharp end of that arithmetic. The tap will keep running, but the assumption that it always will, and cheaply, is finished.

Does any of this actually reach commercial cleaning?

The easy assumption is that hosepipe bans are a domestic matter – roses and paddling pools, nothing to do with cleaning contracts. That’s half right, and the wrong half to rely on. A Temporary Use Ban restricts non-essential hosepipe use, and the list of what counts as non-essential includes cleaning paths and patios, and washing a vehicle. That is a description of a good slice of commercial exterior cleaning. When a ban lands, jet-washing a forecourt or hosing down a bin store with mains water is squarely in scope, and the fine runs to a thousand pounds a breach.

Above the hosepipe ban sits the drought order, granted by government in a serious drought, which can restrict commercial and industrial water use directly. And beneath the legal picture is the plain commercial one: sixty per cent of all business water use in England now sits in regions the regulator calls seriously water stressed, and some companies in the driest areas have started refusing new business water connections altogether.

My honest view cuts against my own trade. A ban that stops a pensioner filling a watering can while a contractor jet-washes a car park all afternoon is a nonsense, and it won’t survive. External cleaning firms should be planning now as though the restrictions already apply to them, because within a few dry summers they will. The contractors still selling weekly mains-fed pressure washing as a default service are going to look, before long, the way the office smoking room looks now.

The jet-wash is where it bites first

Pressure washing is the thirstiest thing most cleaning firms do. A single hose or lance can pour out up to a thousand litres an hour – the water twenty-five people would use taking a four-minute shower each – and a morning spent blasting a service yard gets through a startling amount of it. Worse, most of that water ends up carrying silt and detergent straight down the drain, which isn’t only wasteful. In many cases it’s a pollution breach, because that runoff is meant to go to foul sewer, not the surface drain feeding the nearest river. Water recovery fixes both at once – less taken from the mains, and nothing illicit reaching the watercourse. The jet-wash is the first place scarcity bites, and the first place a sensible operation changes how it works.

Where does the water actually go in a cleaning operation?

If you’ve never metered it, the honest answer is that nobody knows, and that’s the first problem. Water is invisible on most cleaning jobs in a way that materials and labour never are. It shows up as one line on a bill weeks later, long after anyone could tie it to a task. Barely one business water meter in ten is a smart one, so for most operators the usage runs close to unmeasured.

When you do look, the pattern is consistent. External pressure washing is the biggest single draw. After that comes carpet and upholstery extraction, which pushes clean water in and pulls dirty water out by the gallon, and then the daily waste of mopping, where the bucket gets refilled far more often than the dirt requires. Kitchen and washroom deep cleans, taps running throughout, sit somewhere in between. A surprising amount goes on rinsing – rinsing off chemical that was over-applied or slathered on out of habit. Good water, straight down the drain.

Flooding a floor is a habit, not a method

The mop and bucket is where the ordinary waste hides. Most cleaners were taught to work wet, and a flooded floor feels thorough even when it isn’t – it just takes longer to dry and leaves more residue behind. A flat microfibre system, pre-dosed so each pad carries only the water and solution it needs, does the same job on a fraction of the water and leaves the floor dry enough to walk on. The equipment has never been the hard part. The hard part is twenty years of muscle memory that says wetter means cleaner.

What does water-tight best practice look like now?

The good news is that cutting water use in cleaning rarely means cleaning worse. It means cleaning deliberately.

Reclamation is the biggest single saving for anyone doing exterior or carpet work. Modern pressure-washing rigs can capture and filter the runoff and put it back through the lance, cutting fresh-water draw sharply and solving the drainage problem in the same move. Carpet extraction has its own recovery systems. For exterior work there’s also water you needn’t buy at all – a tank capturing roof runoff gives a free supply for washing down that no ban can touch, because it never came from the mains. The kit costs money up front and pays it back in water bills and in not being shut down the week a ban arrives.

I look after a distribution yard over in Charlton where the client had us jet-washing the hardstanding and the bin area every week with a mains hose, straight to the surface drain. Through the dry spell last year we changed it: the wash dropped to fortnightly except for genuine spills, and a machine that recaptures and recycles its own water replaced the mains hose. In between, the yard got swept mechanically. It looks exactly the same as it did. The water going down the drain fell by close to two-thirds, and the client stopped lying awake over what a drought order would do to their operation.

