Cost per Matter as a Buying Standard in IP Operations

An intellectual property operation can be judged from two ends. From the top it looks like cycle time: how long a matter takes from instruction to filing. From the floor it looks like a queue of things waiting for someone to decide something. Neither view produces a purchasing decision, because neither attaches a number to the work.

Cost per matter does. It is the least fashionable measurement available to an IP department and the only one that survives contact with a finance committee.

What the measurement actually is

Cost per matter is the fully loaded cost of moving one unit of work through the operation, from the moment an instruction arrives to the moment the docket confirms the filing. It includes internal time at its real hourly cost, outside counsel and agent fees, official fees, and the share of platform licences attributable to that matter type.

Three of those four are already known to any finance function. The fourth, internal time, is the one almost nobody has, and it is the one that decides whether a purchase pays for itself.

A department that cannot state its cost per matter is not measuring its operation. It is measuring its invoices. The distinction matters because invoices capture only what leaves the building. The largest cost in most IP operations is the time of people who are already on payroll, and payroll does not itemise by matter.

Why the number is worth the trouble

Consider a platform quoted at fifty thousand a year, promising to cut drafting time by thirty per cent. Without a cost per matter, the business case is a percentage applied to an unknown, which is to say a guess dressed as arithmetic.

With a cost per matter, the same case becomes checkable. If the operation handles four hundred matters a year at three thousand two hundred each, and drafting is a quarter of that, then drafting costs three hundred and twenty thousand. A thirty per cent improvement, if it materialises, is ninety-six thousand against a fifty thousand licence and whatever the implementation costs. That is a decision a committee can take.

The same arithmetic exposes the purchases that should not happen. A tool that improves a step accounting for four per cent of the cost cannot repay a six-figure implementation, however impressive the demonstration.

How to build it without a project

The instinct is to instrument everything and wait a year. That is not necessary, and the wait is itself a cost.

Pick one matter type that carries volume. A first filing in a core jurisdiction is usually the right choice, because the path is repeated often enough for a pattern to be visible.

Take twenty completed matters. Twenty is enough to see a distribution, and small enough to reconstruct by hand from existing timestamps.

For each, record the six transitions the work passed through and who held it at each. Attach an hourly cost to each holder, loaded rather than salary alone. Add the external costs already known: agent fees, official fees, translation.

Then divide, and look at the spread rather than the average, because the spread is where the money is. The exercise takes a week of part-time attention. It does not require a new system, and it produces a number that did not exist before.

Read the variance, not the average

The average cost per matter is the figure that goes in the board pack. The variance is the figure that changes how the department is run.

In most operations the distribution is not a bell. It is a body of matters that behave, and a tail of matters that consume several times the median. The tail is rarely explained by technical difficulty. It is explained by an incomplete transfer early in the journey, a client whose approvals arrive unpredictably, or a formalities pass that had to be reconstructed.

Once the tail is visible, it can be attacked directly, and the attack is usually procedural rather than technological. That conclusion is unwelcome to anyone hoping a purchase will fix the problem, and it is the most valuable thing the measurement produces.

What it does to a vendor conversation

A department that arrives at a demonstration with a cost per matter, a distribution, and a named transition where the queue sits, is having a different conversation from the one the vendor prepared for. The question stops being what the platform can do and becomes what it does to that specific number.

Good vendors welcome this. They have built something that improves a definable step, and a buyer who can name the step is a buyer who will succeed with the product and renew. Vendors selling a general promise find the question uncomfortable, which is itself a useful signal.

It also changes the contract. A capability that matters to the business case can be written in as a dated commitment measured against the baseline. Without a baseline there is nothing to measure against, and the commitment cannot be drafted.

The standing version

A cost per matter reconstructed by hand every eighteen months is better than nothing and worse than it needs to be. Where the work is orchestrated across the platforms a team already runs, every transition carries a timestamp and an owner by construction, and the number is a property of the operation rather than a study.

That is the difference between knowing what your operation cost last year and knowing what it costs this week. Only the second one lets you judge a purchase while the purchase is still reversible.