The honest answer to “how long does a deep clean take” is that nobody can tell you from your floor area, and the industry figures explain why. German cleaning-industry reference values put maintenance cleaning of a hard office floor at roughly 250–350 square metres an hour and a Grundreinigung — a deep clean — of the same hard floor at 40–80 square metres an hour. That is not a longer version of the same job. It is a job that runs at somewhere between a fifth and a seventh of the speed, on the same surface, with the same person standing on it.
So the useful planning rule is not a number of hours per bedroom. It is a multiplier: whatever a maintenance clean of a given room costs you, budget roughly four to six times that for a genuine deep clean of it — and then understand which factors push you toward the slow end of that range, because for most households the answer is “all of them.” What follows is how the estimate is actually built in the trade, what the published figures say, and how to derive your own number instead of taking a guess off a service page.
A deep clean is a different job, not a longer one
English-language cleaning content treats “deep clean” as an intensity setting. The German trade treats it as a separate product with a separate purpose, and that framing is far more useful for estimating.
The trade lexicon at gebaeudereiniger-online.de defines Grundreinigung as work carried out periodically that serves the appearance and the Werterhalt — the value preservation — of the floor, and describes it as the thorough removal of residues of every kind, giving as its examples adhering soil and care films: the built-up layer of polish, dressing and product left behind by previous cleaning. German contractor explainers put the interval at roughly once or twice a year and describe it as the most intensive cleaning form there is.
Read that carefully, because it contains the thing most deep-cleaning advice never says. A significant fraction of what a professional deep clean removes is not dirt. It is accumulated product. Every maintenance clean leaves a small residue, that residue holds soil better than the bare surface did, and after a year or two the surface is dull because of what has been put on it rather than what has landed on it. That is why the deep clean is periodic rather than reactive, and it is why the job is slow: dissolving and lifting a bonded film is a chemically and mechanically different operation from wiping loose soil off a clean surface.
It also explains a frustration people report constantly and misdiagnose as their own incompetence — the floor or worktop that looks worse the more diligently it is cleaned. The maintenance clean is not failing. It is depositing.
In the trade, a Grundreinigung of a hard floor often means stripping the finish and re-coating it. That is a deliberate, specified operation on a commercial floor built for it. It is not something to attempt on a sealed or oiled domestic wood floor, a lacquered parquet, or a luxury vinyl plank with a factory wear layer, where an aggressive stripper removes the protection and cannot put it back. Match the intensity to the finish, not to the trade term — and check the flooring manufacturer’s own instructions before using any product described as a stripper, a basic cleaner or a Grundreiniger.
What the published productivity figures actually say
Commercial cleaning is priced from published Leistungswerte — performance values in square metres per hour, by task and by surface. One German cleaning-calculation reference publishes the following set, citing BIV and FIGR industry reference values as their origin:
| Task | Published range (m²/hour) |
|---|---|
| Hard floor, corridor — maintenance | 400–600 |
| Carpet, office — maintenance | 350–500 |
| Hard floor, office — maintenance | 250–350 |
| Stairwells — maintenance | 100–200 |
| Sanitary areas — maintenance | 60–120 |
| Glass, interior | 50–100 |
| Hard floor — Grundreinigung (deep clean) | 40–80 |
Three things fall straight out of that table.
The deep-clean row is the slowest thing on it. Slower than sanitary work, slower than interior glass. Deep cleaning a hard floor is, by these figures, in the same productivity bracket as cleaning a bathroom — and a bathroom is universally understood to be the slow room.
The ratio is roughly five to one. Same floor, same building, same operative: 250–350 m²/h to maintain, 40–80 m²/h to deep clean. Take the midpoints and it is about a factor of five. That multiplier is the single most useful number in this article, because it converts something you already know — roughly how long your normal clean takes — into something you do not.
Room type moves the figure by an order of magnitude before you change anything else. A corridor runs at up to 600 m²/h; a sanitary area starts at 60. A separate German facility-management reference gives the same shape by room: circulation 300–1,200 m²/h, offices 200–700, meeting rooms 160–240, sanitary rooms 60–180, entrance areas 60–160. Whichever set you use, bathrooms and entrances are the expensive square metres and open circulation space is nearly free. Your house is not shaped like an office, but it is shaped like that distribution: the landing costs almost nothing and the bathroom costs most of the morning.
The same reference is blunt about the status of these numbers: industry reference values from BIV and FIGR are orientation values, not guarantees. ISSA, whose cleaning times perform the same function in the English-speaking commercial market, says the equivalent thing about its own figures — they are an estimating guide rather than a time-and-motion study. Use them the way the trade uses them: to get the order of magnitude and the relative cost of tasks right, then adjust for the specific building.
