Why the Best Benchmarking Starts After the Numbers

After more than twenty years working in benchmarking,

I have come to a simple conclusion:

the most interesting part of any benchmark dataset is rarely the numbers themselves

It is the story they tell.

That may sound odd in a sector that often treats benchmarking as a technical exercise. Collect the data. Normalise the data. Compare the data. Produce the table. Identify whether the cost per square metre, meal subsidy, cleaning hour, helpdesk ticket, lifecycle allowance or PPM ratio sits above or below the chosen comparator set.

All of that matters. Poor data produces poor advice. But the real value of benchmarking is not found in proving that one number is higher or lower than another. It is found in understanding why.

That distinction matters more now than ever. FM leaders are operating in a market shaped by inflation, labour pressure, evolving workplace expectations, procurement reform, carbon commitments and ageing estates. Public sector buyers are under intense scrutiny over value for money, while private sector clients are asking more pointed questions about cost, service resilience and the credibility of supplier claims.

In that environment, a benchmark that simply says “you are expensive” is not insight. It is a prompt. The useful work starts when the dataset is used to uncover the operating story behind the result.

A cleaning benchmark, for example, may show that one site is materially above the market range. The lazy conclusion is that the contractor is overcharging. Sometimes that may be true. But the dataset may also reveal a more complicated picture: poor space utilisation, fragmented opening hours, excessive specification, duplicated daytime and periodic tasks, security constraints, or legacy assumptions that have not been challenged since the contract was first let.

The number points to the issue. The story explains the cause.

The same is true in workplace catering. A subsidy figure taken in isolation can create a misleading sense of underperformance. A site with lower footfall, extended service hours and high client expectations will tell a very different story from a dense headquarters building with predictable demand and strong capture rates. Benchmarking should not be used to punish complexity. It should be used to make complexity visible.

This is especially important in PFI and long-term FM arrangements, where the commercial history of the contract often matters as much as the current service cost. A benchmark dataset can reveal whether a price has drifted because of indexation, scope creep, risk transfer, service variation, lifecycle assumptions or a lack of competitive tension over time. In those cases, the question is not simply whether the charge is high. It is whether the charge is explainable, defensible and still aligned to the service being delivered.

That is why forensic benchmarking is different from basic comparison. It does not treat the dataset as an answer machine. It treats it as evidence.

This distinction is becoming more important as procurement becomes more data-led. The UK’s new public procurement regime is placing greater emphasis on transparency, supplier performance and the quality of commercial information. The early data from the new central digital platform is already being analysed for what it can tell us about behaviour, disclosure and market access. At the same time, public procurement remains vast in scale: House of Commons Library analysis puts UK public sector procurement spending at £434 billion in 2024/25.

In other words, the sector is not short of numbers. The risk is that it becomes short of judgement.

Benchmarking can help avoid that, but only if clients and advisers resist the temptation to use averages as blunt instruments. An average is not a standard. A lower quartile result is not automatically good. An upper quartile result is not automatically bad. Context changes everything.

A low-cost FM service may be genuinely efficient. It may also be under-resourced, fragile, dependent on unpaid goodwill, or storing up lifecycle and compliance risk. A higher-cost model may reflect waste. It may also reflect a conscious decision to protect service quality, maintain resilience, support better employment standards, or meet demanding operational constraints.

The benchmark does not remove the need for commercial judgement. It sharpens it.

For FM leaders, the practical implication is clear:

do not commission benchmarking just to win an argument. Commission it to understand the system. The best datasets help leaders ask better questions:

• Why does this service cost what it costs?
• Which variables are genuinely controllable?
• Where are we paying for historic decisions rather than current need?
• What would have to change operationally before the benchmark could move?
• Are we comparing like with like, or merely pretending to?

For suppliers, there is a lesson too. A credible benchmark should not be feared if the service is well run and the cost base can be explained. In fact, good benchmarking can protect suppliers from crude price pressure by showing the operational reasons behind cost. But it also exposes weak narratives.

“That is just the market rate” is not good enough when the data suggests otherwise.

For buyers, the value lies in using benchmarking as a decision tool rather than a procurement ritual. It should inform specification, negotiation, market testing, contract management and future operating models. Done well, it can reveal whether the problem is price, scope, productivity, demand, governance or behaviour.

Every benchmark dataset tells a story. The skill is knowing how to read it.

And after twenty years in this field, I would argue that the organisations that get the most value from benchmarking are not the ones that chase the lowest number. They are the ones that use the evidence to understand what is really happening inside their estate, their contracts and their services.

Because in FM, the number is rarely the whole story.

It is usually the beginning of a much better conversation.