Most labor cost conversations in hotels start with the wrong question: how many people do we need. That question assumes headcount is the lever. It usually isn't. The lever is the shape of demand against the shape of staffing, and the gap between them is where margin actually leaks.
Staffing for the worst hour, paying for it all day
Hotels tend to staff for the busiest plausible moment of a shift, then carry that staffing level through the quiet hours around it. A property might build its front desk schedule around a 3pm-6pm check-in surge, but that same headcount sits mostly idle from 9am to noon. Housekeeping is scheduled to clear a full house by 3pm on a day when occupancy is 68 percent. Room service staffs for a banquet that ended two hours ago.
This is coverage staffing, not demand staffing. It feels safe because nothing gets visibly missed, but it means labor cost is set by the peak, not the average. A property running at 70 percent labor-cost-to-covered-hours efficiency during peaks might be running closer to 35 percent during troughs, and the blended number never shows up clearly on a P&L broken out by department.
The fix is not simply cutting shifts. It's re-mapping shift start and end times against actual arrival, departure, and request patterns pulled from the PMS, rather than against tradition or union-era shift blocks that predate the property's current occupancy mix.
The coordination tax hiding inside headcount
A second, less visible driver of labor cost is coordination work that never appears on a job description. Front desk staff relay maintenance requests by radio, then follow up twice because nobody confirmed the room was fixed. Housekeeping supervisors walk floors to check which rooms are ready rather than seeing status update automatically. A guest calls three times about a late checkout because the first two requests never reached the person who could approve it.
None of this is on anyone's task list, but it consumes real hours. In a mid-size property, an illustrative estimate might put coordination and status-chasing at 45 to 90 minutes per shift, per department, spread across front desk, housekeeping, and engineering. Multiply that across a year and it is the equivalent of a part-time role that produces zero guest-facing output.
The hours a hotel spends managing the handoff between departments are hours it isn't spending on the guest in front of staff.
This is also why adding headcount to fix a service problem often doesn't work. The new hire absorbs into the same broken handoff pattern and the coordination tax simply scales with them.
Why understaffing costs more than overstaffing
Labor cost discipline can overshoot. Cutting a housekeeping shift to hit a budget target looks like a win in the labor line, but if it slows turnaround and pushes back three check-ins, the property may eat a rate concession, a room upgrade, or a negative review that costs more in future bookings than the shift itself did in wages.
A useful mental model: staffing cuts have a visible, immediate savings and an invisible, delayed cost. Recovery costs from understaffing tend to show up in different line items weeks later, which makes them hard to trace back to the original staffing decision. This is part of why departmental cost audits often miss the real driver of margin erosion.
A more accurate way to plan staffing
Rather than setting headcount targets by department, a few properties are shifting to a demand-matched model built on three practices.
- Tiered response instead of dedicated coverage. Cross-trained staff handle a queue of prioritized requests rather than each department waiting for its own idle capacity to be needed.
- Routing based on request type and urgency, not job title. A guest request for extra towels doesn't need to pass through a supervisor before reaching housekeeping; a maintenance issue affecting a VIP stay should jump the queue automatically rather than sit in a general inbox.
- Staffing built from occupancy and arrival data, not last year's schedule template. Shift start times shift with the actual flow of the week rather than staying fixed regardless of midweek versus weekend patterns.
This is the space where automation earns its keep, not by replacing staff but by removing the coordination layer between them. Platforms like Hermes, which connects into a property's PMS to route guest requests and surface relevant guest context directly to the person who needs it, are built around this exact gap: reducing the manual relay work that inflates headcount needs without adding guest-facing capacity.
What to measure instead of headcount
Properties that get this right tend to track a different set of numbers than raw labor cost percentage. Worth watching:
- Average time from request logged to request resolved, by category
- Labor cost per occupied room during peak versus trough hours, tracked separately
- Percentage of staff time spent on coordination and status-checking versus direct guest service, sampled periodically
- Recovery costs (comps, rate adjustments, upgrades) attributable to service delays
None of these numbers are as simple to report as total labor cost as a percentage of revenue, but they are the numbers that actually explain why that percentage moves. Staffing decisions made against demand patterns and coordination data hold up better over a full year than decisions made against a target headcount, because they address the actual shape of the problem rather than its headline symptom.