The FM Cost Ratio Trap: Why the Ratio Is Not the Model

The first question is not whether the ratio is 80/20, 85/15 or 75/25. The first question is: what is the denominator? Until that is answered, the ratio is not a benchmark. It is just a number with a decimal point in a convincing place.

This article is written for the finance, commercial and operations people who have to price these models, defend them, and live with them once the contract is running. The argument is not that ratios are bad. The argument is simpler: the ratio is not the model. A ratio only tells you something once three things are clear, the denominator behind it, the definitions inside it, and the operational reality underneath it.

The ratio is not the model

FM commercial discussions almost always start with a ratio. Someone says the planned to reactive split should be 80/20, or that reactive should sit at 15 per cent of planned, and the conversation proceeds as though the ratio explains the cost model.

It does not. A ratio only shows the mix after the costs have already been classified. It tells you nothing about what sits inside each line, whether the classification is consistent across bidders, or whether the number is built from spend, jobs, labour hours or asset risk. Two contractors can both quote 80/20 and mean entirely different things.

The danger is not that the ratio is wrong. The danger is that the ratio is treated as evidence. In a bid model it becomes a negotiation shortcut. In a live contract it becomes an operational problem. Either way, nobody has yet asked what the number is actually made of.

Percentage of what?

This is where most ratio arguments quietly fall apart, because two different measures get used interchangeably.

Take a simple maintenance stack. Planned maintenance £100,000, reactive maintenance £20,000, consumables £5,000, total £125,000.

Planned as a share of the whole stack is £100k / £125k, which is 80 per cent. But planned as a share of planned plus reactive only, leaving consumables out, is £100k / £120k, which is 83.3 per cent. That second measure is the planned to reactive ratio: 83.3 per cent planned, 16.7 per cent reactive.

There are two different percentages here: planned as a share of the total cost stack, and planned as a share of planned plus reactive only. They are not the same measure. Same estate, same pounds, two different numbers, because the denominator is different.

Now remove the £5,000 of consumables. Watch what happens.

Planned as a share of the total moves from 80 per cent to 83 per cent, because the total shrank from £125,000 to £120,000 while planned stayed at £100,000. Not a pound of planned work was added. The share rose purely because the pie got smaller.

The planned to reactive ratio, meanwhile, has not moved at all. It was 100 to 20 before and it is 100 to 20 after. Removing consumables never touched it.

So the same action, taking out £5,000, makes one measure jump and leaves the other flat. If you do not say which denominator you are using, you can claim the estate got more mature or claim nothing changed, and both statements can be pulled from the same numbers. That is the trap in one worked example: percentage movement can be pure arithmetic, and unless the denominator is named, the ratio is not telling you what you think it is.

There is not one FM ratio. There are several

Part of the confusion is that people say the ratio as though there is only one. There are several, and they are not the same thing. Planned to reactive compares planned maintenance against reactive maintenance, and the confusion is whether consumables sit inside either side or outside both. Labour to materials compares engineering labour against parts and materials, and the confusion is where the contract draws the boundary. Consumables treatment covers filters, belts, lamps, fixings, lubricants and spares, which may sit inside planned, inside reactive, or as a separate line entirely. Life cycle funding covers planned renewal of major assets and components, which may be separately funded or quietly leaning on the maintenance budget.

80/20 might mean planned to reactive. 75/25 might mean labour to materials. Materials and consumables may sit inside planned, inside reactive, or as a separate line, and the ratio changes depending on where the contract puts them. Life cycle may be separately capitalised or hidden inside maintenance. Unless the denominator and the definitions are pinned down, the ratio is not a benchmark. It is just a number.

Standards help here, but only so far. SFG20 structures the planned task regime. BS 8544 and the wider life-cycle costing standards help structure how the maintenance plan and renewal plan sit together. These frameworks help structure the thinking. They do not hand you a universal planned to reactive percentage, and any bid that claims a ratio is the standard is usually reaching for authority the standard does not provide.

Same reduction, different operational story

Because the ratio is not the model, two changes that look identical on paper can mean very different things on the ground.

Suppose you need to take £5,000 out. You could drop the consumables line, or you could trim reactive from £20,000 to £15,000. Both reduce the model by the same £5,000. Both land the total at £120,000. On cost, they are indistinguishable.

Whether either is a saving depends on whether the work removed was genuinely avoidable, and only the person who understands the cost lines underneath can tell which is which.

New-build estates are not mature estates

The same mistake appears in new-build models, dressed up as prudence.

A new estate should not automatically carry mature-estate reactive demand from Year 1. But any early-life reduction has to be evidenced through warranty cover, the defects liability period, planned maintenance from commencement and life cycle funding. Where that position supports it, the reduction is defensible. Where it does not, it is not modelling. It is optimism with a phasing curve drawn over it.

Reliability engineering describes this shape through the early-life, useful-life and wear-out pattern, and a defensible model may phase early-life reactive demand accordingly. The test is not whether the curve looks sensible. The test is whether the position underneath it supports the number.

Life cycle is not a reactive budget

Life cycle is where the whole model either holds together or quietly falls apart.

Life cycle is not reactive maintenance. It is planned renewal. If life cycle replacement is properly funded, it should reduce repeated late-life failures, because assets are renewed before they reach the point of chronic breakdown. If it is underfunded, the problem does not disappear. It reappears as reactive maintenance, later, as call-outs and emergency repairs at a worse unit rate. But the fact that underfunded life cycle pushes up reactive cost does not make life cycle part of the reactive budget. They remain two different things.

So a model that trims life cycle to improve the headline can look cheaper and more controlled on paper while storing up reactive cost that lands in a future year. The ratio improves. The estate does not.

The honest test

This is the point a spreadsheet cannot answer for you.

On the model it is easy to shave five per cent here and there and watch the numbers tidy up. Operationally, that five per cent is real work and real materials on the ground. If the demand was genuinely slack, you have made a clean saving. If it was not, you have not removed the need. You have simply stopped funding it.

Reactive demand you underfund does not vanish. It comes back as emergency call-outs, usually at worse rates, with SLA pressure and faster asset deterioration behind it. Consumables you strip out show up as jobs that cannot be closed for want of a part. The model can make the reporting honest. It cannot tell you which spend is truly avoidable and which you are simply choosing not to see until later. That answer has to come from the operational picture.

So the key test is not whether the percentage looks better. The key test is whether the spend is genuinely avoidable.

The point in one line

FM cost ratios are useful right up to the moment they are treated as the model rather than a summary of it. The ratio is not the model. Name the denominator, define the lines, test the operational reality, and only then does the percentage mean anything at all.