We tested the published cleaning rate benchmarks against their own arithmetic. They don't reconcile.

Every page that tells you what commercial cleaning costs per square foot also tells you what it costs per hour. On the pages we checked, those two numbers cannot both be true — and the gap is as much as 13×.

If you price commercial cleaning, you have looked up a rate per square foot. Everyone does. You finish a bid, you want to know whether your number is sane, and you search for what this kind of building usually costs.

We did the same thing while building a bid calculator — and then, because the calculator had to show its work, we checked the benchmarks against the other numbers printed on the same pages. They don't survive the check. Not approximately: a published rate per square foot and a published rate per hour, from the same article, routinely describe prices that differ by three to thirteen times.

This isn't an accusation of dishonesty. It's a unit problem that has been copied from page to page until nobody states the unit at all. But it matters, because operators are using these numbers to sanity-check real bids on twelve-month contracts.

The test

It's simple, and you can run it on any page that publishes a rate.

1. Take the per-square-foot rate the page publishes.
2. Work out how many labour hours the building actually needs. Use the production rate the industry publishes for that space — commonly 3,000–4,000 sq ft per labour hour for general office — then add restroom fixture time and travel.
3. Divide the money by the hours.
4. Compare the result against the hourly rate the same page publishes.

If a page says a building costs $X per square foot and also says cleaning labour bills at $30–75 an hour, those two statements make a testable claim about how long the building takes. Usually the implied time is impossible.

In Case 1 we use 3,400 sq ft per labour hour — the midpoint of the published office range — plus about three minutes per restroom fixture and 0.4 hours of travel and lock-up. Where the assumption matters we show what it would have to be instead for the published figure to work.

Case 2 deliberately drops the fixture and travel allowance and uses bare production rates. The question there is not whether a figure is possible but how far the answer moves when one assumption changes, and adding the same constant to both sides would only narrow the gap while making the comparison harder to follow.

Case 1 — Housecall Pro

Housecall Pro publishes a rate of "$0.07 to $0.20 per square foot" for standard recurring office cleaning, states that hourly rates "often fall between $30 and $75 per hour", and gives a worked example: a 5,000 sq ft office cleaned three times a week at $607 per visit, which it notes is $0.12 per sq ft — comfortably inside its own band.

So take their example at face value and ask how long 5,000 sq ft takes.

$282.91 per labour hour, against their own published ceiling of $75. Their example bills at 3.8 times the top of the range printed a few paragraphs above it.

You can run it the other way, and it's worse. For $607 to sit inside $30–75 an hour, that 5,000 sq ft office would have to take between 8.1 and 20.2 labour hours — a production rate of 247 to 618 sq ft per hour. The published range for office work is 3,000–4,000.

No assumption of ours rescues this. Even at an implausibly slow 1,500 sq ft/hr the example still implies over $150 per labour hour.

Case 2 — ISSA, and why a rate per square foot isn't a price

ISSA's guidance article publishes "$0.05–$0.25 per square foot", gives a "Common range: $25–$60/hour", and works an example: "20,000 sq. ft. at $0.10/sq. ft. = $2,000/month".

That example settles the unit — in ISSA's usage the per-square-foot figure is monthly, which is more than most published tables bother to say. What it does not settle is how long the building takes, and that turns out to decide everything.

Unlike most sources, ISSA states a production rate of its own. As an example of estimating job length it gives "1 minute per 300 sq. ft. for standard office space" — which converts to 18,000 sq ft per labour hour. So test the example twice: once on ISSA's figure, once on the 3,400 sq ft per labour hour commonly published for basic office work. Same building, same $2,000, nothing else changed.

An 8.82× spread on one contract, and only one of those four rows lands inside the $25–$60 range printed in the same article. The spread is not a coincidence or an artefact: it is exactly the ratio between the two production rates multiplied by the ratio between the two frequencies.

This is the real finding, and it is not that ISSA's number is too high or too low. It is that a rate per square foot is a ratio — a price divided by an area — and it says nothing at all about labour until you supply the two things it doesn't carry: how fast the crew cleans, and how often they show up.

