Buying by the hour

The modern proposal does not sell a thing. It sells the assurance that a thing will work when it is needed. That is a genuine improvement on the old model, and it moves the difficult questions somewhere most buyers are not looking.

What changed

The old model transferred an asset. A state bought vehicles, aircraft or systems and inherited everything that followed: crews, maintenance, spares, obsolescence and disposal. The capital cost was visible and the through-life cost was not.

The new model transfers a risk. The provider carries availability, maintenance and crewing, and charges for the outcome. The state buys serviceable hours rather than hardware.

Neither is better in the abstract. Availability contracting is cheaper to start, cheaper to stop and leaves nothing behind. Ownership is dearer, slower and durable. The right answer turns on whether the requirement is enduring or episodic, which is a question about the threat rather than about the contract.

But only one of the two is usually costed properly. Ownership proposals are scrutinised line by line. Availability proposals are compared on headline rate, and the terms that decide the real cost sit in the variables.

Three charges, and only one is fixed

Almost every availability proposal decomposes this way, whatever the document calls it.

• Mobilisation and recovery. Deploying into theatre, importing, setting up and recovering at the end. Real money, frequently omitted from comparisons, and the reason a short contract costs far more per month than a long one. Always ask for it separately.
• The core availability charge. Payable for having the capability serviceable and crewed against a forecast block of hours or days. This is the number quoted, and it falls very little when usage falls, because most of what it buys is readiness rather than consumption.
• Usage beyond the block. The line that converts a predictable budget into an unpredictable one. Model it against a realistic tempo rather than an optimistic one.

Utilisation is the main rate driver. A forecast block set too low looks cheap at signature and is expensive by month four.

Tiering, and where dependency forms

Modular entry is common and reasonable: a rotational presence, then a permanent one, then an integrated package that adds analysis, additional collection and command and control.

It is worth seeing the mechanism clearly. Moving from rotational to permanent roughly doubles the personnel bill to remove coverage gaps, which is honest and often correct. The third step is the one to watch, because the analytical and command layer is precisely the part a host institution most needs to own, and the part most easily left with the provider indefinitely.

What "all-in" should mean, in writing

A proposal described as a complete package is only complete against a list.

• Personnel in full, with named ratios rather than a headcount, including the rotation behind them.
• Maintenance, engineering and spares, including deep maintenance and what happens when a part requires an export licence.
• Life support: accommodation, feeding, welfare, medical cover and force protection. If the host is carrying any of it, that is a cost transferred rather than avoided.
• Insurance, with the exclusions named. An insurance limit is not a liability cap, and the two are routinely confused in negotiation.
• Training and transfer, with a schedule. Without one the clause means nothing.

The clauses that matter more than the price

Where contracted personnel operate alongside state forces, the governance terms carry the risk that the commercial terms do not. Three are non-negotiable.

Tasking and control. Every task flows from, and is released by, the designated national authority, with no autonomous action by the provider. Stated in the contract rather than in the covering letter, with the post named and a deputy for when it is vacant.

Conduct. Engagement only on positive identification, with proportionality, precaution and abort discipline written into planning and training rather than asserted in a slide.

Record. Every mission logged: who authorised it, under what rules, what was expended, what resulted. An audit trail reconstructed after the first contested incident is worth very little to anyone.

International frameworks exist here. The Montreux Document sets out what states should require of providers, and the International Code of Conduct sets out how providers should behave. Both are a floor rather than a credential.

What is left when it ends

The test of any contracted capability is what remains after the provider has gone home, and most contracts are not designed to pass it.

• Availability of a platform leaves nothing behind. That is not a criticism, but it should be accounted for as rental rather than as capability.
• Trained host personnel remain, if the training was scheduled, examined and resourced. Name the individuals and the standard in the contract, with a withhold against the milestone.
• A tasking and authorisation process is the most valuable residue and the cheapest to secure. Insist it is run by host staff from the outset, with the provider advising rather than operating it.
• Doctrine and documentation remain only if specified as contract outputs, owned by the client and publishable without reference to the provider.

Normalising bids before comparing them

Proposals are rarely comparable as submitted, and most of the difference in headline price is structure rather than value. Restate every bid over the same period including mobilisation and recovery. Price them all against one common usage forecast, then again at fifty per cent above it. List the exclusions side by side, because that is where the difference lives. Convert host nation support into money and add it to each bid. Score the assurance terms separately from the commercial ones, or a cheap bid with weak governance wins on arithmetic.

And cost the null case alongside them, with its consequence stated. Doing nothing is sometimes right, and it is always the benchmark the others have to beat.

Our interest

We help governments write requirements and evaluate bids against them, which is work that exists because these proposals are hard to compare. An article explaining why they are hard to compare is therefore an argument for our own service. Discount it accordingly.

We are not a supplier of operational capability, and we will accept exclusion from any work arising where a client requires it. The advice we give most often in this area is to write the requirement before going to market, which costs nothing and stops you buying somebody else’s product.

Read next · Regional analysis

Haiti and the neighbours

A state collapse exports three separate things on three separate timelines. Most regional responses are built for only one of them.