Two firms take identical units on the same estate in the same month. Ten years later both hand back, and both get a dilapidations schedule with a five-figure cleaning line on it. One pays most of it. The other pays about a fifth, and the difference is a morning one of them spent with a phone camera before the desks went in.

That is the whole argument, and it takes half a day to act on. The reason so few tenants do is that the day you take a lease is the day you are least interested in the day you leave it.

What does a schedule of condition do, and when do you get one?

A schedule of condition is a dated record of the premises at the point the lease is granted. Written descriptions, photographs, sometimes video, prepared by a building surveyor and annexed to the lease itself.

Its purpose is to cap your repairing obligation. Most commercial premises in London are let on full repairing and insuring terms, and a repairing covenant on its own can require a tenant to hand back a building in better condition than they found it, which surprises people every time they hear it. A schedule of condition limits the obligation to the state evidenced in the schedule. Wear that was there when you arrived is not yours to put right.

Get one before the lease is signed. It has to be negotiated at heads of terms, because a landlord’s solicitor has no reason to volunteer it, and once the lease is executed the moment has gone. Older buildings and tired industrial units are where a landlord is most likely to concede one, since both sides know what the alternative argument looks like.

The wording decides whether it works

A schedule bolted onto the back of a lease that does not properly refer to it is decoration.

The lease has to say that the tenant’s obligation is qualified by the schedule, in terms that hold. There is a well-known difference between a covenant to leave the premises in the condition shown in the schedule and one to leave them in no worse condition than shown in the schedule, and the second is the one a tenant wants. That is a job for a property solicitor rather than for me, and it is worth the fee.

Budget somewhere between one and three thousand pounds for a surveyor to prepare a schedule on a typical London unit. Set that against a dilapidations claim and it is not a close call.

Why does the surveyor’s schedule miss the cleaning items?

Because a building surveyor is recording disrepair, and dirt is not disrepair.

A schedule of condition documents cracks, damp, failed seals, spalled concrete, corroded metalwork, worn coverings. It is a structural and fabric document, prepared to a professional standard, and it does its job well. What it rarely captures with any precision is the grime, and the grime is where a large share of the eventual cleaning claim comes from.

I have read schedules that photograph a whole kitchen in one wide shot showing the units are undamaged, with nothing recording that the extract grille was already furred, the splashback behind the hob already greased, the fridge seal already black in the folds. Ten years later a dilapidations surveyor prices a deep clean of all three, and there is nothing in the schedule to say they were like that on day one.

What fair wear hides in a kitchen

Kitchens and washrooms generate more dilapidations cleaning items than the rest of a commercial unit put together, and they are the two areas a condition survey treats most briefly.

Limescale on taps and down urinals in a hard water city. Grout discoloured across a tiled splashback. The seal round a sink gone amber. Extract grilles above head height, which nobody photographs because nobody looks up in a kitchen. Carpet tiles under where the desks will sit, which are the ones a surveyor will later compare against the walkways.

Photograph it dirty, and photograph it in detail, because a wide shot of a clean-looking kitchen is evidence against you rather than for you.

What should you photograph, room by room?

Work the unit systematically and do not stop at eye level.

Floors first, and by type: carpet tiles including the areas that will be covered by furniture, vinyl and its edges and seams, and any hard floor with its existing marking. In an industrial unit, the slab, in detail, including oil staining, joint condition and the marks left by the previous occupier’s racking.

Then walls and ceilings. Scuffs, patched paint, ceiling tiles that are stained or sagging, and one lifted tile showing the state of the void. Then windows and glazing, inside and out, including frames, trickle vents and any existing scratching. Then doors, ironmongery, kick plates and the manifestation on glazed partitions.

Then the two rooms that matter most. Every washroom, close up, taps and cisterns and grout and floor drains. Every kitchen and tea point, including the interiors of any equipment left behind, behind and under any appliance you can move, and the extract grille.

Then the outside. Loading bays, yards, external glazing, bin store, the wall where the previous tenant’s fascia sign was, drains and gullies.

Wide, close, and something for scale

Two shots of everything as a minimum. A wide frame that establishes which room and which wall, then a close frame of the defect itself. Neither is much use alone: the close-up proves the mark and the wide shot proves where it was.

Put something in the close frame for scale where the size matters, a tape measure or a coin. Leave the timestamp on and leave the originals unedited, because an edited file invites an argument about what else was changed. Shoot in daylight and take a second pass with the lights on, since some marks show under one and vanish under the other. A slow video walk-through, narrated with the room names, is a cheap backup that catches things your stills missed.

Three or four hundred photographs on a mid-size unit is a normal number, and the storage costs nothing. Nobody has ever wished they took fewer.

When is day one?

There are two of them, and tenants who do this at all usually miss the first.

Photograph the unit before your fit-out starts, in the state the landlord handed it over. That is the baseline your lease refers to. Then photograph it again once the fit-out is complete and before your people move in, which records what you added and what condition your own works left things in. The second set matters at the other end, when somebody is deciding whether a partition or a floor finish is yours to remove.

Fit-out contractors will not do this for you and are not asked to. The gap between the two sets is where most disputes about who caused what end up living, and on a Cat B fit-out the two sets can be a fortnight apart and show a room that looks nothing like itself.

Photograph it dirty

Now the part that argues against my own order book.

Tenants routinely book a deep clean of a unit before they move in. It is a reasonable instinct, the building looks better for it, and I am happy to quote for the work. What it also does, if you have not photographed the place first, is destroy your evidence and adopt the previous occupier’s dirt as your own starting point. You have cleaned the building to a standard nobody recorded, and you now own the difference.

So my advice is the awkward one. Take the photographs before you let anybody in with a machine, mine included. Then have the clean if you want it. The order costs nothing and it is the single most valuable half day in a commercial tenancy.

I have talked tenants out of booking that clean until the photographs were done, which delays my invoice by a week and has never once cost me the job.

How do you make the photographs count three years later?

By storing them somewhere a person who has left the company can still find, and by pairing them at the other end.

Keep the original files, unedited, with folder names matching the rooms and the date in the folder name. Two copies in different places. Put a note in the lease file saying where they are, because the person who took them will not be the person arguing about them, and a decade is long enough for a laptop, a cloud account and a facilities manager to all disappear.

The pairing is what makes them work. A dilapidations negotiation moves when a tenant can put a day-one photograph and an exit photograph of the same wall side by side. On its own the exit shot proves nothing; against a dated original it settles the question in one move.

The kitchen off Blyth Road

We cleaned a second-floor office off Blyth Road in Hayes, UB3, at the end of a nine-year term last spring. The dilapidations schedule that arrived carried a full deep clean of the kitchen, priced at a number designed to be paid rather than argued with, on the basis that the extract canopy and the tiling were heavily soiled.

They were. They had also been heavily soiled in 2017, and the tenant’s office manager had walked the unit with her phone before the fit-out and taken about sixty photographs, including four of the extract canopy with the grease already visible in the mesh.

The kitchen line came off the schedule inside a fortnight. Nothing else on that schedule moved, and the tenant paid the rest, because the rest was fair. Photograph it dirty, and you only end up arguing about the things you caused.

The office manager, when I mentioned it, said she had done it out of habit from a previous job and had assumed everybody did.

The call comes on a Monday, usually before nine. The office manager has walked in, found the kitchen in the state the Friday leavers left it, and wants to know why it was not cleaned. I ask what the specification says about the kitchen. There is a pause, some scrolling, and then she reads it out to me: kitchen areas to be cleaned as required.

That phrase has cost more money and more goodwill than any other three words in this industry. It commits nobody to anything. It cannot be enforced, cannot be priced straight, and cannot be argued about productively, because both parties can read it in their favour and both are right.

A cleaning specification is a document that tells a contractor what to do and tells a client what they have bought. Most of the ones I am sent do neither.

What is a specification for?

It is the contract’s operational half. The service agreement covers term, price, notice and liability. The specification covers the work, and it is the only part of the paperwork that describes what happens in your building on a Tuesday night.

That makes it your single point of control. Complaints, credits, audits, renewals and disputes all resolve back to the specification, because there is nothing else to resolve back to. A client who has not read theirs is negotiating from memory against a contractor who has read it very carefully.

It also sets the price. Every bidder builds a labour schedule from the specification, so the document determines what you pay before anyone quotes. A tender pack with a good specification and a mediocre service agreement will get you a better contract than the reverse, and legal review time tends to go entirely the other way.

Vague is cheap

Here is the mechanism, and it is the reason this matters more than the tidy-document argument.

A tender goes out with a loose specification. Six contractors price it. Five read it conservatively, allow for the kitchen being cleaned properly every night, and come in at a realistic number. One reads it thinly, allows two and a half hours where the others allowed three and a half, and comes in twenty per cent under.

The thin bidder wins, because on paper all six offered the same service. Vague is cheap, and vagueness in a specification is not a drafting weakness so much as an invitation. The client did not choose a worse service. They wrote a document that could not tell the difference.

Which phrases should never appear in one?

Strike these on sight. As required. As necessary. Regularly. Periodically. To an acceptable standard. When needed. Spot clean as appropriate.

Each of them moves a decision from the document into somebody’s head, and the head it moves into belongs to whoever is under the most pressure at the time. On a well-run site that is a supervisor doing their best. On a badly run one it is a cleaner with forty minutes left and three floors to go.