Indoors the wins are smaller, but they stack: correct chemical dilution so nothing needs over-rinsing, and microfibre pads instead of a slopping mop. Steam and low-moisture methods have their place on the jobs that suit them. None of it is exotic. It’s the difference between a cleaner who thinks about water and one who’s never had reason to.

Don’t dirty the water you’ll only have to clean again

The single shift in thinking worth more than any machine is this: every litre you use, you also have to deal with once it’s dirty. Water you never ran is water you never have to filter or pay twice for. Good exterior contractors now scope a job around capturing and reusing what they can, and treat mains water as the expensive, regulated input it has quietly become rather than something free that falls from the sky.

Is saving water going to cost or save the client money?

Honestly, not much today – and that’s the part the water industry needs to fix. The pricing is upside down: the more water a business uses, the cheaper each litre gets, so the client with the biggest waste has the weakest reason to care. Water is rarely a large enough line on a commercial budget to move anyone on cost alone. The Environment Agency admits as much in its own strategy, which is a strange thing to read in a government document.

The case for using less water rests on risk, not on this quarter’s bill. A business that has already halved its cleaning water use doesn’t flinch when a ban lands or a drought order restricts commercial use. It isn’t the one whose forecourt greens over because the contract was built around a hose that’s now unlawful to switch on.

The incentives are backwards, and it won’t hold

That pricing won’t last, because the sums behind it don’t. The government wants business water use down nine per cent by 2038, and smart meters are landing on commercial supplies in their millions. Metered water that someone can watch hour by hour tends to fall, simply because they’re finally watching it. The clients who get ahead of this – who treat water discipline as normal now, while it’s still voluntary – are the ones who won’t be scrambling when it isn’t. The rest will pay for the delay, one dry summer at a time.

Since 2020 I’ve had the same conversation in a dozen office kitchens. A facilities manager points at a glowing tower in the corner, or a lamp bolted above head height, or a wand still in its box, and asks whether the thing actually works. The honest answer is: sometimes, under conditions the person who sold it rarely spelled out. UV-C is real science with a genuine germicidal effect, and it’s also one of the most oversold products to reach the cleaning industry in years. The whole task is telling those two apart.

What is UV-C actually doing when it “disinfects”?

Ultraviolet light comes in bands, and the germicidal one is UV-C – the short, energetic wavelengths between about 200 and 280 nanometres. Shine enough of it on a bacterium or a virus and it wrecks the organism’s genetic material, so it can’t replicate. No replication, no infection. That’s the whole mechanism, and it’s been used to disinfect water and hospital air for the better part of a century. The workhorse lamp emits at 254 nanometres, and it does the job well.

There’s a catch, and it’s a big one. The same wavelengths that shred microbial DNA are hard on human skin and eyes. A 254-nanometre lamp will give you sunburn and a painful dose of what the eye doctors call photokeratitis – arc-eye, essentially – if you sit under it. Conventional UV-C can’t simply be switched on in a room full of people.

Why the light that kills germs also burns you

This single fact shapes every honest use of UV-C. Because 254-nanometre light is hazardous to people, it has to be kept away from them – run inside sealed air-handling ducts, or mounted high on the wall to treat only the upper air while the room’s convection currents carry microbes up into the beam. Or fired around an empty room after everyone’s gone home. Every legitimate deployment is really an exercise in keeping the light and the humans apart. The moment a product promises to break that rule cheaply, your guard should go up.

Where does UV-C genuinely work?

Against all that, there are still places UV-C earns its keep, and I’ll happily say so. Upper-room systems have decades of evidence behind them in tuberculosis control, where high-mounted lamps disinfect the air people are breathing without touching the people themselves. In-duct UV-C, built into the air handling, is a sensible layer in a building already moving a lot of air. And in hospitals, wheeled robots that flood an empty room with 254-nanometre light after the cleaners have finished do measurably cut surface contamination – one study found high-touch surfaces testing positive dropped from around two-thirds to under a fifth once the robot had run.