Why the ranges are so wide, and why you sit at the slow end
A range of 40–80 m²/h is a factor of two. A range of 200–700 is a factor of three and a half. That width is not sloppiness; it is the honest expression of how much the same task varies, and understanding what drives it tells you where your own estimate should land.
REFA Consulting — the German work-study organisation whose time-study methods underpin a great deal of industrial and service-sector labour planning — sets out how these figures are produced and why they cannot simply be copied between buildings. The methods named are REFA-Zeitaufnahme (direct timed observation) and Multimomentaufnahme (multi-moment random sampling). The conclusion drawn is that cleaning performance is determined not only by the structural conditions of the building but above all by the technical equipment and by the cleaning personnel, and that figures therefore do not transfer one-to-one between objects. The variables it lists include building layout and travel distance between areas, machine and equipment specification, staff qualification, cleaning frequency and method, waste-disposal logistics, and infrastructure such as water access and cleaning cupboards.
ISSA reaches the same place from the other direction. Its cleaning times, in its own account, are built from five components: task, tool, time, total units and training. Two of the five are not about the work at all. They are about what you are holding and whether you know how to use it.
That is the diagnosis for the domestic case, and it is not flattering, but it is not a criticism either. On every one of those variables, a household sits at the unfavourable end:
- Obstruction. Published rates assume a commercial floor plate. A domestic room is dense with furniture, and every item is a lift, a move or a reach-around. The FM reference is explicit that rooms with a lot of furniture, equipment or hard-to-access surfaces require more time — that is precisely why the office range runs from 200 to 700.
- Travel. Commercial rates are measured over large contiguous areas so that walking is amortised across hundreds of square metres. A house is a set of small disconnected areas connected by stairs. You will spend a share of your day carrying things up and down that no published figure accounts for.
- Equipment. The rates assume purpose-built kit — a 14-inch flat mop with a dual-chamber bucket, a machine matched to the floor. Substituting a domestic mop and a single bucket does not cost a few per cent; it changes the task. This is the practical reason to look at what the trade actually buys before a big session rather than after, in the vacuum comparisons and the gear guides.
- Training. The published figures are for someone who has done the task thousands of times and does not stop to decide anything. In Germany, Gebäudereiniger is not a job title but an apprenticed trade: the federal training ordinance, the Gebäudereinigerausbildungsverordnung of 28 June 2019, sets the apprenticeship at three years, structures it around nine occupational competency areas — among them planning and organising work assignments, operating and maintaining cleaning equipment, carrying out cleaning measures, surface care and refurbishment, and quality assurance — and splits the journeyman’s examination into two parts, with Part 1 sat in the fourth training semester. The British equivalent is structured differently but makes the same point about competence being certified rather than assumed: BICSc, the British Institute of Cleaning Science, makes its Licence to Practice a compulsory unit that must be completed before a candidate can progress to any of its Base, Additional, Specialist or Sector Specific units, of which there are over forty.
Nobody is suggesting you need a three-year apprenticeship to clean your kitchen. The point is narrower and more useful: when the published rate is 40–80 m²/h and it was measured on someone with certified competence, purpose-built kit and an unobstructed floor, take the slow end of every range you use, and then add. An estimate built on the fast end of commercial figures is the single most common reason a deep clean overruns by an entire day.
The variable nobody budgets for is the interval
Here is the part that decides whether your Saturday is enough, and it is not in any of the tables: how long it has been since the last one.
The mechanism is straightforward and it is worth stating in physical terms. Fresh soil sits on a surface, held by little more than gravity and static, and comes off with a cloth. Left in place, three things happen to it. It gets pressed into the surface texture by traffic and by wiping. It takes on binder — cooking aerosol in a kitchen, soap and body oils in a bathroom, the residue of cleaning product everywhere — which converts loose particulate into an adhering film. And in the wet rooms it gains a mineral component from evaporating hard water, which is a scale problem rather than a dirt problem and needs an acid rather than a detergent.
Every one of those transitions moves the job from wiping to dissolving, and dissolving costs time you cannot compress by working harder. It is exactly the transition the productivity table describes as the difference between the 250–350 row and the 40–80 row.
This is why the same house takes wildly different amounts of time to deep clean depending on nothing more than when it was last done, and why “how long does it take to deep clean a three-bedroom house” has no answer. A house on an annual cycle and a house that has never had one are not the same job. The second is a restoration, and restorations do not obey productivity figures at all — the first one after a long gap is the outlier, and every subsequent one is dramatically faster.