ISSA's own production-rate figure is worth a second look on its own account. The same list prices restrooms at "10 minutes per 200 sq. ft.", which works out to 1,200 sq ft per labour hour — a 15:1 ratio between the two spaces. And 18,000 sq ft per hour sits four and a half to six times above the 3,000–4,000 commonly published for basic office. A crew clearing 18,000 sq ft of office in an hour is covering about four tenths of an acre every sixty minutes, trash and vacuuming included. We have put that to ISSA directly rather than around them, and they are looking into it.

The same article elsewhere builds a per-visit price from the bottom up — three hours of labour at a $25 loaded rate, plus 15% overhead and 20% profit, reaching $103.50 per visit. That calculation is sound, and at $34.50 per labour hour it sits comfortably inside their stated range. It is the per-square-foot figure above it that cannot be pinned down.

The collision

Here is why the confusion persists. The numbers that circulate are roughly the same digits whichever unit is meant — and the unit is usually missing.

One 10,000 sq ft office, cleaned three times a week. Roughly 3.84 labour hours a visit, 49.8 a month.

The same band of digits, read two ways, differs by thirteen times — exactly the number of visits in a month at three a week. One reading pays a crew $521 an hour. The other pays them $10. Neither is a benchmark; together they are a coin flip.

And a third pattern shows up in consumer marketplace listings, where real quoted jobs — a 6,300 sq ft office at $75 a visit, a 17,000 sq ft office at $175 — work out to roughly $0.01 per sq ft per visit. That is an order of magnitude below the per-visit bands, and lands squarely inside the $25–30 per labour hour you would expect. The real transactions and the published benchmarks are not describing the same world.

What a defensible number actually looks like

Work forward from labour instead. Same 10,000 sq ft office, three nights a week, priced at the hourly range these same sources publish:

Expressed per square foot, that whole defensible range is $0.012–$0.029 per sq ft per visit, or $0.15–$0.37 per sq ft per month.

Look at where that falls. It is far below the published per-visit bands and above most of the published monthly ones. If you had checked your bid against either, you would have been told you were wrong — in opposite directions.

What to do instead

Stop checking bids per square foot. The figure is a useful shorthand between two people who already agree what it means, and close to meaningless as a sanity check on your own number, because you cannot tell which unit the thing you are comparing against was quoted in.

Check per labour hour. Divide your price by the labour hours the job actually needs. That comparison is frequency-independent by construction, it is the one benchmark the sources agree on ($30–75 per labour hour), and it is the only one that knows anything about the work being paid for.

Then do the thing no benchmark can do for you: work out what an hour of your crew costs, loaded — wage plus payroll tax, workers' compensation, PTO and turnover, with overhead spread across the hours you can actually invoice. A market range tells you what buildings like this tend to pay. Only your own cost tells you whether you make money at it.

Method and sources

Figures were read directly from each source page on 5–6 October 2026 and are quoted as published; the ISSA production-rate figure was re-read on 12 October 2026. In Case 1 and the collision table, labour hours use a production rate of 3,400 sq ft per labour hour — the midpoint of the commonly published 3,000–4,000 range for general office — plus 0.055 hours per restroom fixture and 0.4 hours of travel and lock-up. Case 2 uses bare production rates with no fixture or travel allowance, for the reason given above; carrying the allowance through narrows its spread from 8.82× to 4.85× without changing the conclusion. Months are 4.33 weeks (52 ÷ 12).

• Housecall Pro — How to price commercial cleaning jobs
• ISSA — How to calculate commercial cleaning rates
• HomeGuide — Commercial cleaning prices
• Clean Estimator — Commercial cleaning cost guide
• Thumbtack — Office cleaning prices

If we have misread any of these, we want to know and we will correct it in public — support@provebid.com. The arithmetic above uses only figures published on the page being tested, plus the one production-rate assumption stated here, so it should be straightforward to check.