Replace each with a frequency and a scope. Not “washrooms cleaned regularly” but “all washrooms cleaned and restocked once per shift, five shifts per week, with a second check of ground floor washrooms at 13:00 on weekdays”. Long, dull, and worth every word of it. The test for any line is whether two people reading it on a bad Friday would carry out the same work.

Why contractors like a vague specification

I will admit something about my own side of this. When a loose specification lands on my desk, part of me is pleased.

A vague document gives a contractor room. Room to flex hours when somebody is off sick, room to decide that the boardroom does not need doing on a quiet week, room to argue after the fact that what happened was within scope. It also lets a contractor price to win rather than price to deliver, which is a temptation the whole trade lives with.

Firms that intend to do the job properly would rather have a tight specification, because it protects them from the bidder who does not. I have lost tenders to companies whose price only worked because the document let it. A tight specification is the client protecting themselves and, incidentally, protecting the honest bidder.

What belongs in the daily schedule?

Structure it by area rather than by task, because that is how a building is used and how a cleaner moves through it.

Break the site into zones with different requirements: reception and front of house, open-plan office, cellular offices and meeting rooms, washrooms, tea points and kitchens, circulation and stairs, lifts, stores and back of house. Each zone gets its own task list with a frequency against every line, using a plain notation: daily, three times weekly, weekly, monthly, quarterly, six-monthly, annually.

Then say what “clean” means for the items where it is contestable. Desks, for instance. Is the cleaner clearing paper to wipe underneath, or wiping around what is left out? Both are defensible; only one is what you think you are buying. Same with keyboards, screens and telephones, which carry a damage liability that makes many contractors avoid them unless instructed.

List the exclusions explicitly. Consumables. Windows above ground floor level. External areas. Specialist floor treatments. Interiors of fridges, microwaves and dishwashers. Plant rooms and comms rooms. An exclusion in writing prevents an argument; an exclusion assumed guarantees one.

State the hours, not just the tasks

The task list tells you what. The resource line tells you whether it can be done.

Your specification should state the number of operatives on site, the shift start and finish times, the total hours per week, the supervisor’s presence and visit frequency, and the arrangements for absence cover. Without those numbers you cannot compare two bids at all, because the labour schedule is where nearly all of the cost sits.

A client who knows their building has three and a half hours of work in it, and reads a bid offering two and a half, has learned everything they need to know about that bid without opening the pricing page. Vague is cheap, and the hours line is where vagueness gets expensive.

How should periodics be written?

With months against them.

Carpet extraction, hard floor stripping and resealing, high-level work, washroom deep cleans, kitchen deep cleans, upholstery, light fittings, internal glass at height: these are the jobs that keep a building from ageing badly, and they are the first things to fall out of a contract because nothing forces them to happen on a given date.

“Annually” is not a date. “Carpet extraction to all open-plan areas in the week commencing the second Monday in February” is. Name the month for every periodic, say who books it, say whether it sits inside the monthly charge or gets quoted separately, and put the whole schedule on one page that both parties look at each quarter.

The annual that never happens

Watch how it goes wrong, because the pattern is identical everywhere.

Year one, the periodic schedule is fresh and somebody chases it. Year two, the carpet clean is postponed because a project is running on that floor, and it does not get rebooked. Year three, nobody remembers it was in the contract. Year five, the carpet needs replacing at a cost twenty times the extraction that was skipped, and the tenant is arguing about whether that counts as fair wear at lease end.

The remedy is a named person on the client side who signs off each periodic as completed, and a line on the monthly report showing what is due next quarter. It takes four minutes a month.

How do you check any of it is happening?

Audit, jointly, against the document you wrote.

A workable audit is a scored walk of sampled areas, using the same zones as the specification, with defects categorised by severity rather than counted flat. Once a month, contractor and client together, an hour with a clipboard. Score it, agree the actions, date them, review them next time.

What I would avoid is an elaborate penalty regime with points and credits. Those exist mostly to give procurement something to put in a report, and their practical effect is to make a contractor manage the scoring rather than the building. I would rather a client walked the site with me monthly and told me plainly what was wrong. That is a less comfortable arrangement for me than a spreadsheet, and it produces a cleaner building.

Colour coding, chemical dilution control and method statements belong in here too, referenced to a recognised standard rather than described from scratch. Naming the standard is enough. A specification that reproduces half a training manual is one nobody will open twice.

What we rewrote off Mollison Avenue

A two-building site off Mollison Avenue in Enfield, EN3, came to us last year with a specification running to thirty-one pages, most of it inherited from a facilities template and some of it describing rooms the client no longer occupied. The phrase “as required” appeared nine times. There was no resource line anywhere in it.

We rewrote it to eleven pages. Six zones, a frequency against every task, an exclusions list, the hours and shift pattern stated on the first page, and a periodic schedule with twelve named months across two years. The client cut the meeting rooms from daily to three times weekly, because they were used twice a week, and put the saved hours into the washrooms and the two kitchens, which had been short for years without anyone being able to point at a document and prove it.

The monthly charge went up by about four per cent, which the client agreed to before they had seen the new document, on the strength of the hours line alone.

Nobody.

That is the honest answer in most of the buildings I work in, and it is the reason bin stores are the worst-run room in commercial property. The landlord thinks the managing agent has it. The managing agent thinks it sits in the service charge with the waste contractor. The waste contractor collects what is presented and has no view on how it got there. The tenants think it stops being theirs at the office door. And the cleaners, who are the only people who go in there daily, are following whatever instruction they were given by somebody who left two contracts ago.

Meanwhile the law is unambiguous about who is responsible, and it is not the person you would guess from watching the building operate.

Who is the waste producer in a multi-tenant building?

Each tenant is. Waste is produced by the business that generates it, and the duty attaches to the producer at the point of production. A firm on the third floor that fills six sacks a day with paper, coffee grounds and sandwich packaging is a waste producer with duties under section 34 of the Environmental Protection Act 1990, whatever arrangements exist downstairs.

Those duties are to store waste securely, to prevent its escape, to transfer it only to an authorised person, and to describe it accurately enough for the next holder to handle it lawfully. They do not evaporate because somebody else booked the collection.

The bag has an owner, and the owner is the business that filled it.

The service charge does not transfer the duty

Most London multi-lets run a building-wide waste contract, arranged by the managing agent and recharged through the service charge. That arrangement is sensible and it is what I would recommend to any agent. What it does not do is move the legal duty off the tenant.

The tenant is still transferring waste to somebody. If that transfer is not covered by a documented arrangement naming them, they have a gap. The usual fix is straightforward: the building’s waste transfer documentation should list the tenants whose waste is covered, and each tenant should hold a copy of the paperwork covering their own material.

I have asked to see that paperwork in perhaps thirty buildings over the years. It existed and covered the tenants properly in a handful.

What does duty of care require of a tenant who never touches a bin?

Three things, none of them difficult.

Check that whoever removes the waste is registered to do so. Waste carriers appear on the Environment Agency’s public register, it is free to search, and checking takes about a minute. A managing agent should have done this; a tenant relying on the agent should have evidence that they did.

Describe the waste properly. For ordinary office material this is not onerous, but food waste, confidential shredding, electrical items and anything from a workshop or kitchen all behave differently in the paperwork, and a description of “general waste” covering all of it is inaccurate.

Keep the records. Waste transfer notes are retained for two years, hazardous waste consignment notes for three. An annual transfer note covering a regular collection of the same material is permitted and is what most buildings run on, which means the whole obligation amounts to one document a year and a place to keep it.

Where the paperwork breaks in practice

The break is almost always at the handover between building and tenant rather than between building and waste contractor. The waste contractor’s paperwork is usually in order, because that is their regulated business and the Environment Agency knows where they live.

What is missing is anything showing which tenants are covered, from what date, for what materials. Tenants come and go, the schedule attached to the waste contract was written in 2019, and the coffee roastery that took the ground floor unit last year is presenting a waste stream nobody has documented. If an inspector asks, the agent produces a contract and the tenant produces nothing.

The bag has an owner, and paperwork is how that ownership gets proved after the fact.

Why does the shared bin store fail?

Because segregation is a behaviour, not a bin.

Since 31 March 2025, workplaces in England with ten or more full-time equivalent employees have had to present dry recyclables and food waste separately from residual waste. Dry recyclables means plastic, metal, glass, paper and card, with paper and card often kept apart from the rest depending on what the collector will take. Micro-firms with fewer than ten full-time equivalent staff come into scope on 31 March 2027, and the headcount is counted across the organisation rather than per site, so a small London office belonging to a larger company was in scope from the start.

In a single-occupier building this is a manageable internal exercise. In a multi-let with eleven tenants, a shared store and a propped door, one tenant tipping food waste into the dry recyclable bin contaminates the load for everybody. The contaminated load is rejected or recharged at residual rates, and the cost goes into the service charge, where it is spread across the tenants who did it correctly.

Nobody is ever identified, because nobody watches the door.

The residential bins next door

Mixed-use blocks are worse, and London is building a great many of them. Flats and offices in the same structure, often sharing a service yard, with the residential waste collected by the local authority and the commercial waste on a private contract.

Commercial waste in a residential bin is unlawful, and it happens constantly, usually without anybody intending it. A small office on the ground floor of a residential block finds the flats’ bin store nearer than their own, and a cleaner carrying six bags at seven in the evening takes the shorter walk. The council’s collection crew then takes trade waste it is not contracted to take, and the block’s residents pay for the overflow.

