AMC versus reactive maintenance: the real comparison

Fix-on-failure looks cheaper because you only pay when something breaks. The honest comparison counts four things, not one: the invoice, the downtime, the shortened asset life and the risk you are carrying.

By Dr. Dirajen Pullay Marday, DBA, MBA, BEng, CMgr MCMI, CIWFM, MIoD - Managing Director, BFC Ltd. Published 14 August 2026.

What each model actually is

Reactive maintenance, sometimes called run-to-failure or fix-on-failure, means no scheduled attendance. You call a contractor when something stops working, pay for that visit, and wait. An annual maintenance contract (AMC) means an agreed asset list is attended on an agreed frequency for a fixed monthly or annual fee, with reactive call-outs handled at contract rates and contractual response times.

The two are not opposites. Every building needs a reactive capability, because things fail without warning. The question is whether you also buy the planned layer that reduces how often that happens.

The four costs you have to count

1. The invoice

A reactive invoice looks smaller in isolation. It is also more expensive per hour of work, for reasons that have nothing to do with the contractor being opportunistic. An unplanned attendance carries mobilisation for one job rather than a batched route, frequently attracts an out-of-hours or weekend premium, and often involves diagnostic time that a planned inspection would have already done. Parts are bought at short notice instead of being scheduled, and if the part is not on the island the job stalls.

2. The downtime

This is the cost owners consistently leave out, and it is usually the largest. A BPO floor without cooling sends people home. A retail unit with a failed distribution board closes. A serviced office with a burst riser loses tenant confidence, which is a slower and more expensive kind of damage. Downtime cost is specific to your operation, which is why we ask about it during the survey rather than assuming it.

3. Shortened asset life

Plant that runs dirty, unbalanced or under-lubricated does not fail politely at the end of its design life. In a tropical marine climate, a condenser coil left uncleaned works harder, draws more current and runs hotter, and the compressor pays for it. Replacing a compressor several years early is a capital cost that no reactive invoice ever shows you, because it arrives as a separate line in a different budget year.

4. The risk you are carrying

Unmaintained electrical distribution, an RCD that has never been tested, a fire pump that has never been exercised, a booster set with no pressure record - these are not maintenance items, they are liabilities. If something happens, the absence of a maintenance record is the first thing anyone asks for. A planned contract produces that record as a by-product.

A worked comparison

Consider identical assets in two buildings: fifteen split units, three distribution boards, a domestic booster set and eight washroom cores.

Building A, reactive only. Nothing is attended until it fails. Over a year you can reasonably expect several unplanned events: cooling capacity drops through the hot season and two units eventually stop, a blocked drain floods a washroom core after hours, and a booster set coupling starts leaking. Each of those is a separate mobilisation, at least one is out of hours, one causes a partial closure, and the two air conditioning units end up needing gas, coil cleaning and one compressor between them. The compressor was avoidable.

Building B, on an AMC. Cooling is deep-serviced twice a year with filter checks monthly through the hot season, drains are rodded on the planned visit, the booster set is checked and its pressures logged, and boards are inspected and RCDs tested annually. The failures that remain are genuine random failures, they are fewer, and they are attended within a contracted response time by a technician who already knows the site and the asset history. There is no diagnostic learning curve, and the parts likely to be needed are known in advance.

The monthly fee for Building B is a visible, budgetable number. The annual total for Building A is invisible until it has already been spent, and it does not include the compressor or the lost trading hours. In our experience on buildings of any real size the planned route is cheaper across a full year, and the gap widens in year two and three because asset life stops being eroded.

When reactive genuinely is the right call

Planned maintenance is not automatically correct for everything. Reactive is defensible for low-criticality, low-consequence assets: a single lamp circuit in a store room, a tap in a rarely used kitchenette, cosmetic fabric items. It is also reasonable for a building you are about to vacate or demolish, and for assets under an active manufacturer warranty where a third-party intervention would void cover. The mistake is applying the reactive logic to the assets whose failure closes your operation.

The practical answer for most buildings is a tiered asset list. Put the assets whose failure stops the business on a planned schedule with a fast contractual response, and leave the rest to be attended on the next routine visit. That is the cheapest genuinely safe structure, and it is how we build an annual maintenance contract.

How to move from reactive to planned without a shock

Start with a survey and an asset register rather than a contract. You then know what you own, what condition it is in and what the catch-up items are. Separate those catch-up works from the ongoing schedule and price them individually, so the first year of the contract is not quietly funding a repair backlog. Agree priority tiers and response times against real operational consequences. Then set the reporting: what you receive, how often, and with what photographic evidence.