A guest calls the front desk to report that their room's air conditioning isn't working. It's 3pm on a Saturday in July.
This is a failure. But it's not the equipment that failed first; it's the process.
The AC unit didn't fail suddenly. It degraded. Filters got dirty. Coils got fouled. Refrigerant slowly dropped below optimal charge. The unit kept running, consuming more energy per BTU, until the load of a summer Saturday was enough to tip it into a fault condition.
Every step of that degradation was predictable. None of it was predicted.
The cost of reactive maintenance
Reactive maintenance (fixing things when they break) is the default mode for most hotels. It feels efficient because you're only spending money when there's an actual problem. This intuition is wrong.
The true cost of a reactive maintenance call is substantially higher than the labor and parts to fix the problem:
Guest impact. A guest with a broken AC in July is a guest who may not return and may leave a review describing the experience. At $150/night ADR and even a 20% impact on a two-night stay, a single failure event has real revenue consequences that dwarf the repair cost.
Emergency labor rates. A Saturday evening HVAC call goes to whoever is on-call, at overtime rates, and may require an outside contractor. The same repair scheduled during a weekday midmorning costs a fraction as much in labor.
Secondary damage. Equipment that fails catastrophically often causes secondary damage that a pre-failure service call would have prevented. A coil that freezes from low refrigerant can cause compressor damage. A water heater that scales without being flushed eventually fails and may cause water damage. The cost of the secondary damage frequently exceeds the cost of the repair.
Reliability compounding. Assets that receive reactive maintenance have shorter service lives than assets that receive proactive maintenance. The AC unit that gets serviced quarterly outlasts the one that only gets attention when it breaks, often by years.
What proactive maintenance actually requires
The concept of preventive maintenance is well understood. Most hotels have some version of it: filter changes every 90 days, boiler inspection annually, elevator certificate on schedule.
The execution is harder, and the failure mode is consistent: the schedule lives in a spreadsheet or a paper calendar, it's owned by one person who may or may not execute it on time, and there's no systematic tracking of whether it's actually happening.
True proactive maintenance requires three things:
1. Asset inventory. You need to know what you have. Not just "HVAC units on floors 2–6" but specific assets with specific identifiers, service histories, and age. The 8-year-old rooftop unit has a different maintenance schedule than the 2-year-old unit that replaced it, and treating them identically is a scheduling error.
2. Systematic scheduling. The maintenance schedule needs to be generated from the asset inventory and executed as a series of tickets, not a calendar entry that depends on someone remembering to look at a spreadsheet. Each ticket should route to the right person, confirm completion, and capture what was found during the service.
3. Failure pattern tracking. The most valuable data in a maintenance program is repeat failures. A room that generates three AC complaints in a season isn't having bad luck; it's telling you that the root cause was never addressed. Systematically tracking repeat issues and escalating them to root-cause investigation is what separates a maintenance program from a maintenance queue.
The repeat-issue problem
In hotels without systematic maintenance tracking, repeat issues are invisible. The front desk gets three complaints about room 312's AC over a summer. Each complaint generates a separate work order. Each work order gets closed when the immediate problem is resolved. No one sees the pattern.
With ticket history by asset or room, the pattern is visible immediately. Three tickets in four weeks for the same room is an automatic flag: this isn't a maintenance issue anymore, it's an asset issue. The coil needs to be cleaned, the refrigerant needs to be charged, or the unit needs to be replaced. Whatever the root cause is, it needs to be found and addressed, not patched again.
The same logic applies to assets with high frequency-of-failure across the portfolio. If you have twelve rooftop units of the same model and age, and three of them are generating repeat tickets, the other nine should be getting proactive inspections now, not after they fail too.
Building the maintenance rhythm
A functional proactive maintenance program runs on a rhythm: daily checks, weekly inspections, monthly services, quarterly deep maintenance, annual certification and replacement planning.
Each level of the rhythm should generate tickets automatically, route to the right team, and confirm completion before closing. The GM should see a dashboard that shows the rhythm running, not a spreadsheet they have to manually check.
The most important metric isn't how many tickets you closed. It's the ratio of proactive-to-reactive tickets. A mature maintenance program skews heavily proactive. A program that's mostly reactive is a program that's spending too much money, generating too many guest incidents, and shortening its asset lives.
The Saturday in July, revisited
Go back to the guest with the broken AC on a Saturday in July.
In a reactive maintenance program, this is an incident: reactive ticket, emergency call, guest impact, possible review.
In a proactive maintenance program, the same asset had a service visit six weeks earlier. The tech noted the coil showed early fouling, cleaned it, checked the refrigerant charge, and closed the ticket with a note to re-inspect in 30 days. The re-inspection happened. The unit went into July in good condition.
The guest in that room never noticed anything. There was nothing to notice.
That's what a functional maintenance program looks like: not a department that responds to problems, but one that prevents them.
The capabilities behind this dispatch
Where the ideas in this piece become day-to-day operations.