Read that last part again: after the cleaners have finished. The clearest finding across the whole surface-disinfection literature is that UV-C works best bolted onto a proper manual clean, never in place of one. It’s a second pass, and it can’t be the first.

The winners tend to be unglamorous

None of these winning applications looks like the thing being wheeled across an office floor. They’re ducts and ceiling-mounted fittings run to a schedule – infrastructure, not gadgets, scoped by someone who measured the space. The version sold to a general office – a portable tower you roll out at five o’clock, a handheld wand still in its packaging – is where the science and the marketing part company.

What’s really behind the “99.9%” on the box?

A UV-C dose is irradiance multiplied by time – how bright the light is where it lands, times how long it lands there. Both halves matter, and both fall apart in a real room.

Brightness drops with the square of the distance. Double the gap between lamp and surface and you’ve quartered the dose. Robot trials show it plainly: strong kill rates within a couple of metres of the lamp, then a steep fall-off – one set of measurements went from over ninety per cent effective at under three metres to the low forties at five. The desk in the far corner is not getting what the near one gets.

Then there’s shadowing. UV-C travels in straight lines and barely reflects off ordinary surfaces, so anything the light can’t see directly, it doesn’t treat. The underside of a desk, the far side of a monitor, the keyboard tray, the gap behind the bin – all shadowed, all missed. A tower in the middle of the floor cleans the tops of things and leaves the rest.

And “99.9 per cent”? That’s a three-log reduction. The bar for genuine disinfection in a hospital is usually five-log – a hundred-thousand-fold cut, not a thousand-fold – and it’s measured on clean, exposed test surfaces at a set distance, held for a set time. The number on the box is a best-case laboratory result. Your office is not a laboratory.

Dose is the number that’s missing from the box

I was called to a tenant’s office up at Here East in Stratford, where the facilities manager showed me two UV-C towers bought in a panic in 2021, still boxed in a cupboard. She wanted to know whether to start running them. I asked the only question that matters: what dose, at what distance, for how long? Nobody had ever told her, the brochure didn’t say, and without those numbers the towers were furniture. We worked out that to dose the far side of that floor properly she’d have needed to leave each tower running in each zone for the best part of an hour, shifting it by hand between cycles, every single night. They went back in the cupboard.

Is far-UV-C (222nm) the game-changer the reps claim?

The clever answer to the burns-you problem is a different wavelength. Far-UV-C, at 222 nanometres, comes from krypton-chloride lamps, and its selling point is real: the light is so short it can’t push past the dead outer layer of your skin or the tear film on your eye, yet it still reaches bacteria and viruses, which are far smaller. In 2022 the main industrial-hygiene body in the US, the ACGIH, weighed the evidence and raised the permitted daily eye-exposure limit for 222-nanometre light around sevenfold, from roughly 23 to 160 millijoules per square centimetre. Lab and room studies show it knocking airborne bacteria down by well over ninety per cent at levels people can sit under. On the face of it, this is the one that lets you disinfect a room full of working people. The reps love it, and I understand why.

The brochure stops there. What it leaves out is that a 222-nanometre lamp doesn’t only make germicidal light – it makes ozone. Measurements in working offices have clocked meaningful ozone production, pushing indoor levels up by several to twenty-odd micrograms per cubic metre depending on how many lamps run and how many people are in the room. Ozone irritates the lungs on its own, and it doesn’t stop there: it reacts with the volatile organic compounds already drifting round an office to form ultrafine particles and secondary aerosols – the very things sound indoor-air practice tries to cut down.

There’s a wrinkle here that ought to bother anyone in my trade especially. A lot of those airborne compounds come from cleaning products and air fresheners, and the citrus-scented ones are loaded with limonene, which is exactly what ozone loves to react with. So the gadget sold to purify your air can, in a sealed and busy room, quietly turn your lemon-scented surface spray into fine particulate. The ozone dose is usually small, and in a well-ventilated room it may not matter at all. But “may not matter, if ventilated” is a long way from the promise on the stand at the facilities show.

