134. Stop running your cleaning business from memory

Commercial cleaning is a low-margin business.

You don’t need a disaster to lose money.

You just need 15 minutes over-servicing here, a missed visit there, and a supervisor driving across town for cover, with nobody tracking the pattern.

That’s how margin disappears.

Quietly.

When I spoke to Avin Rabheru, founder and CEO of Housekeep, I wanted to understand what proper operational control looks like.

(You can watch or listen to the full conversation at the end of this issue.)

Not the glossy version.

The practical version.

The version that shows up in rotas, routes, proof of service, client communication, and site-level margin.

Most cleaning businesses do not fail because nobody is trying hard enough.

They fail because too much of the operation is invisible.

The rota lives in one person’s head.

WhatsApp becomes the workflow.

Timesheets arrive late, messy, or “adjusted”.

Quality only becomes visible when a complaint lands.

Credits get issued, but nobody fixes the root cause.

One supervisor knows which client is difficult, which cleaner has the keys and which site always runs over.

That’s not an effort problem.

It is a system problem.

And in a low-margin business, system problems become margin problems fast.

One of Avin’s most useful points was also the simplest.

A cleaning business is not just a cleaning business.

It is a network of:

When you buy a cleaning business, you’re buying that network.

Not just the contracts, the revenue and the client relationships.

You’re buying years of accumulated patterns:

That network is either an asset or a liability.

If it is visible, repeatable, and measurable, you can improve it.

If it sits in memory, WhatsApp, and inboxes, it becomes fragile.

Hard to integrate, hard to scale, hard to value.

That’s why operational control matters.

The business isn’t just the client list.

It’s the system that turns contracted hours into delivered outcomes at a margin.

When I asked Avin what technology every cleaning business should use, he gave me four basics:

No AI or software-for-the-sake-of-software.

Basic infrastructure to stop the business from depending on one or two people holding the entire operation together.

That’s the real shift.

From operator-led memory to system-led control.

The biggest commercial lever Avin kept coming back to was density.

Not density as a vanity metric.

Density as economics.

More jobs are close together. Less dead travel. Cleaner shifts. Fewer handoffs. Easier supervision. More predictable quality.

But there is a catch.

Density only creates margin if the routes actually work.

You can have ten sites in the same town and still lose money. Time windows clash, cleaners sit around between jobs, supervisors are plugging gaps, and payroll hours drift away from contracted hours.

This is where a lot of small cleaning businesses leak money.

Not through one dramatic failure.

Through thousands of tiny bits of operational drag that nobody has isolated.

A route that should take three hours takes three and a half.

A site gets 15 minutes of free cleaning every visit because nobody wants the awkward client conversation.

A supervisor spends half a morning arranging cover because absence rules were never defined.

None of these looks fatal on its own.

Together, they eat the margin.

Avin linked density to cleaner economics and service quality.

If you want better cleaners, they need to be properly paid.

If they are going to earn properly, the work needs to be planned.

That pushes you straight into routing, utilisation, and operational control.

My conversation with Avin was in 2023.

There was no talk of AI.

Today, that’s changed.

AI can help with scheduling, exception reporting, inspection workflows, photo review, complaint patterns, and margin analysis.

But only if the operation is already structured.

First, you need to capture the basics:

Then AI has something useful to work with.

Without that, you are just asking software to make sense of chaos.

Automation does not fix chaos.

It scales the structure underneath it.

This is the bit operators need to be careful with.

The temptation is to look for the tool that fixes the business.

But the tool isn’t the operating system.

The operating system is the discipline of capturing, measuring, and improving the work.

Software makes that consistent.

Avin’s framework now sits inside how I think about acquisition diligence.

Before buying a cleaning business, I want to understand the real operating model.

Not just the P&L.

The work.

Because the accounts tell you what happened.

The operation tells you whether it can be improved.

Export every site with:

Then map it.

I’m looking for clusters.

I’m also looking for orphan sites, bad routes and route crossover. Work that looks acceptable on paper but creates drag in reality.

That’s where “wrong work” hides.

Get the actual rota.

If that does not exist, reconstruct it from timesheets, payroll, and WhatsApp.

Then compare:

This is where you find free cleaning.

Over-servicing.

Under-recovery.

And jobs where the margin was never real in the first place.

Ask one blunt question:

Do you track travel time today?

If the answer is no, you already know where to look.

Shadow a few shifts, and you will usually find:

Travel time is one of those leaks that hides in plain sight.

Everyone knows it exists.

Fewer operators measure it.

How often are visits missed?

Who notices first?

How fast does cover get arranged?

In weak systems, the client finds out before the business does.

That’s a churn engine.

It also tells you something important about control.

If the business only knows there is a problem when the client complains, it is not managing quality.

It is reacting to failure.

How does the client know the job was done?

A sign-in book?

A key log?

Photos?

A checklist?

Anything?

If the answer is “trust”, one complaint can turn into margin pain fast.

Trust is good.

Proof is better.

Especially when you are trying to reduce credits, defend pricing, and build a business that doesn’t rely on the founder personally knowing every site.

My deal-gating line is simple:

If I cannot get to these answers in diligence, I treat that as a red flag.

It usually means the margin is narrative rather than measurable.

The first 100 days are not about transformation.

They are about stabilisation.

The urgent work is always:

You have to protect what works before you start changing things.

If the systems workstream waits until later, you’ll hit day 100 with a stable business that remains operationally invisible.

That’s not enough.

So the tech-first workstream has to run alongside stabilisation.

Quietly.

Practically.

Without pretending the business is ready for a full rebuild on day one.

The first job is to make the work visible.

That means:

Not aggressively.

Not in a way that alienates the team.

Firmly enough that the business starts seeing itself.

Once the work is visible, you can start reducing firefighting.

That means:

The aim is not perfection.

The aim is to prevent the same problems from recurring without anyone owning them.

Only then do you start rebuilding the network.

That means:

The goal is simple:

Create capacity without adding more people.

That is where the margin starts to move.

If your scheduling lives in someone’s head, you do not have a people problem.

You have a system problem.

The point is not shiny tools.

The point is control.

Control means knowing what was sold, what was delivered, what it cost, and where the margin leaked.

That’s what protects profit.

It’s also what makes a business easier to buy, improve, and integrate.

Get the work visible.

Then improve it.

Then let technology speed up what is already under control.

My full interview with Avin Rabheru is live on The Growth Lab Podcast.

Listen to on Spotify/Apple Podcasts

Watch on YouTube

That’s all for this week.

Matt Harris

The Growth Lab

If this was useful, forward it to one operator who’s thinking about acquisitions. That’s the best way to help grow this.

The Growth Lab Capital is actively looking at Soft FM businesses across London, the Home Counties, the Midlands and the South East. If you’re an operator thinking about what comes next — even if it’s a few years away — I’m happy to have a conversation. No broker. No pressure. Book a time here.

The Growth Lab newsletter is a playbook for buy-and-build by Soft FM operators. Deals, growth, operations — no theory, no fluff. Subscribe for the next one.

💼 Daily on LinkedIn — Soft FM buy-and-build, deals, and market intel across FM, housing associations, and construction. 4,000+ follow. Join them here.

𝕏 Building this in public on X. Soft FM. Buy & Build. Same operator, less polish. Follow here.