Reactive, Preventive and Predictive

Every maintenance programme is somewhere on a spectrum between fixing things when they break and servicing them on a schedule. Knowing where you sit, and where you should, is the first decision.

Reactive: fix it when it fails

The default everywhere. Something stops working, someone calls a contractor.

Cheapest per intervention and most expensive overall, for reasons that are easy to state and hard to see in a monthly invoice.

The costs are indirect: emergency call-out premiums, product loss when refrigeration fails, lost revenue during downtime, collateral damage when a failing component destroys a working one, and shortened asset life.

A failed compressor costs the repair, the spoiled inventory, the trade during closure, and often the whole unit if the failure was caused by a condenser nobody cleaned.

Where reactive is genuinely correct: low-cost items with no consequence when they fail, assets near end of life, and anything where the preventive work costs more than the failure.

Preventive: service on a schedule

Calendar or usage based. Quarterly refrigeration service, semi-annual hood cleaning, annual fire suppression inspection.

Costs more in visible spend and less in total.

What it actually buys: fewer emergency failures, longer asset life, compliance evidence, and — underrated — a technician looking at the equipment regularly, who notices things.

The failure mode is doing it badly. A preventive visit that consists of a technician arriving, signing a sheet and leaving is a recurring cost with no benefit. This happens more than operators realise, and it is why the task list on the work order matters more than the frequency.

Where it pays clearly: refrigeration, hoods and suppression, HVAC, fryers, ice machines, grease management. High consequence, predictable degradation, known service intervals.

Where it does not: small wares, items under warranty with manufacturer service, and anything where failure is cheap and immediate.

Predictive: monitor and intervene before failure

Condition-based rather than calendar-based. Temperature sensors on refrigeration, energy monitoring, vibration, run-time hours.

Real and increasingly affordable for refrigeration specifically, where wireless temperature monitoring has fallen to a price most operators can justify on food safety grounds alone.

What it genuinely catches: a cooler drifting warm over days before it fails, a door left open, a compressor running longer to hold temperature, a defrost cycle not completing.

What it does not: mechanical failures with no warning signal, anything without a sensor, and the large majority of non-refrigeration assets.

The honest position: predictive maintenance in restaurants means refrigeration monitoring plus, occasionally, energy anomaly detection. Vendors describing anything broader are selling ahead of what the sensors support.

The mix that works

For most multi-site operators:

Preventive for the high-consequence assets. Refrigeration, hoods and suppression, HVAC, fryers, ice machines, grease. Perhaps a dozen items per site.

Monitoring on refrigeration, because the failure is expensive and the signal is available.

Reactive for everything else, deliberately rather than by default.

Review annually based on work order history. An asset generating repeated reactive calls belongs on the preventive schedule or on the replacement list.

The argument to make internally

Preventive maintenance is a line item that increases. Reactive costs are scattered across emergency repairs, inventory write-offs and lost sales, and they do not appear as maintenance spend.

This makes preventive programmes politically fragile. They are the first cut when budgets tighten, and the savings are invisible because they consist of failures that did not happen.

The counter is measurement. Track total cost per asset including product loss and downtime, not maintenance spend alone. That is the only framing in which the comparison is honest, and it requires the work order history to be complete enough to support it.

Where operators typically sit

The common pattern is reactive everywhere except the three things a regulator requires: hood cleaning, fire suppression inspection, and sometimes grease trap service.

That is a compliance programme rather than a maintenance programme. The gap between it and a working one is usually a dozen scheduled tasks per site and a decision to keep them when the quarter is tight.

The twelve assets that justify a schedule

Rather than deciding asset by asset, most operators can start from a standard list and adjust.

Walk-in cooler. Walk-in freezer. Principal reach-in refrigeration. Ice machine. Fryers. Combi oven where present. Dishwasher. Hood system. Fire suppression. Rooftop HVAC. Make-up air unit. Grease interceptor.

Each meets the three criteria: expensive failure, predictable degradation, and preventive work that genuinely interrupts it.

Add water filtration where the supply is hard, because it protects four other assets at once.

Everything else is reactive by decision. Writing that down matters — a documented decision to run something to failure is a strategy, and the same behaviour undocumented is neglect.