My own view here is blunt. I won’t recommend a 222-nanometre system for an occupied office I clean until the ozone and air-chemistry questions are properly settled, and any rep who tells you they already are is selling, not informing. The science is real and the promise is genuine. Promise is a long way from proven-in-your-boardroom.

The ozone the sales sheet forgets

Far-UV-C’s two headline properties pull against each other. To kill enough airborne virus you want plenty of light in the room, and the more light, the more ozone and downstream chemistry. Ventilation fixes the ozone – but ventilation on its own also clears the airborne virus, which begs an awkward question about what the lamp is adding on top. The researchers still thrashing this out in the journals haven’t reached a settled answer. A salesperson quoting you one reassuring figure has simply chosen the study that suits.

So should your London office buy any of it?

For the great majority of London offices, the honest answer is no – or at least, not yet, and not the portable version. The things that reliably lower the germ load in an office are dull and already to hand: fresh air moving through the space, and a cleaner who actually lifts the keyboard and wipes under it. Spend the UV budget on better ventilation and a proper cleaning spec before a penny goes on a glowing tower. Run a hospital or a genuinely high-risk, high-occupancy space and the sums change – validated, engineered systems have a real place there. A normal office is not that place.

Four questions that separate the engineers from the opportunists

If someone is selling you UV-C, four questions do the sorting. What dose does your system actually deliver to a surface two or three metres away, and how long does it need to run to get there? Where’s the independent validation – not a lab certificate for a petri dish, but evidence of a five-log reduction in a room like mine? If it’s far-UV-C, what’s the measured ozone output, and are the lamps properly filtered? And what happens to every surface the light can’t see? A good supplier answers all four without flinching. The rest reach for the brochure, and that tells you what you need to know.

The keys go back on a Friday and the argument starts the following Tuesday. I’ve cleaned enough offices and units on their way out to know the pattern by heart: a tenant thinks they’ve handed back a clean building, the landlord’s surveyor thinks otherwise, and a schedule of dilapidations lands weeks later with a cleaning bill nobody saw coming. Most of that fight is avoidable. It comes down to what got cleaned and whether anyone thought to photograph it. This is the commercial end-of-tenancy clean, and it works nothing like the domestic version.

What does “end of tenancy” actually mean in a commercial lease?

In a domestic let, end of tenancy cleaning is a fairly settled thing – a checklist and a deposit scheme to referee any dispute. Commercial is a different animal. Most business premises in London are let on full repairing and insuring terms, which means the tenant carries the cost of keeping the place in order and handing it back that way. The lease contains a yield-up clause, and it usually says the tenant must return the premises clean and clear of their belongings at the end of the term. “Clean” there means a good deal more than a hoover round on the last afternoon. A surveyor measures it against what the building ought to look like, and any shortfall turns into money.

There’s no deposit scheme to protect you either. The landlord’s route to recovering cleaning and repair costs is a schedule of dilapidations, served by their surveyor, listing every alleged breach with a price beside it. That schedule is an opening position, not a bill you’re obliged to settle in full – but tenants who don’t know that pay it in full nine times out of ten.

Why a schedule of condition changes everything

If one document decides how the end plays out, it’s the one taken at the start. A schedule of condition is a dated, photographed record of the premises on day one, agreed and attached to the lease. Where one exists, the tenant only has to hand the place back in that state – dust and wear that were there at the outset aren’t yours to fix. Where one doesn’t, you’re arguing from memory against a surveyor with a camera, and memory loses. I’ve seen a tenant save five figures on a Wandsworth claim purely because someone had the sense to photograph the grubby carpet tiles before the desks went in.

What does a proper commercial end-of-lease clean actually involve?

Scale is the first shock. A warehouse isn’t cleaned the way a flat is, and neither is a fifteen-thousand-square-foot office floor, and the things that generate dilapidations claims are rarely the things a tenant’s own staff think to touch. The visible floor gets done. The trouble hides above head height and under the furniture.

Kitchens and tea-points are where I find the worst of it. Years of splashback grease behind a microwave nobody ever moved, a fridge seal gone black at the folds. Washrooms tell the same story – limescale furring the taps and hard-water staining down the urinals. The extractor grilles, up where you’d need a step-ladder, sit packed solid with dust. None of it shows until someone runs a finger along it, and a surveyor runs a finger along everything.

