The capacity math: what 10 hours a week is worth to a professional services firm
“It’ll save you hours” is the vaguest promise in automation. Every tool makes it. Very few help you work out whether those hours are worth anything to your firm.
So let’s do the math properly, using your numbers rather than ours.
How many hours is 10 hours a week, really?
Ten hours a week over a working year of about 45 weeks (allowing for holidays and quiet periods) is about 450 hours a year.
A full-time person works roughly 1,600–1,800 hours a year once you take out leave and holidays. So 10 hours a week is about a quarter of a person’s working time.
That’s a useful way to think about it. Automating 10 hours a week of admin is a bit like adding a quarter of a person to the team, without hiring, onboarding or managing them.
How do you calculate what freed-up hours are worth?
There are three honest ways to value the time. Which one applies depends on what your team will do with it.
1. The cost view: what the hours cost you today
Annual cost = hours saved per year × loaded hourly cost
“Loaded” means salary plus taxes, benefits, software and desk costs. As an example, if an administrator’s loaded cost is $30 an hour, 450 hours costs about $13,500 a year.
This is the smallest number, and the least interesting. It assumes you’d save money, which only happens if you’d otherwise cut hours. Most growing firms won’t.
2. The revenue view: if the hours go into billable work
Annual value = hours redirected to billable work × effective hourly rate
If the time comes back to someone who bills, or who’d otherwise be doing client work, the value is much higher. Say a senior’s effective rate is $100 an hour and half the freed time (225 hours) turns into client work: that’s about $22,500 a year. At $150 an hour, it’s about $33,750.
Be conservative here. Not every saved hour becomes a billed hour.
3. The capacity view: if the hours go into new clients
This is the one that matters most for firms with more demand than staff.
New clients you can serve = hours freed ÷ hours each client needs per year
Annual value = new clients × average annual fee
Suppose a bookkeeping client takes about 40 hours a year to serve, admin included, and pays $4,000 a year. Freeing 450 hours could let you take on roughly 10 more clients, worth around $40,000 a year in fees, without a new hire.
Swap in your own numbers. A recruiting agency might count placements, and an insurance broker might count policies or renewals handled. The structure is the same: hours freed, divided by hours per unit of work, times what each unit earns.
Where do 10 hours a week come from?
Rarely from one big automation. Usually from three or four small ones. In an accounting firm, for example, our estimates for a typical 10–100 person firm look like this:
| Automation | Estimated hours saved / week |
|---|---|
| Document chasing | 3–8 |
| Inbox triage and drafted replies | 3–6 |
| Meeting notes to tasks | 2–4 |
| Fee chasing and collections | 1–3 |
These are estimates, not measured results. Your numbers will be different, so work them out with this formula:
Hours saved per week = times per week × minutes each × share automated ÷ 60
Count minutes honestly, including the lookup, the context switching and the logging, not just the typing. And be conservative on the share automated: start at 50–70%. There are always exceptions, and a person should handle them anyway.
For a full walkthrough of the biggest one, read How accounting firms can automate document chasing with AI.
What does it cost to get those hours back?
The other side of the equation is what you spend. Be honest about:
- Build cost: a one-off fee, whether that’s your own team’s time or an outside build.
- Running cost: software, AI usage, and someone looking after it. Automations need monitoring, and they break quietly when tools change.
- Your team’s time: a few hours to map the work, test the drafts and get comfortable.
A good automation pays for itself within months on the capacity view, not years. If the math only works on the cost view, it may not be worth doing yet.
Why decide in advance where the hours go?
This is the step most firms skip. Freed time doesn’t automatically turn into revenue. Left alone, it fills up with other admin, or with more careful versions of the same admin.
Before you automate, write down where the hours will go: “take on 10 more bookkeeping clients this quarter”, “partners stop working Saturdays in January”, “seniors spend Friday afternoons on advisory work”. Then check back after a month.
That’s also how you’ll know whether it worked.
Next step
If you want to run these numbers for your firm properly, workflow by workflow, that’s the first thing our AI Ops Audit does: map where the hours actually go, estimate the savings for each opportunity, and rank them by value, effort and risk. Book a 30-minute call, or get the free Capacity Playbook, which includes a scoring sheet to pick your first automation.
FAQ
Isn't this just a way to justify cutting staff?
Not for most firms we talk to. They have more work than people. The hours go into serving more clients, cutting overtime and giving seniors back time for the work only they can do.
What if the freed hours don't turn into revenue?
That's a real risk, and it's why you should decide in advance where the hours go. If nobody plans for it, saved time tends to disappear into other admin. Tie each automation to a specific use, such as taking on X more clients this quarter.
How do we know how many hours we'd actually save?
Measure before you automate. For each task, count how often it happens and how long it takes, then apply a conservative share for what automation will handle. Our audits do exactly this, workflow by workflow.