Preventive Maintenance at a Resort: Making Recurring Work Actually Recur

Every property has a preventive maintenance schedule. Almost none of them run it as written.

The schedule exists — usually a spreadsheet, sometimes a binder, occasionally a screen in a system nobody logs into. It says the air conditioning filters get checked quarterly and the water pumps get serviced twice a year. And in a quiet month, that mostly happens. Then a full season arrives, reactive work fills every hour, and preventive work becomes the thing that gets done when there’s time. There is never time.

The result is a predictable cycle: preventive work slips, reactive work increases because things fail that would have been caught, and the increased reactive load makes preventive work slip further. Most resorts I’ve seen are somewhere on that loop, and everyone involved knows it.

Why the schedule doesn’t survive contact with a season

Three mechanics, all mundane.

Preventive work has no complainant. A guest reports a broken air conditioner. Nobody reports a filter that hasn’t been checked. Every triage decision between “the guest in 42 is hot” and “the quarterly filter check” resolves the same way, correctly, forever. Without a forcing function, preventive work loses every single time — not through bad judgment but through good judgment applied repeatedly.

The schedule lives outside the queue. If reactive work arrives in one place and preventive work lives in a spreadsheet, an engineer starting a shift looks at the first and not the second. Work that isn’t in the queue where people look isn’t really scheduled; it’s aspirational.

Nothing records the miss. When a quarterly check doesn’t happen, usually no artifact is created. Next quarter comes and the item is due again, and there’s no visible history saying it’s been skipped three times running. The failure is invisible until something breaks, at which point it’s diagnosed as bad luck.

What has to be true instead

Recurring work has to appear as work, in the same place as everything else. Not as a calendar entry, not as a tab — as a task in the queue the engineer already reads, indistinguishable in form from a reactive job. It’s the only version that survives, because it doesn’t ask anyone to look somewhere new on a busy day.

A skipped occurrence has to leave a mark. If preventive work is generated as tasks, a skipped one is a task that’s still open or was cancelled. Either way there’s a record. That’s what turns “I think we’re behind on filters” into something a head of maintenance can put in front of a GM in a budget conversation.

Due dates have to account for seasonality. A quarterly schedule that lands a large batch in your peak month will not be done in your peak month. Better to move it deliberately than to let it fail silently and pretend the schedule is intact.

Inspections belong in the same system. An inspection falling due is structurally identical to a recurring task — a calendar fact that should produce work without anyone remembering. If inspections live in a separate compliance folder, they’ll drift for the same reasons.

In our own boards, recurring work and inspections-falling-due generate tasks the same way a checkout generates a turnover job: the trigger is an event or a date, so the system creates the work rather than waiting for a person to remember. It sits in the same queue as reactive work, on the same board, and it carries a status that outlives the shift.

The one thing that makes the argument winnable

Most heads of maintenance already know all of this. What they lack is evidence, because the missing preventive work is invisible by construction.

Once recurring work is generated as tasks, the backlog stops being a matter of opinion. A scheduled job nobody did is an open task past its due date, and a count of those is something a head of maintenance can put in front of a GM — not a dashboard, a countable queue. Without it, it’s an anecdote about the villa 12 air conditioner; with it, it’s a line item.

That count answers how much scheduled work is overdue. Which unit is failing repeatedly is a different question, and answering it means tying reactive visits back to a specific piece of equipment — the job an asset register exists to do.

What we hold is the work: tasks that appear when they should, in the queue where people actually look, with a status that survives the shift. Buggies are the one asset class we carry that record for: each one has a service history and a ledger of meter readings, and a servicing interval in kilometres or days turns its row red when it is due. Across the rest of the property — chillers, air handlers, split units — the asset register that answers the other question, with parts inventory, depreciation and warranty tracking hung off it, is a full CMMS’s territory, and if the argument you need to win runs on those, buy one. For many resorts under a hundred keys, the queue is the gap that’s actually costing them; for a large property with an engineering department, it’s a component and not the system.

Where to start if you’re behind

Not with the full schedule. A comprehensive preventive program written in one sitting is the thing most likely to be abandoned by week three.

Start with the assets that generate the most reactive visits. You already know what they are without needing a report — the air conditioning units in a specific block, the pumps, the golf-cart chargers, whatever it is at your property. Put recurring work on those and nothing else. It’s a small enough list that it’ll actually run, and it targets exactly the reactive load that’s crowding out everything else.

Then add the compliance-driven items, which have external deadlines and therefore their own forcing function.

Then, only once those two are running for a season, widen it.

The failure mode to avoid is generating more recurring tasks than your team can absorb. A queue full of overdue preventive work teaches everyone to ignore overdue, and once that lesson is learned it applies to the reactive jobs too — which is a considerably worse position than the spreadsheet you started with.

If you want to see how recurring work, inspections, and turnover jobs sit in one queue, book a twenty-minute walkthrough and bring the asset that’s cost you the most callouts this year.