If you manage one of these buildings, the two stores need to be physically separate and obviously labelled, and the commercial store needs to be the convenient one. Design out the shortcut or you will be policing it forever.

Is your cleaning contractor allowed to move the waste?

Within the building, yes. A cleaner carrying bags from an office to the bin store is moving a producer’s waste around the producer’s own premises, which is not carriage.

Off site is a different matter, and here I will say something about my own trade that it would rather I did not. A cleaning firm that loads waste into a van and takes it away needs to be a registered waste carrier, and a substantial number of small operators doing exactly that are not. It comes up most on one-off jobs: a clearance, a strip-out, an end-of-tenancy where the outgoing tenant leaves furniture behind. The cleaning crew loads it up, and where it goes next is a question nobody asks until it turns up dumped on a verge with an invoice in one of the bags.

Ask your contractor for their waste carrier registration number before they remove anything from your site. If they hesitate, that tells you what you needed to know.

What “removal of waste” means in a specification

Cleaning specifications use the phrase constantly and define it almost never.

It should say who moves waste from the office to the store, at what times, into which streams, and who is responsible for presenting the correct containers for collection. It should say who reports a store that is full, blocked or fouled, and to whom. It should say whether the cleaning team handles food waste caddies from the tea points, which is the item most often missed and the one that smells first.

Where none of that is written down, the cleaners do a version of it invented on site by a supervisor, and it works until that supervisor leaves.

Who should own the bin store, and what should the contract say?

One named party, with hours priced against it.

That party can be the cleaning contractor, and often should be, since we are in the building anyway. It can equally be the waste contractor providing a porter, and on a large building with heavy volumes that is frequently cheaper than my hourly rate. I have told agents as much and lost the work, because a dedicated waste porter doing three hours a day is better value than my supervisor doing it badly around a cleaning round.

What cannot happen is the current arrangement in most buildings, where the store belongs to nobody and gets attention when somebody complains.

The store off Chiswick High Road

A managing agent brought us into a mixed-use block off Chiswick High Road, W4, two summers ago. Six commercial units at street level, flats above, and a shared service yard with the residential and commercial bins along the same wall.

The complaint was smell, in July, which is when these complaints arrive. What we found was a food waste caddy from a café being emptied into the residual bin because the food waste bin was consistently full by Tuesday, residual bags stacked outside the containers by two of the offices, and the flats’ bins holding a steady quantity of commercial cardboard.

The fix was not a cleaning fix. Collection frequency for food waste went from weekly to twice weekly, the commercial bins moved to the near end of the yard and the residential to the far end, the store got labelled properly, and each unit got a single side of A4 saying which bin took what. Our contract picked up a fifteen-minute presentation check every morning, priced, written down and assigned to a named person.

The smell went within a fortnight, which is faster than anybody expected and had almost nothing to do with cleaning. What matters more is that when the agent was asked for the building’s transfer documentation last autumn, the schedule listed every tenant and the dates they took occupancy. The bag has an owner, and in that building somebody can now say who.

Walk into a fit-out the week the trades leave and you can taste it. Gypsum sits on the back of your throat, everything horizontal carries a fine grey film, and the glass has that dull bloom that makes a brand new office look second hand. The lights are on, the joinery is in, the place is finished, and it looks wrong.

Getting from that room to the room in the architect’s visuals is three separate jobs, priced separately, booked at different points, and routinely confused with one another by everyone involved. The confusion is where the money goes missing. A tenant assumes the fit-out price included the finished article. The main contractor assumed the tenant was handling the last stage. Somebody ends up paying twice, and quite often it is the client.

What are the three cleans, and which one are you paying for?

The builders’ clean, sometimes called the rough clean or first clean, happens while the site is still a site. It is heavy work: debris and packaging out, plaster and screed splashes off, sheeting lifted, high-level dust knocked down, floors scraped and swept. It is usually carried out by the main contractor’s own operatives and priced into the construction contract.

The sparkle clean is the detailed one. All trades off site, the building empty, and a team working top down through every surface: glass and frames, ironmongery, ceiling grid, light diffusers, skirtings, sanitaryware, floor finishes brought up to their intended appearance. This is the clean that produces the visuals.

The snagging clean, or touch-up clean, comes after the snagging works. Joiners, decorators and electricians return to put right whatever the client marked up at inspection, and they bring their own mess with them.

Where Cat A ends and Cat B begins

London fit-outs run in two stages and the cleaning follows the split. Cat A is the landlord’s base build: raised floor, suspended ceiling, mechanical and electrical services, basic finishes. Cat B is the tenant’s fit-out, which puts partitions, joinery, kitchens and meeting rooms into that shell.

Which means most floors get cleaned, then have holes cut in the ceiling and partitions built through them, then get cleaned again. A landlord handing over a Cat A floor has usually paid for a sparkle clean, and that clean is worth very little to the incoming tenant, because the Cat B works will undo it within a fortnight.

Tenants pay for a Cat A sparkle clean in their rent and then pay for a Cat B one out of their fit-out budget. It looks like duplication and it is not.

Why can’t it all be done in one go at the end?

Because dust falls, and it keeps falling for days after the work that made it has stopped.

Dust falls slowly. Fine gypsum and screed stays suspended in still air for a long time and gets disturbed again by every door opening and every trolley wheeled across the floor. Clean a partition on Tuesday while a first-fix electrician is drilling above the ceiling grid and you have cleaned nothing. The material relocates onto the surface you have just finished.

There is a second reason, and it is about what dust does to finishes. Cement and plaster residue left on glass and polished stone for weeks is alkaline and abrasive, and it etches. Screed splash left on new vinyl hardens into something that then needs a scraper, and a scraper on new vinyl is a decision nobody wants to make. The rough clean exists to get contaminating material off surfaces early, before it becomes damage.

The dust in the ceiling void and the floor void

Two spaces on every London fit-out go uncleaned as a matter of routine.

Above the ceiling grid, on top of the tiles and along the cable trays, there is whatever the construction left behind. Nobody sees it. It stays there for the length of the lease, feeding into the return air path, and it turns up years later on a dilapidations schedule or during the next strip-out when someone lifts a tile and a decade’s worth of construction dust comes down on the desks.

Below, the raised floor void collects offcuts, screws, cable ties, sandwich wrappers and a startling depth of fines. Floor tiles get lifted for a data pull three years on and the void is filthy, which matters because the void on many buildings is part of the air distribution route. Ask for both voids to be included in the sparkle clean specification. They almost never are unless somebody names them.

What separates a sparkle clean from a wipe round?

The specification, and whether anyone reads it with a torch afterwards.

A proper sparkle clean works top down through the whole volume. Ceiling grid faces and tile edges, light diffusers inside and out, ductwork grilles, and the tops of partitions and door frames where the last of the airborne dust lands. Then vertical surfaces, glass, ironmongery, switch plates. Then joinery, including the inside of every cupboard and drawer, which trades treat as somewhere to leave things. Then sanitaryware, with labels and protective film removed and adhesive residue taken off rather than left as a grey smear. Then floors, last, with the right treatment for the finish.

Silicone smears on glass, plaster in the track of a sliding door, paint speckle on ironmongery, labels still on the taps: those four failures account for most of the snagging items a cleaning contractor gets marked up on.

The glass is where fit-out cleans get ruined

More glazing is damaged during fit-out cleaning than at any other point in a building’s life, and once it is damaged the only remedy is replacement.

Cement and plaster splash on glass bonds and etches. If it is left long enough, the mark is in the glass rather than on it. The temptation at that point is a blade, and blades on modern glazing are how you turn a stain into a permanent scratch pattern. Coated glass, low-emissivity units and anything with an applied film will not tolerate scraping at all, and a lot of London glazing is coated in some way the cleaning operative cannot see.

The answer is chemical rather than mechanical, applied early, before the splash cures. Which means the rough clean matters more than anyone gives it credit for, and a fit-out contractor who skimps on it is storing up a glazing replacement bill for somebody.

Manifestation film applied over glass that has not been properly cleaned is its own small disaster. The dust is now permanently under the film.

When should each clean be booked, and what goes wrong with the dates?

Rough clean while the works run, with at least one pass after first fix and one after the messy wet trades. Sparkle clean once every trade is off site and no further construction work is planned, which is later than the programme claims and needs a day or two of settling time after the last activity. Snagging clean after snagging is signed off, and before furniture arrives if you can manage it.

What goes wrong is the programme. Fit-out programmes slip at the front and never at the back, because the client’s move-in date is fixed by a lease expiry somewhere else. The compression lands on the last fortnight, which is where the cleaning sits.

So the sparkle clean gets booked for Thursday, a joiner comes back on Friday to hang a door that arrived late, and the desks land on Monday morning on top of a floor that has sawdust on it.

The snagging clean nobody budgeted for

Of the three, the snagging clean is the one that goes unfunded most often, because it appears on nobody’s original scope. The construction contract covers the builders’ clean. The fit-out budget covers the sparkle clean. Then snagging generates a fresh round of drilling and painting, and the office picks up a new coat of fine dust in the week before the staff arrive.

Put a figure in the budget for it at the start. Something in the region of a fifth of the sparkle clean cost, held back and spent at the end, saves the argument about who pays for it in the week everyone is at their worst.

Furniture install is the same story: cardboard, protective film and the dust off a hundred boxes, all landing after the clean that was meant to be final.