Then the floors. Carpet tiles in a long-occupied office are often past cleaning and into replacing, and there’s a real argument to be had over whether that’s fair wear or a genuine breach. Vinyl needs stripping and resealing, not mopping. And in the industrial units, the floor is the whole game.

I cleaned a light-industrial unit off Coronation Road in Park Royal last year where the outgoing tenant had run a small fabrication business for a decade. The concrete was a map of oil stains, forklift tyre marks and the pale shadows where racking had stood bolted down since before they arrived. They’d assumed a jet-wash would sort it. It didn’t – oil that’s soaked into unsealed concrete for ten years doesn’t lift with water and good intentions. We degreased it in stages across three days, and even then some of the deep staining stayed. The lesson the tenant took away, too late to spare them, was that the floor should have been sealed the week they moved in.

The outside counts too, more than tenants expect. Loading bays and yards get inspected. So does the external glazing, and the ghost of a fascia sign left behind as a rectangle of clean brick and a row of drilled holes when the tenant’s board finally comes down. A clean that stops at the front door leaves a surveyor plenty to write up.

The bits every tenant forgets

High-level cleaning is the great blind spot – the tops of partitions and the ductwork nobody’s looked at since the fit-out, furred grey with dust. Floor boxes are the other one: those recessed power-and-data points sink into the carpet, fill with dust and biscuit crumbs, and get missed completely because the desks sat on top of them for years. Then the glass, which shows every fingerprint – the manifestation strips on partitions, the smears at door-handle height. This is the layer that separates a clean a surveyor signs off from one they mark up, and nine times out of ten it’s the layer a general office clean skips.

Where do outgoing tenants get caught out?

The single biggest mistake is treating the clean as the last job instead of the first. Tenants spend the final week shifting furniture and cancelling the broadband, then send two people round with a vacuum on the last morning and call it done. By the time the schedule arrives, the building’s been relet or stripped and there’s no way left to prove what state you handed it back in.

I’ve watched this next part happen too many times to call it coincidence. Managing agents put cleaning items on dilapidations schedules that they know a tenant’s own photographs would knock straight off – the “professional deep clean throughout” line, priced at a figure designed to make you flinch. They do it because most tenants pay rather than argue. I think that’s closer to a shakedown than a claim, and the one thing that reliably stops it is evidence the tenant took the trouble to gather on the way out.

A receipt proves you paid, not that you cleaned

Tenants wave invoices at managing agents as though a receipt from a cleaning firm settles the matter. It doesn’t. A receipt proves money changed hands. It says nothing about the state of the third-floor kitchen the day you left. What holds up is a dated set of photographs – wide shots and close-ups of every problem area, taken the morning the keys go back. Nine times out of ten the tenant who brings photographs to the negotiation pays a fraction of what the tenant with a receipt pays. Do the clean, then prove it. The proving is half the job.

What should landlords and managing agents actually ask for?

Landlords and their agents have a fair interest here. A building handed back filthy is genuinely harder to relet, and a tenant who’s wrecked a tea-point should put it right. The trouble starts when the clean becomes a way to hand the landlord a better building than the one they let out.

That’s betterment, and the law won’t have it. A landlord can require the premises back in the condition they were let in, not upgraded at the departing tenant’s expense. If the carpet was mid-life and tired when the tenant took the keys, they don’t owe you a new one. Section 18 of the Landlord and Tenant Act 1927 caps what a landlord can actually recover at the drop in the building’s value caused by the breach – so if you’re refurbishing the whole floor anyway, the cleaning you’re claiming for may be worth nothing once the sums are done. That’s the part agents tend to skate over: the headline figure on a schedule and the figure a landlord can actually recover after Section 18 are often two very different numbers.

Good agents know this and scope realistically. The smart instruction to a cleaning contractor is plain: get the building to relettable standard and document it. A defensible, photographed clean is worth more at settlement than an inflated one that falls over the moment it’s challenged.

