Run it with less chaos · Every business
Capacity Planning for Service Businesses: How Full Is Full?
A calendar at 100% looks like success, yet decades of queueing research show that a service business runs best with some room left in the week, and the useful question is how much.
At a glance
- Capacity planning is the habit of comparing the hours a business can actually sell with the hours customers want, by day and by hour, so the owner knows where the week is full, where it is empty and how much room to keep.
- Running full has a cost that most owners never see: a simulation study in the BMJ found that hospitals averaging 85% of their beds began to see shortages, and that at 90% or more they could expect regular shortages and periodic crises.
- In a simple queueing model, the wait for the next opening is about 5.7 times the length of a service at 85% busy and 19 times at 95% busy, so the last few points of fullness cost far more than the first.
- Measure booked hours against bookable hours for each day of the week, aim for a working range around 80% to 85% overall, and decide in advance what the spare 15% is for: same-day customers, rebooks, overruns and the waitlist.
- For an illustrative four-chair salon at 78% overall, closing the gap to 85% is worth about $46,800 a year, and a day-by-day split shows which days hold that room.
Part 1 | Full is a range, not a number
Most owners judge capacity by feel, and the feel comes from the busiest day of the week. We're slammed; we need another person. That sentence is often true on Saturday and false on Tuesday, and a business that hires for Saturday pays for Tuesday as well, which is why capacity planning starts with a count rather than a hire.
The count has two parts. Bookable hours are the hours a provider is on the schedule and free to serve a customer, after breaks, cleanup, admin and travel come out; booked hours are the hours actually sold. Divide one by the other and the result is capacity utilization, the share of sellable time that turned into service, and the number only becomes useful when it is split by day of the week and by hour of the day, because a weekly average hides the two problems owners actually have, which are turning customers away at peak and paying for empty time off peak.
The research on how full is too full comes mostly from hospitals, which have measured the question with more care than any other service industry. A stochastic simulation published in the BMJ modeled emergency admissions at a typical hospital and found that risks of a bed shortage became visible at about 85% average occupancy, and that a hospital running at 90% or more could expect regular shortages and periodic bed crises; the authors concluded that some empty, staffed capacity is a necessary cost of a reliable emergency service rather than waste.
A salon chair or a service van is not a hospital bed, but the mechanism is the same in any business where customers arrive unevenly and services run long or short. Demand arrives in clumps, so a schedule that is full on average is overfull on its busiest hours, and the overflow shows up as waits, rushed services, overtime and customers who book somewhere else.
Part 2 | Why the last ten percent costs the most
Queueing theory, the mathematics of waiting lines, explains why fullness behaves this way. A chapter on queueing analysis in healthcare, written for operations researchers and hospital planners, sets out the central result plainly: as average utilization rises, average delays rise at an increasing rate, with an "elbow" past which even a small increase in utilization produces a large jump in delay, and greater variability in arrivals or service times moves that elbow to the left.
The simplest queueing model, one provider with random arrivals and random service lengths, makes the curve concrete. The average wait in line equals the service time multiplied by utilization divided by one minus utilization, so a provider who is 60% busy produces waits of about 1.5 service lengths, while the same provider at 95% produces waits of 19.
In a booked business the wait does not appear as a line in the lobby; it appears as the number of days until the next opening. A client who wants a Saturday color and hears "three weeks out" is experiencing the right-hand side of that curve, and the same chapter notes that larger operations can run at higher utilization with the same delays, which is one reason a two-person shop needs proportionally more room than a twelve-chair salon.
The same research gives a useful sense of scale. In the chapter's worked obstetrics example, a unit sized so that only about 1% of patients would wait for a bed ran at 71.5% occupancy, and hospital planners in the United States long used 85% as a target even though actual average occupancy at nonprofit hospitals sat closer to 66%.