Does every fit-out need all three?

No, and I turn down work on this basis several times a year.

On a small Cat B job, say a floor under 5,000 square feet with light partitioning and no wet trades, one properly resourced clean after the trades leave, plus a touch-up before occupation, is sufficient. Selling a three-stage programme on a job that size is selling stages. The point of the sequence is to manage dust generated over weeks by multiple trades, and a small refit does not generate it.

What I will not do is a sparkle clean on a site that still has trades working. The client pays for a finish that is undone within a day, then pays somebody again the following week. If your programme has slipped, move the clean and move the furniture delivery, and if the furniture cannot move, accept that you are buying a rough clean and a touch-up rather than a sparkle.

What happened off Featherstone Street

A tenant took a fifth floor off Featherstone Street, EC1V, last spring, 9,000 square feet, Cat B with a decent joinery package and a coffee bar. The programme lost eleven days to a switchgear delivery and the move-in date could not shift.

We were booked for the sparkle on a Wednesday with the client’s staff arriving the following Monday. On the Tuesday I walked the floor and there were two decorators still on the meeting rooms and a shopfitter waiting on a glass panel. Doing the clean as booked would have been theatre.

What we did was a heavy second rough clean on the Wednesday, the sparkle on the Saturday once the shopfitter had finished, and a two-person touch-up on the Sunday evening after the furniture install. It cost about fourteen per cent more than the original quote and was the cheapest version available by then.

Dust falls for three days after the last drill goes quiet. Everything in the programme has to be arranged around that, or it gets arranged around it anyway, at the client’s expense.

Every tender document I read promises the client a fresh start, and every one of them is written by someone who knows perfectly well that the same six people will be mopping the same floor on the Monday after the changeover. That gap between what the market sells and what the law delivers is the single most expensive misunderstanding in commercial cleaning procurement, and clients walk into it two or three times a decade.

TUPE is usually explained as employment protection, which it is. It is also, from where a building manager sits, a set of commercial facts about what you are buying, what you cannot buy, and what your outgoing supplier can do to you on the way out.

Does TUPE apply when you change cleaning contractor?

Almost always, and the mechanism is worth knowing by name. The Transfer of Undertakings (Protection of Employment) Regulations 2006 cover business transfers and, separately, service provision changes. A service provision change is what happens when a contract for a service moves from one provider to another, or comes back in house, or goes out for the first time.

Cleaning is the textbook case. There is an organised grouping of employees whose principal purpose is serving that client at that building, the activities carried out after the changeover are fundamentally the same activities, and so the staff assigned to the site transfer automatically to the incoming contractor on their existing terms, with their continuous service intact.

You cannot contract out of it. A clause in your tender inviting bidders to supply their own team does not disapply the regulations, and a bidder who cheerfully agrees to that clause is either inexperienced or planning something you would rather not be party to.

What the client signs up to without signing anything

The client is not the employer and does not transfer anybody. What the client does is trigger the whole thing by awarding a contract, and then live with the consequences in three places.

Price, because the incoming bidder has to fund whatever terms arrive. Continuity, because your building keeps the team it had rather than getting a new one. And risk, because a botched transfer generates tribunal claims, and the parties to those claims will be reaching for whatever the tender documents and the service agreement said about co-operation and indemnities.

The people do not change. Hold that thought through everything below.

What arrives with the transfer, and what does it cost you?

Contracts of employment transfer as they stand. Hourly rates, contracted hours, shift patterns, notice periods, accrued holiday, continuous service dating back to whenever that cleaner first started on your site under whichever contractor. Liabilities transfer too, including outstanding grievances, disciplinary processes and live tribunal claims connected with the employment.

The incoming contractor cannot tidy this up. Changing terms to the employee’s detriment because of the transfer is void, and levelling everyone down onto a house contract is precisely the thing the regulations prevent. So a bidder inheriting a cleaner on an above-market rate with twelve years’ service and an enhanced sick pay arrangement has to carry it, and it goes into their price.

Two consequences for the client. Your new contract will not be cheaper than the arithmetic of the transferring payroll allows, whatever anybody’s sales director says. And if your outgoing contractor has been generous, or careless, that generosity is now yours to fund.

The list, and why it grows

Outgoing contractors are required to provide employee liability information to the incoming one no later than 28 days before the transfer. Names, ages, terms, disciplinary and grievance history over the past two years, claims, collective agreements.

Watch that list. In a market this competitive, a contractor who has just lost your site has a modest interest in making the site expensive for whoever won it, and lists have a habit of arriving longer than expected. The area supervisor who visited fortnightly appears as assigned to the building. Somebody on long-term sick materialises. A mobile operative who covered your washrooms on Thursdays turns out to have been dedicated to you all along.

Some of those additions are legitimate and some are chancing it, and the test is whether the person was in fact assigned to the organised grouping serving your site. Get the list early, question it in writing, and make sure your current contract obliges the incumbent to provide accurate information within a stated period. That clause costs nothing to insert when you sign and is worth a great deal when you leave.

What should the client do before the tender goes out?

Sort three things before a single bid document leaves the building.

Establish the true staffing picture. Headcount, contracted hours per person, hourly rates, shift patterns, length of service, and any enhancements. Bidders cannot price your site properly without it, and a tender issued without it produces quotes that all get revised upwards after award, which wastes everyone’s spring.

Decide your wage position and say so. If you want the London Living Wage on your site, put it in the specification rather than hoping. The rate rose to £14.80 an hour for 2025-26 and accredited employers had until the first of this month to have it in place across their workforces, contracted staff included. A client who mandates it gets it priced properly. A client who mentions it in the interview and leaves it out of the specification gets the version somebody hopes they will not check.

And leave time. Which brings me to the mistake I see most.

Six weeks between award and start

The transfer needs a measures letter, a consultation with the affected staff or their representatives, employee liability information at least 28 days out, and enough breathing room for the incoming contractor to meet the team, sort payroll details, arrange vetting and order kit.

Clients routinely award a cleaning contract three weeks before they want it live, because procurement ran late and the start date was fixed in January. Failure to inform and consult carries a protective award of up to thirteen weeks’ pay per affected employee, and while that liability sits with the contractors rather than with you, the disruption lands squarely on your floor.

Six weeks between award and mobilisation is comfortable. Four is workable. Anything under three is a decision to have a bad handover.

Why does the new contractor’s first month go badly?

Because the team arriving on day one has just watched their employer lose the contract, and every one of them assumes they are next.

That is not paranoia. Incoming contractors do sometimes trim hours in the first fortnight, quietly, hoping nobody counts, and the trade knows it. So your cleaners turn up on the Monday demoralised and job-hunting, and the ones with options start leaving in week three. The people do not change, but the ones who leave take the building knowledge with them, and that is where the standard wobbles. A handover that loses three of seven cleaners in a month has cost you more than the tender saved.

There is a second reason, less discussed. The new contractor has inherited a rota built by somebody else for a specification that has probably drifted from what the building now needs. Working out what is going on takes a month of visits, and during that month things get missed.

The induction that gets skipped

The single best predictor of a smooth handover is whether the incoming contractor spends real time with the transferring team before day one rather than after it.

Not a letter. A supervisor in the building, walking the floors with the people who have cleaned them for years, asking what the specification gets wrong. Guaranteeing hours in writing in the first week rather than leaving people to wonder. Getting names right.

It costs a few hundred pounds of supervisor time and it is the difference between keeping five of six and keeping two. I have never regretted spending it and I have twice regretted rushing it.

Should you change contractor at all?

Here is the part that argues against my own tender pipeline.

Most of the sites I am invited to bid for do not have a contractor problem. They have a specification problem, or a price problem, or a communication problem, and all three are cheaper to fix with the incumbent than by running a procurement exercise and a transfer. If the cleaning has slipped because the hours were cut two renewals ago, changing supplier changes nothing at all, because the same hours and the same people arrive under a different logo. You will have spent three months and a great deal of goodwill to buy yourself the situation you already had.

Change contractor when the problems are managerial rather than structural: no supervision, no cover, no response to complaints, invoices that never match the specification. Those do improve with a new firm. A shortfall in hours does not, unless you also fund the hours.

The government launched a call for evidence on reforming these regulations in April, closing in July, so some of the detail above may move over the next couple of years. The underlying arithmetic will not.

What we did off Fenchurch Street

We took over a mixed-tenant block off Fenchurch Street, EC3M, last year, inheriting seven cleaners and a specification nobody had reviewed since 2019. The list we were given had nine names on it. Two of them, a mobile supervisor and a window operative, were not assigned to the site in any meaningful sense, and we said so in writing before the transfer rather than after.

The people do not change, so we spent the money on the handover instead of the mobilisation paperwork. The seven who did transfer got their hours confirmed in the first week and a supervisor who walked the building with them before day one. Six are still there. The specification we inherited turned out to have two of the eight floors being cleaned nightly that were empty, and one that was cleaned twice a week and shouldn’t have been.

We moved the hours across rather than handing them back. The client’s invoice stayed the same, which is not a story anybody puts in a tender document.

A seated adult breathes out somewhere around 20 litres of carbon dioxide an hour. Put forty of them on a floor with the fresh air turned down to save on heating, and by three in the afternoon the room is running at two and a half times the concentration of the street outside.

