How Many Employees Should I Schedule Each Shift at My Coworking Space?

Signals you actually need this

Most operators do not run this math until something breaks, and the breakage is remarkably consistent across spaces. Here are the tells that your shift schedule is set by habit rather than by demand.

Tours are getting handed off to whoever is closest. A booked tour is a revenue event. If your front-desk person is abandoning the desk mid-check-in to walk a prospect through the phone booths, you are one person short on that block — not because the desk is drowning, but because tours and desk coverage are two different jobs colliding in one body. Track how many tours got run by someone who was scheduled for something else. If it is more than one or two a week, the block carrying your tour load needs its own headcount line.

Packages pile up past noon. Mail and parcel handling is the most reliably underestimated task in a shared workspace. It arrives in waves — usually a morning carrier drop and a mid-afternoon one — and it does not wait. When members start Slacking the community channel asking whether their laptop stand arrived, the intake task is losing to check-in traffic. That is a staffing signal, not a process signal.

Your evening block runs the same headcount as your morning block. This is the single most common symptom of habit scheduling. Access-control data almost always shows a 3:1 or 4:1 spread between weekday-morning peak and weekday-evening trough, but the schedule shows two people on both. You are paying morning wages for evening traffic. The fix is not cutting the evening person — one person is the floor for any staffed block, because someone has to be able to open a door and answer a question — it is recognizing that the morning needed a third body you never scheduled.

Event setup happens after hours, unpaid or unplanned. If your community manager is staying late to break down a 30-person member mixer because the schedule ended at 6:00, you have an events load that is not represented anywhere in your coverage math. Event blocks are additive. A shift carrying setup, run, and teardown for a 30-person event needs a dedicated person for roughly the 90 minutes before and 45 minutes after — that is not front-desk time, and it cannot be borrowed from front-desk time without the desk going dark.

Members stop asking staff for things. This is the quiet one, and it is the most expensive. When a member gives up on flagging down a community manager and just leaves the printer jammed, you have crossed the ratio without seeing a single complaint. Churn signals are lagging indicators; unasked questions are leading ones. Ask your leads directly: how often this week did someone look like they needed help and you couldn't get to them?

You cannot answer "how many people are here right now?" from data. If the honest answer is a shrug or a guess, the ratio math has nothing to divide into. Access-control logs, badge swipes, Wi-Fi association counts, and desk-booking records are all acceptable proxies. Pick one, pull a trailing four to eight weeks, and average by day-of-week and time block. Do not average across the whole week — Monday morning and Friday afternoon are different businesses.

The same signals show up in adjacent operations. A gym reading badge-ins, a car wash reading bay throughput, a branch bank reading teller queue depth — all of them are solving the identical problem: a variable demand curve met by a fixed-cost roster. Coworking just happens to have unusually clean data, because everyone badges in.

What good looks like vs. bad

A bad schedule is flat. A good schedule is shaped like the building's actual traffic. That is the whole difference, but it plays out in specifics worth naming.

Bad: two people on every weekday block, 8:00 to 6:00, because that is what the last manager did. Total: 20 person-hours a day. The 9:00–noon window is underwater, the 3:00–6:00 window is staff standing around, and no one owns tours or event prep. When someone calls out sick, the whole day is a scramble because there is no defined minimum.

Good: blocks sized independently. Morning peak of 90 on-site members at a 50:1 ratio gives two desk staff, plus one tour-and-events lead because that block carries four booked tours and a lunch-and-learn setup — three people, 8:00 to 1:00. Midday settles to 60 on-site: one desk staffer plus the lead who stays through the event teardown. Evening runs 20 on-site: one person, closing checklist, out at 7:00. Total lands near 16–18 person-hours, but the hours are where the work is.

The delta is not really the two person-hours saved. It is that the morning stopped being understaffed. Flat schedules almost never save money on net — they misallocate. You get the cost of the slow hours and the service failure of the busy ones simultaneously.

Good practice: one defined minimum, always. Whatever the math says, no staffed block goes below one person. Below one you do not have a service ratio, you have an unattended building. If a block genuinely draws eight people, the honest answer is either "one person, and they also do the deep-clean checklist and inventory count" or "this block should not be staffed and we go keycard-only with on-call escalation." Both are real options. Pretending 0.4 of a person is on duty is not.