The planning consequence is uncomfortable but freeing: if you have never done this, the first pass will take substantially longer than any figure in this article predicts, and that is not a sign you are slow. It is a sign you are paying down an interval. Budget the first one generously, split it across sessions, and expect the second one to be a different experience.
Preparation is a separate job, and the standards say whose job it is
Almost every domestic deep-clean overrun has the same shape: the cleaning was fine, but two hours vanished into moving things.
The commercial sector solved this by writing it into the contract, and the clearest example is a standard written for the highest-stakes cleaning there is. DIN 13063:2021-09, Krankenhausreinigung — Anforderungen an die Reinigung und desinfizierende Reinigung in Krankenhäusern und anderen medizinischen Einrichtungen (“Hospital cleaning — Requirements for cleaning and disinfecting cleaning in hospitals and other medical facilities”), published in September 2021, is described by DIN as laying down concrete requirements for the service provider and the service recipient, covering the interface and responsibilities between them alongside quality management and monitoring.
That is the structural insight worth stealing. A serious cleaning specification does not treat the building as a passive object. It names what the client has to do — clear the surfaces, provide access, make the water and power available, decide what is allowed to be moved — and it makes that a precondition rather than an afterthought, because if it is not written down, it happens on the clock.
In a house, you are both parties. So do the client’s half first, as a separate task, ideally on a different day:
- Clear every horizontal surface in the rooms you intend to do. Not tidy — clear. Into boxes, out of the room.
- Decide in advance what does not move. The wardrobe, the fridge, the sofa. Write it down. Mid-job is the worst possible moment to start negotiating with yourself about whether to pull the cooker out.
- Assemble everything before you start. Every trip to a cupboard is dead time, and it is the one inefficiency that scales with how much you are trying to get done. The reasoning behind carrying the whole job with you is in what professional cleaners do differently.
- Buy or wash enough cloths for the whole session. Running out mid-job means either stopping or continuing with a loaded cloth, and a loaded cloth redeposits rather than collects — see the two-cloth method.
Preparation done properly takes a real hour or two. Counted separately, it is a manageable task. Counted inside the deep clean, it is the reason the deep clean took eleven hours.
Building your own estimate
Here is the derivation, stated openly so you can adjust any step of it. Two inputs, both from published figures: the maintenance rate for the room type, and the roughly five-to-one deep-clean multiplier implied by the 250–350 versus 40–80 comparison.
- Measure or estimate the floor area of each room you intend to deep clean. Only those rooms. A deep clean of a whole house in one session is a project, not a Saturday.
- Take the slow end of the published maintenance rate for that room type. Sanitary 60 m²/h; hard floor with furniture 250 m²/h; stairs 100 m²/h. Slow end, always, for the reasons in the section above.
- Divide area by rate to get the maintenance time. Area ÷ Leistungswert = time. That is the trade’s own formula and it is the whole of the arithmetic.
- Multiply by five for a deep clean. Four if the room is uncluttered and was deep cleaned within the last year; six or more if it is dense with furniture or has never been done.
- Add preparation separately — clearing, assembling, moving — and do not fold it into the cleaning number.
- Add nothing for the slow chemical jobs, but sequence them first. Oven, limescale, grout: these need product sitting on them for a period the label specifies, and that period is spent whether you stand there or not. Apply them at the start of the session and they cost you no additional time at all. Apply them at the end and they cost you their full duration.
Worked through for a small bathroom of roughly 6 m²: at the slow end of the sanitary maintenance range, 6 ÷ 60 = about six minutes for a maintenance clean, which sounds about right for a quick wipe-round. Multiply by five and you get around thirty minutes of actual deep-cleaning work — before preparation, before descaling dwell time, and before anything you have to move. That is why a bathroom deep clean that “should take twenty minutes” reliably takes ninety: the cleaning was half an hour, and the other hour was everything the estimate left out.
Run the same arithmetic across the rooms you actually intend to do and you will get a total that is larger than you hoped and considerably smaller than the open-ended dread that made you put it off. That is the point of estimating. The specific procedures for each surface — what to use on grout, on an oven, on a mattress — sit in the deep-cleaning section; this article is only about the clock.
Hours are the wrong thing to be measuring
Worth ending on the observation that the most sophisticated system in the field refuses to specify time at all.