That is the number behind every stuffy meeting room complaint I have ever been asked about, and it has nothing whatever to do with cleaning. I get asked anyway, because the cleaner is the only supplier most tenants see, and because the word hygiene has stretched over the last few years to cover things it was never meant to cover. So this is a piece about where the line falls between my trade and the mechanical engineer’s, written by someone who has an obvious commercial interest in drawing that line generously and is going to draw it straight instead.

What is a CO2 reading telling you?

Carbon dioxide at office concentrations is a proxy rather than a hazard. People exhale it, ventilation dilutes it, so the level in a room tells you how much outdoor air is arriving relative to how many people are in there. It is the cheapest useful measurement in building services.

The reference points are well established. CIBSE takes 600 to 800 parts per million as the mark of a reasonably well ventilated room, and treats anything above 1,500 ppm as a very poorly ventilated space. The HSE’s position runs alongside it: readings consistently over 1,500 ppm in an occupied room mean the ventilation needs attention, and roughly 1,000 ppm corresponds to about 10 litres per second per person. That figure is also the whole-building standard in Part F, which asks for 10 litres per second per person or one litre per second per square metre, whichever comes out higher. The HSE’s Approved Code of Practice sets a lower floor still, saying fresh air should not normally fall below 5 to 8 litres per second per occupant.

Outdoor air in London sits somewhere over 400 ppm, so you are never starting from zero.

1,500 ppm is not a poison reading

Tenants who buy their first monitor tend to panic at the wrong thing. A reading of 1,400 ppm is not making anyone ill in the toxicological sense. Occupational exposure limits for carbon dioxide sit at thousands of parts per million, far above anything an office will produce.

What that reading tells you is that the air in the room is being recycled through the lungs of the people in it, and that everything else in the air is concentrating at the same rate. Volatile compounds off the furniture, cooking smells, whatever came in off the road, the aerosols from the person with a cough by the window. Carbon dioxide is the receipt for all of it.

The air is not a surface, and that distinction runs through the rest of this article.

Can cleaning change the air in an office?

Yes, at the margins, and the margins are worth having.

Most of what floats around an office floor has been on the floor first. Settled dust is a reservoir: skin cells, fibres, the fine grit that walks in on shoes, and the particulate that came off the South Circular and through the door. Footfall lifts it, and a badly run cleaning operation lifts more of it than the occupants do.

Damp dusting rather than dry keeps particulate captured instead of redistributed. Microfibre holds fines that a duster flicks into the air. Regular attention to the horizontal surfaces nobody sees, particularly above head height, removes the reservoir before it can be disturbed. Kitchen and washroom extract grilles clogged with grease and lint reduce the extract rate they were designed for, and cleaning them restores something a mechanical contractor would otherwise be called out to diagnose.

None of that changes the ventilation rate by a single litre per second. It changes what is available in the room to be circulated by it.

The vacuum is the machine that matters

If a client wants one cleaning intervention that shows up in a particulate reading, it is the vacuum specification.

A cylinder machine with a tired bag and a poor exhaust filter collects the coarse material and fires the fine fraction straight back out at ankle height, which is precisely the fraction that stays airborne longest and reaches deepest into the lung. HEPA-filtered machines with sealed bodies do not do this. The seal matters as much as the filter grade, since air that bypasses a filter through a badly fitting housing has not been filtered at all.

Ask your contractor what machines are on your site and what the filtration is. It is a fair question, most specifications do not answer it, and the difference in cost between the right machine and the wrong one is a few hundred pounds spread over several years.

Where does cleaning make the air worse?

Here is the part of the trade that does not get discussed at facilities exhibitions.

Cleaning chemicals are a source of volatile organic compounds. Solvent-based products, aerosol sprays, and anything strongly fragranced put measurable quantities of material into the air of a room, and a floor cleaned at seven in the evening with the air handling switched off at six sits in its own emissions until the plant restarts in the morning. The first arrivals breathe the concentrated version.

Trigger sprays make it worse than it needs to be by aerosolising product that could have been applied to a cloth. Decanting into a damp cloth and wiping, rather than misting a surface at head height, is a change that costs nothing and takes a real quantity of airborne material out of the room. Where a building runs air cleaning devices as well, the interaction between what those produce and what we spray deserves more scrutiny than it gets.

The scent that gets mistaken for hygiene

My industry sells smell as evidence of work, and I think it is one of the worse habits we have.

Plug-in air fresheners, scented washroom units, citrus-heavy floor products: all of them exist because a client walks in and judges the clean by the nose. What they are doing chemically is adding compounds to the air of a room that was already short of fresh air, and calling the result improved. Limonene, the compound behind most citrus fragrance, is reactive, and the chemistry it takes part in indoors is not something anyone can wave away as pleasant.

A properly cleaned office smells of nothing. That is the target, and it is a harder one to hit than a floor that smells of lemons.

What should be measured, and what should be ignored?

Buy a CO2 monitor and take it seriously. Buy one with a non-dispersive infrared sensor, which measures carbon dioxide directly. The cheap end of the market sells devices reporting something labelled eCO2 or equivalent CO2, which is inferred from a volatile compound sensor rather than measured, and the numbers those produce in a room where somebody has just cleaned or opened a marker pen are close to meaningless.

Log over a fortnight rather than glancing at a display. What you want is the shape of the day: where the curve sits at nine, at eleven, after lunch, at four, and whether it ever comes back down. A room that starts at 500 ppm and finishes at 1,600 ppm every day has a ventilation rate that was designed for fewer people than are now sitting in it.

Ignore anything sold on the promise of sterilising your air, at least until it comes with dose figures and independent room-scale data rather than a laboratory certificate.

Where the sensor goes

Not on a desk beside somebody’s face, where their own breathing dominates the reading. Not next to a window or a door, where dilution flatters it. Not on top of a warm server cabinet.

Roughly breathing height, a metre or so off the floor, in the occupied body of the room, away from supply diffusers. One sensor per distinct space, since a meeting room and the open floor outside it behave nothing alike. Put a sensor in the wrong place and you will get a number that is precise, repeatable and about the wrong volume of air.

Who should you call about an air quality complaint?

Not me, most of the time.

Persistent stuffiness, headaches through the afternoon, a floor that everybody describes as airless: those are ventilation problems and they belong to a mechanical services contractor. The questions to ask are whether the air handling unit is running at its designed fresh air fraction or has been dialled back for energy, when the supply filters were last changed and to what grade, whether the extract is balanced against the supply, and whether the occupancy of the floor has risen since the system was commissioned. That last one catches a lot of London buildings, where a floor designed for eighty people now holds a hundred and thirty desks.

Spend the money there before you spend a penny with me. The air is not a surface, and a cleaning contract is not a ventilation strategy however it is dressed up.

What we changed off Clarence Street

A tenant on a second floor off Clarence Street in Kingston, KT1, called us about air quality last autumn, convinced the cleaning was the problem. It was not. Their monitor was showing 1,700 ppm by mid-afternoon in a room with sealed windows and an air handling unit somebody had set back years earlier.

We did make three changes on our side. Machines with sealed HEPA filtration replaced two elderly vacuums. The plug-in fragrance units came out. And the cleaning slot moved to start while the plant was still running rather than an hour after it shut down, which cost the landlord about forty minutes of fan power a night.

The 1,700 ppm did not move at all. That took a controls engineer and a fresh air damper that had been shut since 2019. The air is not a surface, and the cleaner is rarely the person holding the answer.

Before you do anything else about a lease that ends this year, go and stand in the middle of your unit and look down.

Not at the walls, which will be repainted by somebody for a known price. Not at the roller shutter, which either works or gets repaired. Look at the slab, and specifically at the parts of it you have not seen in years – under the racking, behind the pallet stacks, in the strip where the forklift turns. That is where your dilapidations bill lives. In a warehouse or a light industrial unit, the floor routinely accounts for more of the final settlement than every other item on the schedule put together, and it is the one part of the building that a tenant has almost never thought about since the day they moved in.

Why does the floor decide the bill on an industrial handover?

An office hands back on the strength of its finishes. Carpet, paint, ceiling tiles, all of them replaceable at rates a surveyor can look up. An industrial unit is mostly a steel frame, a roof and a slab, and two of those three are not yours to damage.

The slab is. It is the only surface in the building that takes the entire weight of your operation for the whole of the term, and unlike everything else it cannot be swapped out at the end for a few thousand pounds. A slab that has been abused reads as abused, and a surveyor pricing the remedy is pricing either a specialist clean, a resin coating over the top or, at the far end, sections cut out and recast. Those are three very different numbers and the difference between them is decided by what you did in years one and two, not by what you do in the final fortnight.

Nobody in the trade puts it this way, so I will. The slab remembers.

What a surveyor does in the first ninety seconds

They walk the aisles. Not the perimeter, the aisles, because that is where the traffic ran.

Then they look for four things. Dark patching under where machinery or a delivery bay sat. Black arcs and scuffs at the turning points. Pale rectangular shadows where racking uprights stood. And joints – the sawn or formed lines across the slab – checked for spalling at the edges, which is the crumbling you get when hard forklift wheels have hammered an unsealed joint for a decade.

Ninety seconds gets them to a category. Cleanable, coatable, or structural. Everything that follows in the schedule is a refinement of that first judgement, and the tenant who has never looked at their own floor with those four questions in mind arrives at the negotiation with no idea what category they are in.

What happens to a concrete slab over a ten-year tenancy?

