How many hours can automation actually save your business? The honest maths

The short answer

For most businesses, automation realistically recovers somewhere between a meaningful slice of one full-time person and several, depending on headcount and how much of the week is repetitive administration. The honest way to find out is arithmetic on your own numbers, not a vendor's slide, and the arithmetic is simple enough to do on a napkin.

This article walks through it: the unit that makes savings claims comparable, a clearly hypothetical worked example, where the hours actually hide, and how to tell an honest estimate from a fantasy.

Start with the unit: 162.5 hours

"Save hundreds of hours" means nothing without a unit. The unit that makes time savings legible is the full-time equivalent, or FTE: what one full-time person works in a month.

The maths: a standard UK full-time week is 37.5 hours. Multiply by 52 weeks in the year, divide by 12 months:

37.5 × 52 ÷ 12 = 162.5 hours per month.

So one FTE is 162.5 hours a month, and every savings claim should be converted into it. "This will save you 80 hours a month" becomes "this frees roughly half a person". If anything the conversion flatters the human side, because 162.5 is a contracted figure that ignores holiday and sickness; real worked hours are fewer. Using it makes your equivalence estimate conservative, which is where an honest estimate should sit.

A worked example: the 20-person office

Take a deliberately ordinary, hypothetical office of 20 people, and a deliberately modest assumption: each person loses 7.5 hours a month to repetitive admin. That is under 25 minutes a working day, which most office workers would call an underestimate the moment they think about their own inbox.

Now apply the unit: 150 ÷ 162.5 = 0.92. Call it 0.9 of a full-time person. This ordinary office is spending nine tenths of an entire salary's worth of hours, every month, on work that software does without complaint, and nobody can see it because it is spread across 20 people in 25-minute slivers.

The numbers are an example, not a promise; your real figure might be half that or double it. The method is the point: count, multiply, divide by 162.5.

Where the hours actually hide

These hours survive every efficiency drive for a simple reason: they never appear in one place. No job description mentions them, no timesheet has a code for them.

• Double entry between systems. The enquiry arrives by email, gets retyped into the CRM, again into the job sheet, again into the invoice. The same ten fields, four keyboards, every single time.
• Chasing status. "Has the engineer been?" "Did the payment land?" "Where is that application up to?" Questions a connected system answers itself, currently answered by a person walking across the office.
• Manual reminders. Appointment confirmations, certificate expiry warnings, payment nudges: sent when someone remembers, missed when they do not.
• Report assembly. Every Friday, someone copies numbers out of three systems into one spreadsheet so that Monday's meeting can look at figures that were true on Thursday.

Nobody's timesheet says "retyping". The hours are real; they are simply spread too thin for anyone to see them in one place.

Why one umbrella beats point solutions

Look at where the hours in the table come from: almost all of them live in the gaps between systems. Retyping exists because the inbox does not talk to the CRM. Chasing exists because the job sheet does not talk to the diary. Report assembly exists because nothing talks to anything.

A point solution patches one gap and leaves the rest. Buy a reminder tool and you still retype; buy a form tool and you still assemble reports. Each tool saves a sliver, and each adds another system for the others to not talk to.

When departments are built as one connected platform, the savings compound instead: an enquiry typed once flows to the diary, the job, the invoice and the report without being touched again. This is the premise behind how we work at AevoMind: build each department properly, connect them under one umbrella, and staff the repetitive layer with AI employees who work inside the connected system rather than around it.

Honest estimate or vendor fantasy?

Automation marketing has a credibility problem, and it earns it. Here is how to tell the difference.

A fantasy estimate has these tells:

• Big round numbers with no unit and no working: "save thousands of hours".
• Counting 100% of a task when only part of it can be automated.
• Assuming every saved hour instantly becomes billable, productive time.
• No mention of exceptions, escalations or a bedding-in period.

An honest estimate starts from your actual volumes, converts everything into FTE at 162.5 hours, discounts for the work that still needs human judgement, and states its assumptions so you can check them. It should survive you changing the inputs.

That last part is why we put the arithmetic in your hands: the calculator on our platform page uses exactly this method, hours in, FTE out, assumptions visible, so you can run your own headcount rather than take our word for it.

What to do with the hours: the real decision

Suppose the maths says 0.9 FTE, as in the worked example. Nothing happens automatically next; you face a choice with exactly two honest branches. Run the same workload with fewer people, or keep the team you have and take on more business without hiring. The first shows up as lower costs, the second as growth with a flat wage bill. Both are legitimate; the mistake is not choosing.

Where to start counting? For most businesses the phone is the loudest leak, which is why the receptionist is usually the first role automated; we have written separately about what an AI answering service can and cannot do. And when you want the counting done properly, across every department rather than a napkin, book an assessment and we will do the arithmetic with you.