NS-INSTA 800-1:2018, published by Standard Norge as Rengjøringskvalitet — Del 1: System for å fastsette og bedømme rengjøringskvalitet, is the Nordic system for establishing and assessing cleaning quality, used across Denmark, Finland, Iceland, Norway and Sweden. Standard Norge’s own description states that it is independent of the cleaning method, of the system used, and of the cleaning frequency. NHO Service og Handel, the Norwegian service-sector employers’ association, states the same point plainly: the standard says nothing about how the cleaning is to be carried out. What it specifies is the acceptable amount of residual soil, across six quality levels from 0 to 5, where 5 is the highest and 1 the lowest, and level 0 has no practical significance.
So a Nordic cleaning contract can specify the result and stay silent on the hours, the method and the schedule. The contractor is free to be fast. What is not negotiable is what is left behind.
For a household that is the right way round, and it resolves the anxiety underneath the original question. You are not trying to spend the correct number of hours. You are trying to reach a defined end state in a room and then stop. Decide before you start what “done” means for each room — a list you can look at, not a feeling — and the duration becomes a consequence rather than a target. It also stops the two failure modes that waste the most time: polishing something already finished, and drifting into an adjacent room because the current one has no defined end.
Pick a level and hold it. Not every room has to be at the same one; the Nordic system explicitly allows different surfaces in the same room to carry different quality levels, and it permits more residue in difficult-to-access locations than in easily accessible ones. The top of a wardrobe is allowed to be dirtier than the desk. That is not a compromise you are making. It is how the professional assessment system is built.
Frequently asked questions
How long does it take to deep clean a whole house?
There is no defensible single figure, and any page that gives you one from bedroom count alone is guessing. Build it from the rooms instead: take the slow end of the published maintenance rate for each room type, divide the room’s area by it, multiply by about five for deep-clean intensity, and add preparation as a separate line. The variable that moves the total most is not the size of the house but the time since the last deep clean, because soil that has been left to bond has to be dissolved rather than wiped.
Why is a deep clean so much slower than a normal clean?
Because it is a different physical operation. A maintenance clean removes loose soil sitting on a surface. A deep clean removes soil that has been pressed into the surface, bound by grease or body oils, mineralised by hard water, or sealed under a film of previous cleaning product. German trade figures put maintenance cleaning of a hard floor at 250–350 m²/h and a Grundreinigung of the same floor at 40–80 m²/h — roughly a fifth of the speed.
How often should a deep clean actually happen?
German contractor explainers describe a Grundreinigung as a once- or twice-yearly operation, separate from and additional to the scheduled maintenance cleaning underneath it. That is a commercial building with daily maintenance, so treat it as a floor rather than a ceiling for a home on a weekly routine. The useful test is not the calendar but the mechanism: when surfaces stop responding to a normal clean, the soil has bonded and the interval has been exceeded. Frequency planning by room and surface is covered in the routines hub.
Should I deep clean the whole house in one go or spread it out?
Spread it, by room. The published productivity figures assume continuous work on a single task type in a single area, and every switch between rooms and tasks costs setup and travel that no figure includes. Splitting by room preserves that continuity, gives each session a defined end state, and avoids the specific failure of a house that is uniformly half-done. A house with three finished rooms and three untouched ones is in a better position than one with six rooms at sixty per cent.
Does buying better equipment actually make it faster?
Materially, yes, and the trade’s own accounting says so. ISSA builds its cleaning times from five components and two of them — tool and training — are about what you are holding and whether you know how to use it rather than about the work itself. REFA’s analysis reaches the same conclusion: cleaning performance is determined above all by the technical equipment and the personnel, which is why published rates cannot be transferred between buildings one-to-one. The published mopping figure assumes a 14-inch flat mop and a dual-chamber bucket, not a string mop and one bucket of increasingly dirty water.
How do I know when a room is actually finished?
Define the end state before you start, in writing, per room. The Nordic quality standard NS-INSTA 800 does exactly this: it specifies the acceptable amount of residual soil across six levels and says nothing at all about method, frequency or hours, which means “done” is a property of the room rather than of your effort. It also grades easily accessible and difficult-to-access locations separately and permits more residue in the latter — so a written definition of done is allowed to say that the top of the wardrobe is out of scope.
Is a professional deep clean faster than doing it myself?
Faster in elapsed hours, generally — that is what the training and equipment components in the productivity figures are measuring. Whether that is worth it is a separate question this site does not answer, because it depends on a local market and this is not a pricing publication. What the figures do tell you is where the difference comes from, and it is not effort: it is purpose-built kit, certified competence — a three-year apprenticeship in Germany, a compulsory Licence to Practice before any other unit in the British BICSc scheme — and the absence of any decisions being made mid-job.