Industrial slabs in London are usually power-floated concrete, laid smooth and left bare. Bare is the operative word. A power-floated finish looks sealed and is nothing of the sort: it is a dense but porous surface with an open capillary structure, and everything liquid that lands on it starts moving downwards.

Untreated concrete also dusts. Traffic abrades the surface paste, which is why a busy warehouse develops that fine grey film on every low shelf, and the dusting itself opens the pores further. A slab five years into heavy use is thirstier than the same slab on day one.

Then there is the loading. Point loads from racking legs, dynamic loads from a laden forklift crossing a joint at speed, chemical attack from whatever your operation spills. All of it accumulates, none of it reverses.

Oil is not a stain on the floor, it is inside it

This is the part tenants find hardest to accept, and I have had the conversation in a dozen units. Engine oil, hydraulic fluid and gearbox oil that has sat on unsealed concrete for years has wicked down into the slab, in some cases twenty or thirty millimetres.

Nothing applied to the surface reaches that. A degreaser lifts what is in the top couple of millimetres and the rest sits below, and within a fortnight it migrates back up and the stain reappears, paler and softer at the edges. A poultice does better, because it draws by capillary action as it dries rather than working downwards, and repeated poultice applications will pull a surprising amount out of a slab. What they will not do is deliver a floor with no history.

A tenant who ran a fabrication business or a fleet workshop needs to know this before they buy anybody’s remediation quote. The honest ceiling on cleaning a deeply oiled slab is a substantial improvement and a visible ghost. Any contractor promising better than that is either coating over the top or has not seen the floor.

Which marks come off, and which ones never will?

Sort the floor into categories before you price a thing.

Rubber transfer from forklift tyres comes off. It looks alarming – long black arcs that read as burns – and it is polymer deposited on the surface, not damage to it. The right rubber-mark remover and a rotary with the correct pad will take it. Using the wrong pad, which is the common error, burnishes it into the surface and makes it permanent.

Surface grime, the general grey of a decade, comes off with a scrubber-drier and the right detergent. Line marking comes off, though epoxy aisle lines need mechanical removal and that means grinding, which changes the texture of the concrete under the paint and leaves its own shadow.

Efflorescence, the white bloom you get where moisture has moved through the slab, is a symptom rather than a mark, and cleaning it without addressing the moisture achieves nothing. Spalled joints and cracks are not cleaning at all. They are repair, they belong to a different trade, and they belong in a different section of the schedule. The slab remembers, and it distinguishes between what you spilled on it and what you did to it.

The racking shadow nobody can remove

Racking is the one that catches everybody. Bolted-down uprights sit in place for the whole term, and the floor beneath them stays clean and unabraded while everything around it wears and greys. Take the racking out and you get a grid of pale rectangles across the unit, sharper than any stain.

There is no cleaning solution to this. The rest of the floor is worn and the protected patches are not, so the only route to an even appearance is bringing the whole slab back to a consistent state, which means grinding or coating the lot. That is a five-figure job in a mid-size unit, and it starts as an aesthetic complaint about some pale squares.

The fixing holes are separate again, and they are a making-good item under most yield-up clauses.

What does it cost to put right, and what would it have cost to prevent?

Rough figures on an 18,000 square foot unit, which is about 1,670 square metres.

A thorough industrial floor clean – degrease, scrub, rubber removal, several passes – runs to something in the region of £8,000 to £15,000 depending on how bad the contamination is and how much of it needs poulticing. A resin or epoxy coating over the whole slab, which is what gets specified when cleaning cannot deliver an acceptable appearance, sits nearer £60,000 to £100,000 once you include preparation. Cutting out and recasting damaged bays is priced by the square metre and by the misery.

The sealing job in week one

Sealing that slab when you moved in would have cost you a fraction of any of those numbers. A lithium silicate densifier with a decent penetrating sealer over a floor that size is a job of a few days and a few thousand pounds, applied to clean bare concrete before a single pallet lands on it.

The densifier reacts with free lime in the concrete and hardens the surface, cutting the dusting. The sealer closes the pores, so oil sits on top and gets mopped up rather than travelling down. A sealed slab still marks and still wears. What it does not do is absorb ten years of hydraulic fluid into its capillary structure, and that single difference is most of the gap between a five-figure handover and a six-figure one.

Almost nobody does it, because week one of a new tenancy is chaos and the slab looks fine.

Should you clean the floor at all before you hand it back?

Sometimes the answer is no, and I will say so on site even though it costs me the work.

If the slab is deeply oiled across a large area, spending £14,000 on remediation that leaves a visible ghost buys you very little ground in the negotiation. The surveyor still marks it up, and you have paid twice. In those cases the sensible move is a proper condition survey, a photographic record, a specialist opinion on what is achievable, and a negotiated cash settlement based on the diminution in value rather than a cleaning invoice. I have told clients in Barking to keep their money and instruct a building surveyor instead. The slab remembers, and no amount of my invoicing changes what it remembers.

Where cleaning does pay is the middle ground – surface grime, rubber, localised staining, the units where the floor is tired rather than saturated. That is most units, and there the money works hard.

February is the worst month to find out

Cold is the practical trap on a winter handover. Most degreasers and poultice systems need the substrate above about five degrees to work properly, and an unheated warehouse off River Road in Barking, IG11, in the last week of February is not above five degrees. Resin coatings are worse: below their minimum cure temperature they either skin badly or fail to cure at all.

So the March lease expiry has to be planned in January, with heaters hired and a slab warmed for a day before anything is applied. A tenant who rings a contractor in the final fortnight of a winter term will be told either that it cannot be done properly or that it can – and the second answer is the one to worry about.

“Your quote is nearly double the one underneath it, and I can’t see what the difference is.”

He was right that he couldn’t see it. Nothing in either document showed him where the money went – two pages of service descriptions, a monthly figure at the bottom, and no way of telling which one was arithmetic and which one was a guess. Cleaning is sold as a service and priced as a payroll, and the gap between those two things is where most bad contracts are signed.

So here is the payroll, opened up.

What does £2.40 a square foot buy in a London office?

Around £2.40 per square foot per year is where a five-night daily clean lands on a mid-size London office floor. Call the working range £1.80 to £3.20. The low end is an open-plan floor with one washroom core and a bin at every desk cluster. The high end is a cellular layout with meeting rooms, two kitchens, a client-facing reception and a managing agent who inspects.

On a 10,000 square foot floor, £2.40 is £24,000 a year, or £2,000 a month. That figure covers a cleaner in the building five evenings a week for something like three and a half hours, a supervisor across the site, the equipment, the insurance, the training and vetting, and whatever margin the contractor has managed to hold on to.

It does not cover a great deal else, which is the second half of this article.

The hour is the unit, not the square foot

Square footage is a convenient way to compare two quotes. It is not how anyone prices the work. What a contractor builds is a labour schedule, and the square footage only tells them how many hours to put in it.

For standard daily office cleaning – desks, bins, washrooms, kitchens, a vacuum through – a working cleaner covers somewhere around 2,500 to 3,000 square feet an hour. Break the floor into small rooms and that rate falls off a cliff, because doors, corners and separate bins eat time that open floor does not. Add a second washroom core and it falls again, since washrooms are the slowest square footage in any building by a distance.

The hours are the money. Everything else on the invoice is a rounding error by comparison, and any conversation about price that does not start with the hours on site is a conversation about nothing.

Where does the money go once you have paid the invoice?

Take the £24,000 apart.

Direct labour, including everything that sits on top of the hourly rate, runs to something between £16,000 and £17,000 of it. Call it seventy per cent. Supervision, mobile cover and account management take around £2,000 – the supervisor who walks the site, the person who finds you somebody when your regular cleaner’s child is ill. Materials, machinery and equipment replacement come in near £1,200. Insurance, training, DBS checks, COSHH documentation and the rest of the compliance overhead take another £1,000 or so.

What’s left is the margin, and on a competitive London daily contract it sits between five and ten per cent. That is thinner than most clients assume when they are pushing for another five off.

Look at the shape of that breakdown rather than the individual figures. Seventy per cent of what you pay walks into your building wearing a uniform. The hours are the money, and a contractor who tells you they have found savings in procurement or efficiencies in their operating model has found them in the rota, whatever the covering letter says.

The on-costs nobody quotes separately

A cleaner on the London Living Wage costs £14.80 an hour. That is not what they cost the contractor.

Holiday pay adds 12.07 per cent for staff on the accrual method. Employer National Insurance runs at 15 per cent on earnings above a secondary threshold of £5,000 a year, which was cut from £9,100 in April 2025 and is frozen until the end of the decade. That cut mattered enormously in this trade and went almost unremarked outside it. Cleaning runs on short shifts, and a part-time cleaner earning £8,000 a year used to sit under the old threshold entirely. Now most of that wage is inside the charge.

Add pension contributions where the earnings trigger auto-enrolment, sick pay, uniform, and the paid time a supervisor spends inducting somebody new. The £14.80 becomes something closer to £18 before a mop has moved. When a client tells me the maths of a quote doesn’t add up, this is usually the part they have left out.

What is not in the price?

Consumables are the first surprise. Paper towels, toilet rolls, hand soap, bin liners, sanitary units – these sit outside almost every daily cleaning specification in London and get billed on top, at whatever the contractor’s mark-up happens to be. A busy 100-desk floor can run to £3,000 or £4,000 a year in washroom consumables alone. Two quotes are not comparable until you know whether either of them includes this.