Part 3 | Measuring your own week
The capacity map takes one ordinary week to build. List every provider, write down bookable hours for each day, then pull booked hours from the calendar for the same days; four weeks gives a steadier picture, and the result is a small grid of percentages that shows exactly where the week is tight and where it is slack.
| Stylists on the schedule (Assumption) | 4 |
|---|---|
| Bookable hours per stylist per week (Assumption) | 36 |
| Bookable hours per week (Calculated (4 × 36)) | 144 |
| Booked hours per week (Assumption) | 112 |
| Capacity utilization (Calculated (112 ÷ 144)) | 78% |
| Hours available at an 85% target (Calculated (144 × 85%)) | 122.4 |
| Room to fill per week (Calculated (122.4 − 112)) | 10.4 hours |
| Average revenue per service hour (Assumption) | $90 |
| Weeks open per year (Assumption) | 50 |
| Revenue in that room per year (Calculated (10.4 × $90 × 50)) | $46,800 |
Illustrative example; replace every assumption with your own numbers.
The overall figure is only the starting point. If Friday and Saturday are at 100% and Tuesday and Wednesday are near 55%, the business does not have a capacity problem; it has a timing problem, and the fixes are scheduling fixes such as moving a stylist's day off from Saturday to Tuesday, a reason for flexible customers to choose midweek, and a waitlist for the full days, rather than a new hire.
Deciding what the spare room is for keeps it from being wasted. A sensible split for most appointment businesses is a few protected slots for same-day and urgent customers, a short buffer for services that run long, and open time the front desk can offer to the waitlist or to regulars who are due back, released on a set schedule if nobody claims it.
How you compare
- Hospitals: planners have long aimed at 85% occupancy, and the BMJ model found shortages beginning near that level; a service business averaging above 85% across the week should expect long lead times and turned-away customers at peak.
- Dental practices: in the American Dental Association's third-quarter 2025 survey of 770 dentists, 34.8% said they were not busy enough and could have treated more patients, while 30.4% were either too busy to treat everyone or overworked, a reminder that one industry can hold both problems at once.
- Your own baseline: your day-by-day utilization grid for the last four weeks, with the highest and lowest day circled, is the comparison that matters most.
Questions to ask this week
What are our bookable hours, really? Take scheduled hours and subtract breaks, cleanup, admin and travel before dividing anything.
Which day is over 90% and which is under 65%? The gap between those two days is usually the first project.
How far out is the next opening on our busiest day? A lead time of more than two weeks on one day and same-day openings on another is a timing problem.
What is our spare capacity for? Write down who gets the protected slots and when unclaimed ones are released.
The next step is to pull four weeks of bookings, build the day-by-day grid and circle the fullest and emptiest days before making any decision about hiring.
About this research: This article draws on a stochastic simulation of hospital bed use published in the BMJ, a chapter on queueing analysis in healthcare from an edited operations research volume, and the American Dental Association Health Policy Institute's quarterly economic survey. The queueing figures in Exhibit 1 are calculated from a standard textbook formula and show the shape of the curve rather than a forecast for any one business. The salon example is illustrative.
Sources
- Bagust, A., Place, M. and Posnett, J. W. Dynamics of bed use in accommodating emergency admissions: stochastic simulation model. BMJ 319:155–158, 1999.
- Green, L. Queueing Analysis in Healthcare. In Hall, R. (ed.), Patient Flow: Reducing Delay in Healthcare Delivery, Springer, 2006.
- American Dental Association Health Policy Institute. Economic Outlook and Emerging Issues in Dentistry, Q3 2025. 2025. 816 dentists; 770 answered the busyness question.
Related reading: How to Stay Booked in the Slow Season Without Discounting · What One Empty Appointment Slot Costs Over a Year · Booking Rules for Salons: Buffers, Service Times and Double-Booking · Utilization Rate: The Number Behind a Busy Week That Still Loses Money · Staff Scheduling: Building a Weekly Schedule That Matches Demand · Running a Waitlist That Fills Cancellations
Work with osNOVA on capacity and scheduling
We help owner-run service businesses map how full each day really is, then set up the schedule, waitlist and booking rules that move customers from the overbooked days to the open ones, starting from what you want more of, whether that is fuller weeks, shorter lead times for regulars or a calmer Saturday.
Email hello@osnovaops.com about your business
Or leave your email and we will reach out
Tell us what kind of business you run and what you want more of.
Not ready to talk? osNOVA runs free live events for owners.