Practice

The Real Cost of Taking Over a Matter

What you look for when taking over a matter is not a document. It is decisions, and the context in which they were made. No model holds it.

Let us put a figure on a pain everyone knows without measuring. A lawyer takes over a matter left by a departed colleague. How long before they are truly operational on it? Rarely less than two days, often more. Two days spent opening the shared directory, reading the pieces in the order they were filed rather than produced, reconstructing why such a position was taken, calling the colleague, guessing, sometimes starting over.

Let us do the math, because it is in doing it that one measures the size of the item. On a team of twenty people, with turnover, absences, transitions, this cost counts in weeks per year, perhaps in months. Each departure triggers a takeover, each takeover costs its two days, and those two days are paid at the rate of a qualified professional who, during that time, produces nothing billable. Multiply by the number of takeovers in a year, and you get a considerable cost item, entirely invisible.

For this cost appears on no invoice. It is the most expensive of costs: the one you do not see, because it is diffuse, spread over dozens of partial takeovers, and carried by no accounting line. A firm that could quantify what it loses each year in matter takeovers would discover a figure that would justify, on its own, a serious investment in the only thing able to reduce it: a memory that survives people’s departure.

What is lost in a matter takeover is not time. It is a cost no one quantifies.

What you think you take over, and what you really do

To understand where this cost comes from, one must confront two accounts. What a firm thinks it transmits by transmitting a matter is the documents: they are there, archived, retrievable, complete. On paper, the takeover should be immediate, since everything is available. It is the first account, the reassuring one, and it is accurate: no document is missing.

What one really takes over is something else entirely. For a matter is not made only of documents; it carries a decisional layer, made of trade-offs, motives, intentions, that lives nowhere in usable form. It is in scattered emails, handwritten notes, a partner’s memory, a Tuesday-evening conversation no one transcribed. This second layer, the only one that matters for taking over, is precisely the one that does not transmit with the shared directory. The second account, the one that decides the real cost, is almost empty.

The gap between the two accounts is the whole problem. One has all the documents and understands nothing, because what gave the documents meaning was never captured. Why was this clause accepted when it seems unfavorable? Because it was traded for another, earlier, in a discussion of which no written trace remains. The document carries the decision; it does not carry its reason. And it is the reason, not the decision, that must be reconstructed to take over, which explains why the takeover costs two days where merely reading the pieces would take two hours.

It is objected that modern tools, with augmented retrieval, find any passage in any document in a second. That is true, and it is unrelated to the problem. Retrieval operates on the first account: it finds documents and passages, the very ones that were never missing. The decisional layer, by definition, is not written in the documents; it is the reason the documents say what they say, and that reason, if it was not recorded, is found nowhere.

No retrieval, however capable, will find why a clause was accepted in this matter when it was refused in another, if that reason was never recorded. One can find both clauses; one will not find the trade-off that separates them. It is the difference between a library and a witness: a library gives you all the documents, a witness tells you what happened. Retrieval is a perfect library, and what a firm looks for when taking over a matter is not a library, it is a witness.

Retrieval finds what is written. The cost of takeover comes from what never was.

One can refine the calculation of the gap, because it explains why current tools reduce this cost almost not at all. A firm that equips itself with an excellent document search engine improves the first account, the one that had no problem, and does not touch the second, the one that costs two days. It pays to accelerate a reading that was not the bottleneck, and leaves intact the real bottleneck, the reconstruction of meaning. This is why so many firms find, after investing in augmented retrieval, that their matter takeovers still cost as much: they optimized the account that was already fine.

A memory that depends on no one

If the cost comes from the decisional layer being lost, then the only way to remove it is to retain that layer at the moment it forms, and not to hope to reconstruct it after the fact. It is this layer that MAX carries: a persistent legal memory that retains decisions, their context, their sequence, the positions taken, the trade-offs made, and that makes this memory active in each new interaction with the matter. A memory that depends neither on a colleague’s availability, nor on the quality of a filing, nor on the luck that an email was kept.

The stake far exceeds the comfort of takeover, and this is where the invisible cost reveals itself to be a strategic stake. A firm’s value lies not in its documents, but in its accumulated decisions: it is this capital that allows not making the same mistake twice, holding a coherent position from one client to the next, capitalizing on experience rather than letting it leave with those who go. Each costly takeover is the symptom of a capital leaking, for lack of a layer to retain it.

An AI layer that does not hold this accumulation holds nothing durable, whatever the quality of its generation. It produces fast, and forgets just as fast, leaving the firm to repay, at each takeover, the cost it thought it had eliminated by equipping itself. Conversely, a memory that retains turns each departure from a loss into a mere change of hands: the matter continues, because what made it hold never left with the person.

This shift of the gaze, from the document to the decision, is what distinguishes a memory layer from mere improved archiving. To archive better is to file the first account more finely; to retain the decision is to constitute the second as it happens. The first approach makes documents easier to find; the second makes the matter genuinely resumable, because it keeps not what was written, but what was decided and why. It is the only one that tackles the cost where it lies.

A firm’s value is not in its documents. It is in its decisions, and that is what takeover makes it pay for.