Periodics are the second, and the larger, surprise. The daily clean keeps a building presentable. It does nothing for the carpet, which needs extraction once or twice a year, or the vinyl, which needs stripping and resealing, or the kitchen extract canopy, which needs cleaning to a schedule your insurer may care about more than you do. Windows are almost always a separate line. So is the annual high-level dust.

Budget for the periodics separately or they will arrive as a series of unwelcome quotes across the year, each one presented as an exception, each one entirely predictable. A sensible planned maintenance schedule on a 10,000 square foot floor adds perhaps £3,000 to £5,000 a year on top of the daily contract, and it is the difference between a floor that ages gracefully and one that needs recarpeting two years early.

Above head height is always extra

The daily specification stops at the height a cleaner can comfortably reach standing on the floor. Everything above that line – partition tops, ductwork, light diffusers, the ledges above the door frames, the ceiling void grilles – is out of scope unless somebody wrote it in.

Nobody notices for about three years. Then the office gets a refit, or the lease ends and a surveyor arrives with a torch, and the dust that accumulated across those three years turns into a line on a schedule of dilapidations. The daily contract was never going to touch it. The annual high-level clean nobody bought would have.

How does the quote underneath yours come in a third cheaper?

The hours are the money, so a cheaper quote is a quote with fewer hours in it. There is no other lever of any size. Materials and margin together are barely a quarter of the price, and no contractor can find thirty per cent in there.

What they can do is put two and a half hours a night on the schedule where the building needs three and a half, and hope the client doesn’t count. They can price the labour at the statutory National Living Wage of £12.71 rather than £14.80, which is legal, and which produces exactly the staffing churn you would expect at that gap. They can leave the supervisor off the site and manage the account from an office in Croydon by phone. They can strip out cover, so a sickness absence means your floor gets missed and nobody rings to say so.

Each of those is a real decision by a real contractor, and each of them shows up on your floor within about four months.

Four ways the hours disappear

Watch for a specification with no hours in it at all – “cleaning to be carried out to the required standard” is a phrase that commits nobody to anything. Watch for a headcount without a shift length. Watch for washrooms folded into the general area rate, since washroom time is the first thing a short-priced schedule steals from. And watch for consumables priced generously against a suspiciously thin labour line, which is a contractor telling you where they intend to make their money back.

A quote that names the hours per night, the supervisor’s visit frequency and the periodic schedule can be argued with. A quote that names none of them cannot be compared to anything.

Is your building being cleaned more often than it needs?

Here is the part that costs me money to write.

A great many London offices are on five-night contracts for buildings that hold three days of people. Occupancy settled after the pandemic into a pattern everyone in facilities now recognises – full Tuesday to Thursday, sparse either side – and the cleaning specifications largely did not move with it. Friday night on a floor that held eleven people is not a cleaning requirement. It is a habit with an invoice attached.

I would rather sell four nights done properly than five done thin, and I say so at tender more often than my sales figures would like. Move the money instead of cutting it: four full nights, a proper Monday reset after the weekend, and the saved night’s budget put into the periodic schedule that everyone underfunds.

What we did off Jamaica Road

A client of ours took 14,000 square feet in a converted warehouse off Jamaica Road, SE16, and inherited a five-night specification from the previous tenant. Their people came in Tuesday, Wednesday and Thursday. We dropped Friday nights, put the hours into a longer Monday and a quarterly carpet extraction that the building had never had, and the annual cost came down by about eleven per cent. The floor looks better than it did on five nights, because the carpet is now cleaned and the desks were never the problem.

The one thing I would not touch was the washrooms. Those get done every day the building is open, whether four people used them or ninety, and that line is not negotiable at any price per square foot.

No square mile of London has changed more in my working life than the Isle of Dogs. I’ve cleaned there long enough to have watched the whole arc – the empty concrete, the towers going up, the trading floors filling and then half-emptying, and whatever it’s turning into now. People outside the trade see the skyline. What I see is a cleaning market that was invented from nothing in thirty years, nearly written off, and quietly being rebuilt into something different. The skyline tells one story. The cleaning tells another.

What was here before the towers?

Within living memory the Isle of Dogs was docks – the West India Docks, opened in 1802, that made London the busiest port on earth and then, when container shipping moved downriver, died. By 1980 the docks had closed and the peninsula was acres of derelict water and empty quay. There was no commercial cleaning market here to speak of, because there was almost nothing worth cleaning.

Then came the London Docklands Development Corporation in 1981, an enterprise zone with generous terms, and a developer’s bet that the City would overspill east. One Canada Square – the tower with the pyramid top that every Londoner can picture – opened in 1991, and Canary Wharf rose around it through the decade. The Jubilee line arrived in 1999 and the Elizabeth line in 2022, each one making the place easier to reach and harder to ignore.

The scale of the bet is easy to forget now. In the early eighties this was one of the poorest, most cut-off corners of the capital, and the notion that global banks would one day headquarter here read as fantasy. The towers built a cleaning market where none had existed.

How a derelict dock became a cleaning market overnight

For a firm like mine, the change created something that hadn’t existed: a concentrated cluster of Grade-A office towers, all needing cleaning to a standard the old City barely knew. Millions of square feet of marble and glass appeared where there had been rubble, and every bit of it needed a contract. A cleaning market didn’t grow here. It was switched on.

What kind of cleaning market did Canary Wharf actually create?

The market Canary Wharf built is unlike anywhere else I work. The towers are enormous single-occupier or few-occupier buildings, which means a handful of colossal contracts rather than a street of small ones. You don’t win the cleaning of a fifty-storey bank with a good quote and a van – it’s a facilities-management contract worth millions and run through a national FM company, with a specification the length of a phone book.

The work itself is specialist from top to bottom. The glass is cleaned from cradles and abseil ropes hundreds of feet up. The marble atriums need trained stone care rather than a mop. The trading floors are cleaned overnight to be spotless before the markets open, which means a night workforce most of the City never sees.

There’s a whole ecosystem behind it that the public never notices. Cleaners on the bank floors are security-vetted like the staff they clean around. The estate itself – the streets and the malls under the towers – is privately managed and privately cleaned, so even the pavement outside runs on a contract. It’s a self-contained cleaning economy, and it operates to its own rules.

My honest opinion, and the big FMs won’t like it: the sheer size of these contracts is the worst thing about the local market. A single-supplier deal across a whole tower locks out every capable independent in London and hands the work to whichever national can shave its margin thinnest – and you can read the result in the details a giant contract stops noticing. Scale buys consistency. It rarely buys care. The details are where you feel it – the corner that’s technically in spec and visibly tired, month after month.

Why an independent can’t touch the towers

It isn’t about quality, and it never was. A good London independent could clean any floor in Canary Wharf to a higher standard than the contract that holds it. What they can’t do is absorb a thousand TUPE’d staff and price a five-year deal thin enough to win the tender. The barrier is the shape of the contract, not the skill of the cleaner. That’s why the towers have always belonged to a small club of national firms, and always will.

Did hybrid working really gut the market?

For about three years the story everyone told was that Canary Wharf was finished. Hybrid working had emptied the towers, the argument went, and an office district built for a five-day commute made no sense in a two-day one. The tide had gone out, and plenty of people wrote the place off for dead.

From where I stood it was never that simple. Desk occupancy did fall, and it settled into a pattern the whole trade now recognises – rammed Tuesday to Thursday, quiet at the edges of the week. That changed the cleaning, but it didn’t end it. A floor used by three hundred people on Wednesday and ninety on Friday still has to be cleaned to the same standard on both, and a half-used building throws up its own problems – the kitchen that’s a flashpoint on the busy days, and the washrooms that can’t be scaled down just because Friday is quiet. If anything the uncertainty made the specs fussier for a while, not looser – nobody wanted to be the building that cut cleaning the year everyone had become alert to the state of a door handle.

The Tuesday-to-Thursday office

What actually happened was a change of shape. Daily deep cleaning gave way to smarter scheduling – heavier where the people were and lighter where they weren’t, with more daytime presence so tenants could see the place being looked after. The contracts that struggled were run by people who read “fewer desks” as “less cleaning” and cut the spec to match. The ones that held read it as “different cleaning” and re-planned around the new rhythm. The bins, I can tell you, never got as light as the headlines promised.

What is the market becoming now?

The obituary turned out to be premature. Canary Wharf had its best leasing year in a decade in 2025, footfall hit levels central London would envy – one January day during its winter light festival drew three hundred and twenty-five thousand people – and last autumn JP Morgan announced a new three-billion-pound headquarters on the estate for twelve thousand staff, the biggest office building London has seen in a generation. HSBC, once the symbol of the exodus, stayed put. The tide came back in.

It came back as a different sea, though, and that’s the part that matters for cleaning. Canary Wharf is deliberately no longer just banks. By 2028 barely half of it will be finance, down from over ninety per cent in the nineties. The rest is filling with life-sciences labs, shops, hotels, leisure and – most of all – homes. More than three thousand people already live on the estate, a number set to double.

That residential growth is where the work has quietly moved for a firm like mine. I picked up a communal-areas contract in a new tower on Marsh Wall, over on South Quay, a couple of years back – the lobbies and lifts, the gym, the roof terrace, the bin rooms that a thousand flats fill faster than any office ever did. It’s seven-day work, it’s relentless, and it exists because the Isle of Dogs stopped being a place people only came to work. A district that empties at night needs one kind of cleaning. A district that lives there needs another.