Good practice: separate the roles before you separate the bodies. Front desk is check-in, packages, member questions, printer and AV triage, and day-pass intake. Community/events is tours, meeting-room turnover, event setup and teardown, member programming, and the walk-the-floor pass. These are different work with different interruption profiles. Once you name them as two roles, the question "do we need a third person Monday morning?" answers itself — you are not adding a generic body, you are covering an unowned function.

Bad practice: staffing to membership count instead of on-site count. A 300-member space does not have 300 people in it. Utilization for a typical flex-membership base runs well under half of the roster on any given day, and hot-desk-heavy plans run lower still. Divide your *on-site* number by the ratio. If you divide total membership, you will triple your labor line and wonder why the margin vanished.

Bad practice: one ratio forever. The 50:1 figure is a starting yardstick, not a law. Recalibrate it quarterly against two things: how often your leads report they couldn't get to someone, and how many member-help tickets or Slack requests went unanswered past 30 minutes. Drifting to 40:1 during a heavy onboarding month is correct. So is stretching to 60:1 when your membership matures into heads-down regulars who badge in and disappear into a phone booth.

Reading the diagram in practice: you run it once per block, per day-of-week. A five-day week with three blocks a day is fifteen passes, which sounds tedious and takes about twenty minutes with a spreadsheet in front of you. You do it once properly, then you are adjusting deltas rather than rebuilding.

Real cost and ROI ranges

The staffing decision is a labor-line decision, so it is worth putting real ranges around it — with the caveat that wages vary enormously by market and you should substitute your own numbers.

The labor line itself. Community-manager and front-desk roles in a shared workspace are hourly or salaried-hourly positions. Take your fully loaded hourly cost — wage plus payroll taxes plus benefits plus paid time off, which typically lands meaningfully above the base wage — and multiply by scheduled hours. That is the number the ratio math moves. A single unnecessary person-hour per weekday is roughly 260 person-hours a year. At any realistic loaded rate, that is a line item you would notice on a P&L.

The asymmetry that matters. Overstaffing costs you the wage. Understaffing costs you a tour that did not convert, a member who churned, and a renewal conversation that got harder. The wage is a known, bounded cost. The other is unbounded and invisible. This asymmetry is why the ratio should be a service floor rather than a stretch target — you tune it toward coverage and let the schedule shape absorb the savings, instead of tuning it toward thinness and hoping nothing breaks.

Where the money actually comes back. Three places, in rough order of size:

*Tour conversion.* A booked tour that gets run properly by someone whose job that block is running tours converts better than one squeezed between check-ins. If your space converts tours at some baseline rate and a dedicated tour block moves that rate at all, the math is not close — a single additional membership recovered per month typically exceeds the cost of the incremental staffing that made it possible. Track tours-run-by-assigned-owner as a metric and watch the conversion delta yourself; do not take a generic number for it.

*Retention through response time.* Members do not churn over one jammed printer. They churn over an accumulated sense that nobody is minding the place. Response time to member requests is the closest proxy you can actually measure, and it degrades sharply once you cross the ratio. The retention value of a single saved membership is the monthly rate times the remaining tenure you would have gotten — for most spaces, a number in the hundreds to low thousands.

*Meeting-room and event revenue.* Rooms that get turned over promptly get booked again. Rooms that sit dirty until someone notices lose the next slot. If meeting-room and event revenue is a meaningful share of your top line, the person who does turnovers is paying for themselves out of utilization, not out of goodwill.

Cost of the tooling layer. You do not need software to do this math — a spreadsheet and your access logs are genuinely sufficient for a single location. If you do want a scheduling app, the market splits along pricing model, and that split matters more than feature lists:

*Per-location pricing* favors a single space with a large part-time roster. If you run twelve part-timers across a week, paying per location rather than per head is a large structural discount. Several vendors in this category offer a free single-location tier with unlimited employees, which is often enough to run publishing, availability, and time-clock for a first space.

*Per-user pricing* favors a lean, stable crew. Four full-timers on a low per-user monthly rate is cheaper than any per-location plan. It stops being cheaper the moment you add a bench of weekend part-timers.

*Demand-driven tiers* — the ones that ingest a booking or occupancy feed and suggest coverage — cost more per seat and are worth it only once you have enough sites or enough volatility that hand-running the division is a real burden. For one space with three blocks a day, the suggestion engine is solving a problem you can solve in twenty minutes a quarter.

