You cannot buy an outcome you cannot specify

Outcome-based maintenance contracting is an appealing idea, and mostly a sound one. Instead of paying a contractor for the hours worked and the materials consumed, the owner specifies a service level and pays for the asset being kept at it. Responsibility for method moves to the party best placed to choose it.

The appeal is strongest for the organisations least equipped to use it. An owner with weak internal capability looks at outcome contracting and sees a way to buy competence. That is the reading that goes wrong.

Five things the owner has to know first

A service level cannot be specified, priced, or verified unless five questions already have answers, and they are the owner's questions, not the bidder's.

What the asset is, and where. What condition it is in, assessed rather than assumed. What intervention it needs, to what standard. When, at what interval or trigger. And what that costs, as rates applied to real quantities rather than allowances applied to estimates.

Answer those and an outcome contract is a genuinely better instrument. Leave them open and something predictable happens.

Unpriceable risk gets priced anyway

A bidder facing an undefined asset base has three options. Price the uncertainty, which makes the contract expensive. Price optimistically and argue later, which makes it adversarial. Or decline to bid, which narrows the field to the two firms willing to do one of the first two.

None of these is the contractor behaving badly. A contractor asked to guarantee an outcome across an estate nobody has catalogued is being asked to underwrite the owner's ignorance, and will charge for it.

There is a verification problem too. A service level regime depends on measuring performance against a baseline. Without a defensible record of what was there and in what state at contract award, disputes about whether a defect is pre-existing or a failure of maintenance have no factual anchor. They are settled by whoever argues better.

The funders say the same thing

This is not a minority view. The World Bank, which has done more than anyone to promote output and performance-based road contracts, is explicit that implementation carries prerequisites, among them a strong governance model and genuine client buy-in to asset management as well as to the contracting model itself.

Read that carefully. Asset management is listed as a precondition of the contract, not a benefit of it.

What an asset management system actually is

The term gets used loosely, often to mean a software package. In the standards it means something else entirely.

ISO 55001 sets out requirements for a management system: a set of interacting elements that establish asset management policy and objectives and the processes to achieve them. Underneath the policy sits a Strategic Asset Management Plan, which converts organisational objectives into asset management objectives and sets the approach for the plans below it. Asset management plans then say what will actually be done, to which assets, when.

A note on wording, because it causes confusion in tender documents. Older practice, and a good deal of current usage, says "asset management strategy". ISO 55000 and ISO 55002 accept that a strategic asset management plan may be called by other names, so nobody using the older term is wrong. But the standard's term is SAMP, and the 2024 revision made it a distinct requirement in its own right rather than something inferred. If a specification is being written against the standard, it is worth using the standard's vocabulary.

What changed in 2024, and why it matters here

ISO 55001 was revised in 2024, drawing on a decade of use across more than fifty countries. Four of the changes bear directly on outcome contracting.

The Strategic Asset Management Plan became its own requirement. Previously implied, now specified, including what it should contain and what it should bring about. An organisation cannot claim a system while skipping the document that links what it owns to what it is trying to achieve.

Externally provided processes and services were brought inside the system. What the 2014 edition called outsourcing is now framed as externally provided processes, products, technologies and services, and service providers are treated as participants in the asset management system rather than as parties outside it. An outcome contract is not a substitute for the system. It sits within it.

Data and information were separated and clarified. The revision distinguishes the two in an asset management context and treats configuration as a basic capability supporting decisions. Having files is not the same as having information you can decide with.

Knowledge became an explicit requirement. This one is new, and it is the most quietly significant. The standard now requires an organisation to determine the knowledge needed to operate its asset management system, put processes around it, and keep it current as requirements change.

The knowledge clause deserves its own paragraph

Anyone who has watched a capable improvement programme evaporate will recognise why that clause was added.

A review identifies deficiencies. A sponsor understands them and backs the work. The sponsor then moves to another portfolio, and within a year the understanding of what was wrong, where the information sat, and why the last attempt stalled has left the organisation entirely. The next programme starts from the beginning and pays full price for the same discoveries.

Treating knowledge as something to be deliberately retained rather than something that happens to reside in whoever is currently in post is a governance decision. The 2024 revision now asks for it directly.

The order that works

Outcome contracting is not the problem, and nothing here argues against it. The problem is sequence.

• Establish the system before the contract. Policy, a strategic plan that links assets to objectives, and asset management plans beneath it.
• Build the information base to a standard you can state, and grade its confidence honestly rather than presenting inference as survey.
• Then specify the service level, because now it can be described, costed and measured.
• Keep the capability afterwards. A contract does not maintain your knowledge of your own estate. Under the 2024 revision, maintaining that knowledge is a requirement of the system, not a nicety.

An owner that does this buys a genuinely better instrument, and can tell whether it is getting what it paid for. An owner that reverses the order signs a contract whose central terms it cannot verify, and finds out at the first dispute.