From trading floors to lab benches and lobbies

Every new use brings its own standard. A life-sciences floor needs a discipline closer to a hospital than an office, and a hotel runs to hospitality’s hours and expectations. The winter festivals and the weekend crowds leave public spaces that have to be reset by dawn. A festival that pulls three hundred thousand people through in a day leaves a mess no office ever generated, and it has to be gone before the Monday commuters arrive. Life-sciences tenants bring cleanroom-adjacent protocols and materials rules an ordinary office cleaner has never met, with sign-off and documentation to match. The cleaning market here is broader and busier than it was at the height of the banking years – it’s just spread across seven days and a dozen kinds of building instead of concentrated in Monday-to-Friday towers.

Where does that leave a cleaning contractor on the Isle of Dogs?

The honest answer is that the two markets barely touch. The towers still belong to the national FMs, walled off behind contracts no independent will ever bid for, and that hasn’t shifted since 1991. What has changed is everything around them. The residential blocks, the smaller offices on Millharbour and Marsh Wall, the life-sciences overspill, the seven-day leisure and events work – that’s the ground a London independent can actually win, and there’s far more of it than there used to be. Read the Isle of Dogs as a single office market and you’ll either chase towers you can’t win or write off a district that’s busier than it has ever been. Read it as a dozen markets stacked on one peninsula and the openings are obvious.

The work is in the edges, not the towers

The firms that have done well on the Isle of Dogs are the ones that stopped staring at the skyline and looked at the edges. The mistake is to treat the towers as the market; they’re a market, and a closed one. The real opening, all through the ups and downs, has been the buildings the giants don’t want and the hours they don’t work – the residential and the weekend work the big contracts can’t be bothered with. The tide here has always come in and gone out. The contractors who last are the ones who stopped trying to own the towers and learned to work the shoreline instead.

A tag appears on the front of a building overnight, and by nine the next morning I’ve usually had the call. The client wants it gone, wants it gone today, and assumes it’s a simple matter of scrubbing. Sometimes it is. Often it isn’t, and the difference comes down to two things nobody thinks about until the paint is already on: what the paint is, and what it’s stuck to. Get either wrong and you don’t remove graffiti so much as trade it for a worse mark that won’t come off at all.

What actually makes graffiti so hard to remove?

Graffiti isn’t a single substance. Aerosol paint, permanent marker, paint pens and acid etching each behave differently, and each needs a different answer. The paint is only half the problem, though. The other half is underneath it.

A non-porous surface – glass, or a sealed metal panel – is the easy case, because the paint sits on top and has nowhere to hide. Porous surfaces are where it gets hard. Unsealed brick and Portland stone drink solvent-based paint straight into the pores, so the colour you see on the surface is only the top of what’s actually there. Scrub the face clean and the stain wicks back out a week later like a bruise coming up.

Acid etching is its own miserable category, and it isn’t really a coating at all – the tagger has bitten damage into glass or polished stone, so there’s nothing to lift. On glass it usually means polishing out or replacing the pane. Whoever did it has essentially carved the building, and no solvent brings that back.

Why the wall matters more than the paint

Before I touch anything, I work out what I’m standing in front of. Porous or sealed, natural stone or painted render, sound or already crumbling. That reading decides everything that follows, because the same remover that lifts paint off glazed brick in minutes will drive it deeper into soft limestone. The tagger picked the wall at random. The person removing it doesn’t get to.

Why is speed the thing that matters most?

If there’s one lesson I’d press on a building manager, it’s that graffiti is a race, and the clock starts the moment the paint dries. Fresh paint hasn’t fully cured or sunk in, so it lifts with far less aggression than a tag that’s been baking on a south-facing wall for three weeks. Wait a month and a job that needed a mild solvent now needs something that risks the surface beneath it.

There’s a second reason to move fast, and it’s about the next tag rather than this one. A wall cleaned within a day or two rarely gets hit again. Taggers want their work seen, and a surface that swallows their effort overnight is no fun to paint. Leave a tag up for a fortnight and you’ve advertised the spot as unmanaged – I’ve watched a neglected loading bay out toward Shepherd’s Bush become a gallery in that time, one tag turning into six.

The 48-hour window

The outfits that manage graffiti well across big estates all run to the same rough rule: get it off within about forty-eight hours, every time, no debate. It sounds excessive until you’ve seen a quiet service yard turn into a canvas. The forty-eight-hour habit costs a little every week and saves a fortune across a year, because you never let the surface reach the state where only the harshest, most expensive methods will touch it. Speed is cheaper than it looks, too, because the fast jobs are the easy jobs – you’re beating the clock and the chemistry at once, before either has set hard.

What methods do professionals actually use?

The toolkit runs from gentlest to harshest, and the craft is starting as gentle as the job allows and stopping the moment it works.

Chemical removers come first, matched to the paint rather than grabbed off one shelf – a clinging gel for vertical brick, or a poultice pasted onto porous stone and left to draw the stain out as it dries before it’s scraped away with the colour trapped inside it. For masonry, superheated water systems – the trade knows them by names like DOFF – deliver steam at low pressure that softens paint and floats it off without hammering the surface. Gentle abrasive methods, swirling a fine media in a soft mix of air and water, handle the jobs where nothing milder will. Most removers also need managing while they work – kept wet and given time to dwell, then rinsed and neutralised so nothing keeps eating at the surface after you’ve packed up. The times on the tin are a starting point, not gospel; brick in July and stone in January behave nothing alike.

And then, at the far end, the pressure washer.

I’ll be blunt here, because I clean up the results too often. A high-pressure lance taken to a stone or brick façade is vandalism with a machine. It blasts the graffiti off and takes the face of the stone with it, leaving a permanent bright patch and opened-up pores that drink the next tag deeper than before. It blows out the mortar joints too, and starts the clock on water damage. Half the ruined façades I get called to were wrecked not by the tagger but by the last firm’s jet-wash. High pressure has its place on tough, non-porous hardstanding. On a building’s face it’s almost always the wrong answer.

Why the pressure washer is the last tool, not the first

Professionals reach for it last for one reason: it’s the only method that can’t tell the difference between the paint and the wall. A solvent works on the chemistry of the paint; steam softens and lifts it. A high-pressure jet does neither – it just removes material, with no idea whether that material is aerosol or four-hundred-year-old limestone. The skill in this trade is picking the mildest thing that works, not the strongest thing you own.

How do you get graffiti off heritage stone without wrecking it?

Heritage stone is the job that separates people who know this trade from people with a van and a machine. Portland stone, the pale limestone half of grand London is built from, is soft and porous, and often listed – which means the wrong method can be unlawful as well as damaging.

I was called to a Portland-stone office frontage on a Bloomsbury side street where someone had sprayed a metre-high tag across the ground floor, and the managing agent’s first instinct – understandably – was to get the nearest firm to blast it off by lunchtime. That blast would have etched a bright, rough scar into the stone that outlasted everyone involved. What the façade needed was patience: a solvent poultice, tested first on a hidden corner, then worked in small sections to pull the paint out of the pores over hours rather than seconds, followed by gentle superheated steam. It took the best part of two days. Two days on a busy frontage also means hoarding the work off from pedestrians and watching the weather, so a poultice doesn’t freeze or bake before it’s drawn the paint out. When we finished, you couldn’t tell anyone had ever painted it.

That last part is the real test, because the trap with porous stone is the ghost – the faint outline that lingers after the colour’s gone, where paint sank unevenly or a heavy hand bleached one patch paler than the rest. The whole skill is avoiding the ghost. It’s why the test patch comes before the confidence, and why anyone promising a heritage façade cleaned in an hour is promising you a ghost.

The ghost, and the test patch that saves the facade

No two stones take a stain the same way, so the small square you clean first – somewhere nobody looks – tells you what the whole wall will do before you’ve committed to it. Skip the test patch and you learn the remover lightens the stone at the same moment the client does. On a listed building it’s worth a call to the conservation officer as well, since gentle methods usually pass without fuss while aggressive ones can land you in real trouble. The stone was there long before the tag and it’ll be there long after – the job is making sure the removal doesn’t outlast both.

Do anti-graffiti coatings actually pay off?

Every client on a repeatedly hit wall asks about coatings, and the honest answer is that it depends. They come in two broad types. Sacrificial coatings are a clear, wax-like layer that lifts off with the graffiti when you clean it, carrying the paint away; you then reapply. Permanent coatings seal the surface so future paint can’t bond and wipes off with far less effort, and they last for years.

On a façade that gets tagged every few weeks – a quiet side elevation near a night-time route – a coating earns its money fast, both by cutting each removal to a wipe and by shielding the substrate from the removers themselves. On a frontage hit once in five years, you’re paying to solve a problem you don’t really have. And on heritage stone there’s a further catch: many coatings trap moisture in stone that needs to breathe, so the cure turns out worse than the tag. A permanent coating can also dull or slightly cloud a polished surface, which a client who cares about the look of a glass-and-stone entrance notices at once. There’s no free protection; every coating is a trade.

Sacrificial versus permanent, and when neither is worth it

One thing no coating changes is where the mess goes. The stripped paint and spent remover coming off a wall is contaminated waste, and it can’t be hosed into the nearest surface drain to the river – it belongs in a controlled catch and a proper foul-sewer route, the same as any other trade effluent. A firm that lets the remover sheet off down the gutter into the drain is telling you exactly how it treats everything else.