*Compliance features* are priced as a premium and are not optional if you operate in a fair-workweek jurisdiction or run 24/7 access with overnight coverage. Predictive-scheduling penalties for late schedule changes are a real exposure in several US cities, and the guardrails in the heavier platforms exist because that exposure is expensive. Check your local rules before deciding this is a nice-to-have.

Payback timing. The method itself pays back in the first month you apply it, because reshaping a flat schedule is free — you are moving hours, not adding them. Tooling payback depends entirely on roster size. A twelve-person part-time roster on a per-location plan typically justifies itself against the time saved on schedule publishing and swap management alone, before you count any coverage improvement.

How it plugs into your workflow

The ratio math is not a one-time exercise; it is a small recurring loop that hangs off systems you already run. Here is where each piece attaches.

Upstream: your access control and booking systems. This is the data source, and it is the only genuinely non-negotiable input. Badge or keycard events give you on-site headcount by timestamp. Desk-booking and meeting-room systems give you committed load. Export a trailing four to eight weeks, bucket by day-of-week and block, and take the average — or, if you would rather protect against understaffing, take the 75th percentile instead of the mean. Peaks are what break service; averages hide them. Most operators should use the mean for planning and the peak for sanity-checking their minimum.

Sideways: your CRM or tour pipeline. Booked tours live in whatever system your sales or membership pipeline runs in. That is a RevOps handoff in miniature — the pipeline knows about demand that the scheduling system cannot see, and the coverage plan has to reflect it. If you have three tours booked Thursday afternoon, Thursday afternoon needs a tour owner, regardless of what the badge data says about headcount. Pull the coming week's tour calendar every Thursday or Friday when you build the next week's schedule.

Sideways: your events calendar. Same logic, bigger blocks. Member programming, tenant-hosted events, and external room rentals each carry setup, run, and teardown. Attach a staffing requirement to the event when it is booked, not when it is imminent. The cleanest way to do this is to make "staffing owner" a required field on the event record — if nobody can be named, the event does not get confirmed.

Downstream: the published schedule and the clock. Once the counts are set, publishing is pure logistics — get the schedule onto phones, handle availability and swaps, and capture actual hours worked. Any competent scheduling app does this. The important discipline is that swaps must not silently break the coverage math: if two people trade and the Monday-morning block drops from three to two, someone should see that before it happens.

Downstream: the labor-cost review. Compare scheduled hours against actual clocked hours monthly. Systematic overage on a block usually means the block is genuinely busier than your average suggested — that is your recalibration signal, and it is more reliable than anyone's memory of how the week felt.

The review cadence that keeps it honest. Weekly, you adjust for known events and known absences — that is a ten-minute pass. Monthly, you compare scheduled versus actual hours and look for blocks that consistently ran over. Quarterly, you revisit the ratio itself against member feedback and unanswered-request data. Annually, or whenever you cross a membership threshold or add a floor, you rebuild the block model from scratch rather than editing the old one.

Absence buffer, handled properly. Do not pad every block by half a person. Pad the roster instead: keep one or two cross-trained part-timers who can cover any block on short notice, and know in advance which blocks degrade gracefully (a quiet Friday afternoon) versus which do not (Monday morning, or any block with a booked tour). When a callout hits, you are moving a known person into a known gap rather than improvising.

Where this generalizes. The same loop — demand signal in, ratio applied, coverage out, actuals compared, ratio recalibrated — is what a service desk does with ticket volume, what a warehouse does with pick volume, and what a sales org does when it sizes a team against pipeline. The vocabulary changes; the division does not. If you already run any kind of capacity model elsewhere in the business, you can borrow its cadence wholesale and just swap the numerator.

Related questions

How do I set the initial ratio if I have no data?

Start at 50 on-site members per staffer and run it for four weeks while logging every instance where a member waited or a task slipped. If those instances cluster in one block, that block's ratio is wrong. Adjust that block, not the whole model.

Should day-pass guests count toward the headcount?

Yes, and weight them slightly heavier. A day-pass guest needs onboarding, a Wi-Fi walkthrough, and orientation — meaningfully more attention than a badge-in regular. Counting each guest as 1.5 members for ratio purposes is a reasonable working adjustment until your own data suggests otherwise.

Does a 24/7 access space need overnight staff?

