---
title: "The contract that quietly went underwater · BlastoClean"
description: "A janitorial contract signed two years ago is still invoicing the same figure every month. That is exactly why nobody notices when it stops paying."
url: "https://www.blastoclean.com/blog/the-contract-that-quietly-went-underwater"
---

# The contract that quietly went underwater

August 14, 2026 · 4 min read · The BlastoClean Team

Ask a janitorial owner which of their contracts is the worst one and you will
usually get a fast, confident answer. Ask them to prove it and the room goes
quiet.

That is not carelessness. It is the shape of the business. A commercial
contract is signed once and then runs for years, and nothing about it looks
like it is changing: the same invoice goes out on the first of the month, and
it gets paid. Every visible signal says the account is fine.

What changes is the **time**.

## How a good contract goes bad without a single event

A building bid at three hours a night is bid honestly. Then:

- The tenant on the second floor fits out an open-plan area and adds forty
  desks.
- The client asks, reasonably, if the crew can start doing the stairwells.
- One of your two people leaves and the replacement is slower for six months.
- Somebody starts pulling the recycling separately because the client asked
  nicely one Tuesday.

None of those is a renegotiation. Not one of them generates an email anybody
would keep. But four hours a night at $30 an hour loaded is $600 a month more
than three hours was — and if the invoice never moved, that $600 came out of
your margin.

The contract did not fail. It **drifted**, and drift has no event to notice.

## Why your P&L will not tell you

A profit and loss statement answers "did the company make money". It is the
wrong grain for this question in two directions at once.

**It rolls sites together.** A property manager with ten buildings is one
customer, one contract and one invoice. Eight of those buildings can be
comfortably profitable while two bleed, and the account as a whole still looks
healthy. The two that bleed are the two you would have re-bid at renewal, if
you had known.

**It rolls months together.** The month the second floor got fitted out is the
month everything changed, and a yearly figure has no way to show you which
month that was.

You need the same money, cut by **site and month**. That is the only grain at
which "this used to work and now it doesn't" is visible.

## The three numbers

**Revenue.** The agreed price of the visits actually done. Billed, not
collected — this is a question about whether the *work* pays. A client who is
slow to pay is a real problem with a different report.

**Labour cost.** Hours actually worked, times what an hour actually costs you.
Both halves of that sentence matter.

*Actually worked*, because the whole point is the gap between what you bid and
what it takes. If your crews clock in and out, you already have this. If they
don't, this is the reason to start.

*Actually costs you* means **loaded**: wage, plus payroll taxes, plus workers'
comp, plus insurance, plus the share of the van, the vacuum and the chemicals.
An owner who costs jobs at the hourly wage concludes that every contract in
the book is profitable, because at the wage alone almost every contract is.
That is not optimism, it is the wrong arithmetic — and it is the single most
common way a cleaning company grows itself into trouble.

**Margin.** The difference. Negative rows are the list.

## What honest reporting refuses to do

Two refusals are worth more than the report.

**If you have not recorded what an hour costs, it should not invent one.**
Revenue and hours are still worth seeing. A margin computed against a number
nobody stands behind is worse than no margin, because you will act on it.

**A visit nobody timed should not be costed from its schedule.** If a visit
was booked for three hours and nobody recorded what it took, the honest answer
is that it is uncosted. Treating the booked duration as the actual one assumes
the very thing you are trying to measure — and it would report a healthy
margin on precisely the visits that are running long.

Which is why any report like this has to tell you how much of the month it
actually covered. "Nine of twelve visits costed" is a usable answer. A margin
with no idea behind it is not.

## What you do with it

Nothing dramatic, and that is the point. You look at it once a month, and
three or four times a year it tells you something:

- A site went from 34% margin to 11% over five months. Go and see why.
- A site has been negative since March. It is either re-bid at renewal or it
  is a decision you are now making deliberately instead of by accident.
- A site is at 48%. Do not touch it, and find out what is different about it.

The contract you re-bid at renewal because you had the numbers is worth more
than the one you kept because nobody looked. Underbid contracts do not blow
up. They bleed, quietly, for years — and the only thing that finds them is
looking at the right grain, on purpose, on a schedule.