The betterment trap

I get asked more often than you’d think to quote for a clean that’s really a refurbishment wearing a disguise – strip the lot out and bill the departing tenant for a building that looks new. I don’t play that game, and the agents worth working with don’t ask me to. A contractor who’ll dress up betterment as dilapidations is a contractor who’ll hand over a document that comes apart the second the tenant instructs their own surveyor. The credible clean is the one scoped to the lease, not to the landlord’s wish-list.

Can you get through a commercial move-out without the dilapidations fight?

Mostly, yes – and the tenants and agents who manage it tend to do the same two things. They agree the standard up front, and they inspect together.

The walk-round that settles it before it starts

The clean that never turns into a dispute is the one where both sides walked the building together before the keys changed hands. The landlord’s agent and the outgoing tenant in the same rooms on the same afternoon – ideally with the contractor who did the work standing alongside them – agreeing what’s clean and what still needs attention while there’s time to fix it. Half an hour of people sorting it out in person, in the actual tea-point, beats three months of surveyors’ letters. Book the clean early enough that a week sits between finishing and handing back, so there’s room to put right anything the walk-round throws up. The tenants who leave it to the last morning are the ones I hear about later, still arguing the toss six months after they moved out.

The best cleaner I ever employed stayed with me for eleven years. She knew which office kept confidential files out on the desk overnight and where the spare bin bags lived on the third floor, and she knew it without being told twice. When she retired it took three people the better part of a month to cover what she’d done on her own. That’s the thing nobody tenders for: continuity. And in London commercial cleaning, continuity is rare.

Why do cleaners keep walking out of commercial contracts?

Turnover in this trade runs high for reasons that stop being mysterious the moment you’ve built a rota yourself. Pay sits underneath most of it. A cleaner on the statutory minimum, in a city where a monthly travelcard swallows a chunk of the week’s wages, is doing sums that don’t add up. When a warehouse round the corner offers forty pence more an hour, the maths makes the decision for them. I’ve lost good people to a supermarket depot and a hospital portering job in the same month, both for pennies more and daylight hours.

Then there’s the way the job treats people. Cleaners are often the only workers in a building told to be invisible – finished and gone before anyone important arrives. Invisibility by design wears at you. A supervisor who never learns your name, a head office that only rings when something’s been missed, a client who steps over the mop bucket without a word: it adds up to a message, and the message is that you don’t much matter.

I’ll say the thing most of my competitors won’t. Zero-hours contracts are the single biggest driver of churn in this industry, and any contractor still handing them out is telling you plainly how they see the people doing the work. You can’t ask someone to build childcare and travel around hours you won’t guarantee, then act surprised when they’ve gone by spring.

The shift that eats your evening

The hours are their own problem. Offices get cleaned when they’re empty, so the work lands before seven in the morning or after six at night. A lot of contracts run as split shifts – two hours at dawn, two at dusk, with a dead gap in the middle too short to go home and too long to sit around unpaid. That pattern grinds people down. They tolerate it while they must and drop it the instant a single daytime job comes up, and I don’t blame one of them.

What does turnover actually cost the contract?

The visible cost of replacing a cleaner is the advert and the interview. That’s the small part. The real bill hides in the fortnight after someone leaves, when the site runs short-handed and the people still there either pick up the slack or let the standard slip. Either way, that costs.

Then you recruit. In commercial work that usually means a DBS check and, for secure sites, a vetting process that can run weeks before the new person sets foot in the building. Through those weeks you’re covering a Farringdon floor with agency staff at a premium, or with your own supervisors doing shifts they shouldn’t have to. That costs too – and more than the wage you were trying to save.

The part nobody prices in is knowledge. A cleaner who’s worked a building for two years knows the boardroom gets used on Thursdays and that the client is particular about the glass doors. A replacement knows none of it and takes months to learn. In the gap, complaints go up, and a client who fields three complaints in a month starts wondering whether the contract’s worth renewing. Lose the contract and you’ve lost far more than a cleaner.

And there’s the redoing. A half-trained starter misses the underside of the desks or leaves streaks on the partition glass, a client photographs it, and now a supervisor drives across town to put it right on a Saturday. The job gets paid for once and done twice. On a tight-margin contract, a handful of Saturdays like that is the difference between a site making money and losing it.