Usually not. Most 24/7 spaces run staffed hours plus keycard-only access outside them, with on-call escalation and camera coverage. Overnight staffing becomes necessary when you have physical security requirements, regulated tenants, or enough overnight traffic to generate real service demand.

How does this change with multiple floors or locations?

Each floor with its own entry point and amenity cluster is effectively its own block model — you cannot share one desk person across two floors and call it covered. Multi-site groups should build per-site models and then look for shared roles (roving events lead, floating cover) across them.

What if my busiest block is also my most expensive to staff?

That is normal and it is not a reason to understaff it. Look instead at whether the peak can be flattened: shift package intake to a quieter window, move recurring tours off the busiest hour, or stagger start times so coverage ramps rather than steps.

FAQ

What is the single most important factor in deciding how many staff to schedule?

Your agreed service ratio — how many on-site members one staffer can serve well. That number, typically somewhere in the 40 to 60 range depending on how high-touch your community is, is the foundation of every shift calculation. Without it you are not scheduling, you are guessing and calling it a plan.

How do I know how many members will be on-site during a given shift?

Pull historical check-in data from your access-control system or booking platform. Look at the same day of week and the same time block across the past four to eight weeks and take the average. That headcount is what you divide by your service ratio. Never average across the whole week — Monday morning and Friday afternoon behave nothing alike.

What if my space has a lot of events or meeting-room bookings?

Add staff for setup, teardown, and guest management as a separate line from front-desk coverage. A quiet block with 20 members on site needs one person, but if that same block carries a 30-person event, you need a second body owning the event end to end. Event load is additive, never borrowed from the desk.

Should I schedule the same number of Employees for every Shift of the day?

No, and flat schedules are the most common mistake in this category. A morning peak may need three people while the evening block needs one. Build each block independently off expected on-site headcount plus its own task volume — tours, mail, packages, room turnovers, events.

What if my members are particularly high-need or low-need?

Adjust the ratio, not the method. A high-touch community with heavy day-pass traffic and constant onboarding runs closer to 40 members per staffer. A mature, heads-down membership that badges in and works quietly can stretch toward 60. Calibrate from your own unanswered-request data over a quarter rather than adopting someone else's number.

How often should I revisit the Coworking staffing plan?

Weekly for known events and absences, monthly against scheduled-versus-actual hours, and quarterly for the ratio itself. Rebuild the whole block model from scratch whenever you cross a meaningful membership threshold, add a floor, or change operating hours — editing a stale model tends to preserve its original wrong assumptions.

Sources

• https://www.bls.gov/ooh/ — U.S. Bureau of Labor Statistics Occupational Outlook Handbook, for wage and role data on front-desk and customer-service occupations.
• https://www.bls.gov/news.release/ecec.nr0.htm — BLS Employer Costs for Employee Compensation, for fully loaded labor cost versus base wage.
• https://www.dol.gov/agencies/whd/flsa — U.S. Department of Labor, Fair Labor Standards Act overview covering hours worked and overtime.
• https://www.sfgov.org/olse/formula-retail-employee-rights-ordinances — San Francisco Office of Labor Standards Enforcement, predictive-scheduling ordinance requirements.
• https://www.seattle.gov/laborstandards/ordinances/secure-scheduling — City of Seattle Secure Scheduling Ordinance, advance-notice and schedule-change rules.
• https://www.sba.gov/business-guide/manage-your-business/hire-manage-employees — U.S. Small Business Administration guidance on hiring and managing employees.
• https://hbr.org/2015/11/the-truth-about-open-offices — Harvard Business Review on shared workspace dynamics and utilization behavior.
• https://www.jll.com/en-us/insights — JLL research insights on flexible workspace and office utilization trends.
• https://www.cbre.com/insights — CBRE insights on office and flexible workspace occupancy data.

Related on PULSE

• [How Many Attendants Should I Schedule Each Day at My Car Wash?](/knowledge/tl0067)
• [How Do I Figure Out How Many People to Schedule Each Day and at What Times for My Single Store?](/knowledge/tl0002)
• [How Many Sales Reps Do I Need to Hire for My Logistics Company?](/knowledge/tl0058)
• [How Many Salespeople Do I Need to Hire for My Car Dealership?](/knowledge/tl0052)
• [How Many Producers Do I Need to Hire for My Insurance Agency to Grow My Book?](/knowledge/tl0015)