Unique requests: what falls outside a standard scope, and its price

Every provider's service list describes the requests they have already solved. The interesting ones are the remainder: a permission that does not normally exist, an object that is not for sale, an arrangement nobody has assembled before.

These are priced separately from everything else because they cannot be estimated from experience. What is being quoted is the search, not the outcome.

The three are routinely confused. A feasible request is one where the obstacle is money or effort. An expensive one is feasible with a large number attached. An impossible one has an obstacle that is not commercial at all: a law, a safety rule, a person who simply will not agree.

A provider worth retaining says which of the three applies at the outset.

The failure mode in this category is a firm accepting an impossible request and spending three months proving it.

Usually as time plus a success fee, because the alternative, a fixed price for an uncertain outcome, is unattractive to both sides. Expect the time component to be payable regardless of result.

Where a provider offers a purely contingent fee on something genuinely difficult, look closely. Either they know it is easier than it appears, or the fee reflects the risk of doing the work for nothing.

Flexibility on anything that is not the point. A request fixed on a specific date, a specific venue and a specific person is three constraints; relaxing any one of them frequently converts impossible into merely difficult.

And time. Almost every unusual request that succeeds does so because it was raised early enough for someone to ask a favour rather than to demand one.

A provider cannot price what it cannot picture, so the first job is to convert an ambition into a specification. Three answers usually do it: what would count as success, what is genuinely fixed about it, and what the deadline is for finding out it will not happen. The third is the one people skip, and it is the one that limits the cost.

The answers also decide who should do the work. A request that turns on a relationship belongs with a firm that already holds one in that sector; a request that turns on research and persistence belongs with whoever will do the hours cheaply. Paying a relationship premium for a research problem is the most common way this category becomes expensive for no reason.

Providers vary enormously in how they respond to being asked for a specification. A firm that pushes back with its own questions is doing the work; a firm that agrees immediately and quotes a round number has not begun.

The honest reasons are worth recognising, because they are also the useful ones. A request may be declined because it would damage a relationship the firm depends on, because the obstacle is legal rather than commercial, or because the effort cannot be estimated closely enough to quote against. All three are better outcomes than acceptance.

A refusal that arrives with a reason and an alternative is a sign of a firm working inside real constraints. A refusal with no explanation usually means the request touched something the firm would rather not discuss, which is itself information.

The failure mode to watch for is neither acceptance nor refusal but drift: a request that is never formally declined and never progresses. Setting the abandon date at the outset is what prevents it, and a provider who resists setting one is telling you something about how the engagement will run.

Very few requests are unusual all the way through. Most are an ordinary arrangement with one difficult element attached, and separating the two changes the price: the ordinary part is quoted from experience and the difficult part is quoted as a search. A provider who prices the whole thing as exceptional is charging the search rate for work that has a standard cost.

That separation is also what makes an unusual request comparable between firms. Two quotes for an undefined ambition cannot be compared; two quotes that both break out the routine coordination, the search, and the success fee can be.