The bill that never reaches the invoice

Every departure pulls a manager off the work that actually grows a business and drops them into recruitment and paperwork. Multiply that across a portfolio running thirty per cent annual churn and you’ve got a manager who does little else. Meanwhile the team that stays watches the churn and reads it correctly – if everyone keeps leaving, this must be a bad place to work – and a few of them start looking too. Churn breeds churn. I’ve watched a stable site come apart in a single quarter because two good people left within a fortnight of each other and took the mood of the place with them. The replacements cost me nothing on paper and a small fortune in practice – six weeks of wobble on the floor and a supervisor stretched so thin I nearly lost her as well.

How does churn show up on the client’s floor?

From the client’s side, turnover doesn’t announce itself as turnover. It shows up as inconsistency. One week the desks are spotless, the next the bins get missed, because a new starter hasn’t been shown the routine and nobody had time to teach them properly. The client doesn’t know they’re watching churn. They know the standard wobbles, and they start to lose faith.

There’s a trust cost sitting behind that. A building’s cleaning team holds keys and after-hours access to floors full of laptops and confidential paper. Every time the team changes, the circle of people with the run of your office at ten at night gets redrawn.

A revolving door is a security problem

There’s a reason the vetting process exists, and constant turnover puts it under real strain. When you’re desperate to fill a shift on a secure site, the temptation is to cut a corner – start someone provisionally, chase the paperwork later. Good contractors don’t. I’ve turned down last-minute cover rather than put an unvetted person into a solicitor’s office overnight. A client rarely sees that decision, but it separates a contractor who takes the keys seriously from one who doesn’t. Get that wrong once and you’re not managing a complaint any more. That costs in a way no invoice captures.

What do the good contractors actually do differently?

Everything above is fixable, and the contractors who manage it tend to do the same handful of things. They guarantee hours, so a cleaner knows what’s landing in their account at month’s end and can plan around it. Each site gets a named supervisor who actually knows the team, not a voice on a phone. There’s a proper induction, too – a walk round the site with someone who already knows it, rather than a laminated sheet and a bunch of keys – so a new cleaner starts the first shift knowing where the risers are and which floor the fussy tenant sits on. And they build somewhere to go – a good cleaner can become a supervisor and then an account manager, so the job turns into a career with a floor under it rather than a stopgap. When people can see a way up, they stay to climb it.

Pay is the argument I keep having

I pay the London Living Wage, and I argue about it with prospective clients more often than I’d like. The London Living Wage sits at £14.80 an hour for 2025, set by the Living Wage Foundation against the real cost of living in the capital. A client looking only at the monthly invoice sees the London Living Wage contractor as the expensive one. What they don’t see is that the cheaper quote has already priced in turnover. The low wage guarantees people leave, and the client pays for that leaving in missed bins and unfamiliar faces instead of in pounds on the invoice. I’d rather charge a bit more and send the same person to your building every week for five years.

Can you really build a stable cleaning team in London?

You can, and I have, but it takes deciding to. Stability comes from a dozen dull choices made in the staff’s favour rather than the invoice’s – luck has little to do with it. The awkward moment where it’s won or lost is the handover, when a contract changes hands.

The TUPE handover nobody plans for

Under TUPE, when a new contractor takes over a site the existing cleaning staff transfer across on their current terms. On paper that protects people. In practice it’s the most fragile point in the whole business, because the incoming contractor often inherits a demoralised team who’ve watched their old employer lose the contract and assume they’re next. The good contractors read that fear and move fast to settle it; the bad ones confirm it by trimming hours in the first week.

We took over a nine-floor block off Wellesley Road in East Croydon a little over two years ago – a mixed-tenant office with a cleaning team of six, every one of them braced to be managed out within the month. The previous contractor had run them on zero-hours and a rotating cast of supervisors. I did the opposite. I guaranteed everyone’s hours in writing in the first week and moved the whole team onto the London Living Wage. One supervisor went in and stayed. Five of those six are still with us. The sixth left to train as an electrician, which is the one kind of leaving I’m glad about.