News in Practice

How many subscriptions does it take to run a single matter?

Ask your finance manager, then ask whoever actually runs the matters. The two answers never coincide, and the gap between them is the subject.

For partners, general counsel and business managers who own a tools budget. How to take stock of what is genuinely used, and what the sum of subscriptions does not tell you.

The question looks like an accounting one and it is not. Your finance manager will answer by listing budget lines: a research platform, a document management system, an e-signature service, a practice management tool, an email platform, perhaps a contract analysis solution. Whoever runs the matters will answer something else, and the gap between the two answers is the only figure that counts.

Taking stock through the budget

It is easy to obtain and it is misleading, for three reasons. Start with the most obvious.

It counts licences, not use. A licence bought for fifteen people and used by four costs the price of fifteen and delivers the service of four. That gap is common, it appears nowhere, and it is the first line of savings available in most organisations.

It ignores whatever has no line. Undeclared individual use, free tools, subscriptions expensed by a partner who did not want to wait. That category is systematically underestimated and is sometimes the most used of all.

It says nothing about overlap. Two tools bought two years apart, for two distinct needs, may have converged functionally without anyone noticing. The budget counts them twice because they are two lines; it cannot see that they partly do the same thing.

It also ignores commitment terms. Two lines of identical value do not mean the same thing if one is cancellable monthly and the other committed for three years. That information never appears in an expenditure table and it governs every arbitration.

Taking stock through the matter

Take a transaction handled last month and follow what the person running it actually opened.

The result is generally surprising in its number, and more surprising still in its distribution. On a mid-sized deal one easily counts six to eight distinct environments: document management for the exhibits, email for correspondence with the other side, an external data room provided by the seller, a word processor with its versions, a research platform opened three times, a tracking spreadsheet built by hand, and the practice management system for time recording.

Add what is not a tool and functions as one: the shared tracking file, the precedents folder on the server, the personal notebook where the client's positions are recorded. Those objects are part of the stack, they carry part of the matter's information, and they appear in no inventory.

None of these environments knows about the others. That, rather than their number, is what produces the real cost.

The budget counts lines. The matter counts crossings from one environment to another, and that is where the money goes.

How to carry out the exercise

The method that yields a usable result consists of three questions put to three different people, and the gap between their answers is more instructive than any one of them alone.

To the business manager: which lines appear in the budget, for how many licences, and since when. That answer is available within the hour.

To a practitioner, on a specific matter: what did you open last week. That answer requires reconstruction and is incomplete by nature, which does not matter: what they forget to mention is what they use most automatically.

To the team, collectively and without consequence: what do you use that does not appear in the budget. That is the hardest question to have answered honestly, and the only one that reveals the third category. It presupposes being asked in a way that requires nobody to justify themselves.

Comparing the three answers is generally enough to open the discussion: what finance believes it is funding, what is actually opened, and what nobody had declared never coincide.

The three items with no budget line

These are not hidden costs in the sense of concealed spending: they are workloads that nobody has ever traced back to the stack that produces them.

Transfer. Each crossing from one environment to another requires selecting, exporting, re-establishing context. The action takes anywhere from a few seconds to several tens of minutes depending on what is being moved, it repeats, and it appears on no invoice.

Version checking. As soon as a document exists in two places, someone has to establish which one governs. That work is invisible as long as it is done well, and very expensive on the rare occasions it is not.

Mental mapping. Knowing where things are is an acquired skill, it takes months to build, it does not transfer, and it disappears with the person. That item explains why a reinforcement takes so long to become useful on a live matter.

A fourth item deserves adding, less obvious: the coordination load. Once a stack reaches a certain size, someone in the organisation becomes the de facto owner of each tool without that role appearing in any job description. They answer questions, train joiners, deal with the vendor. Across five tools, that amounts to a part-time role spread over three or four people.

These items share a property: they grow faster than the number of tools. Two environments create one boundary, three create three, four create six. That is why a stack that remained manageable at three components ceases to be so at five, with no event marking the transition.

What the exercise almost always reveals

Four findings recur with a regularity worth flagging.

The first is a gap of one to three between licences paid for and users active on at least one tool in the stack. It is no sign of negligence: it results from departures, changes of scope and automatic renewals nobody has reopened.

The second is the existence of at least one tool whose purpose nobody can any longer state. It sits in the budget, it renews, and the person who requested it has left.

The third is an undeclared individual use, often free, generally the most used of all. It costs nothing and it is the real governance question.

The fourth, and the most interesting, is that the practitioner spontaneously mentions arrangements that are not tools: a shared file, a spreadsheet, a precedents folder. Those objects are part of the stack and never figure in it.

What the exercise allows you to decide

Three decisions, and none of them is to consolidate everything.

The first is to check licence scope, which requires only a count and a comparison, and which produces an immediate result.

Remove what is not used. That is the simplest gain and it is purely budgetary: it changes nothing about the work. It is also the only one nobody argues about.

Identify overlaps and choose. Two tools that partly do the same thing force every user to decide which to use, which produces divergent practices within the same team.

These first two decisions are purely budgetary and can be taken without changing anyone's work, which makes them easy to obtain from a partnership.

And distinguish, among what remains, what must stay specialised. An exhaustive research platform, a citator, a subscription commentary service, an alerting system answer needs that a general-purpose layer does not cover and is not meant to cover. A firm that removes them to simplify its stack discovers the error on the first matter where exhaustiveness mattered.

A fourth decision sometimes arises and deserves naming: to change nothing, knowingly. An organisation whose stocktake shows a coherent, used and proportionate stack has no reason to act, and the exercise will have served to establish that rather than assume it.

The right conclusion of a stocktake is therefore almost never to reduce everything to a single tool. It is to know what each line actually buys, which is the precondition of any decision, including the decision to leave things as they are.