Every lawyer in Big Law learns very early to measure work in fractions of an hour. Six minutes, ten, or fifteen, depending on the firm. Open the matter, describe the task, record the time, move on. At some point, the act becomes so familiar that nobody really stops to look at it. The timesheet becomes part of the furniture of legal practice.
That familiarity is exactly what makes it dangerous.
A bit of background
The timesheet feels natural because lawyers have lived with it for generations, but it began as something much narrower: a way to understand cost. In 1913, in Boston, a Harvard-trained lawyer, Reginald Heber Smith, had to manage roughly two thousand cases a year for people who could not pay. He needed to know what the work required, what it cost, and how resources should be allocated. Like many managers of that era, he looked to the factory. Taylorism, time-and-motion studies, the stopwatch. Industrial efficiency entered the law office.
That origin still matters. The timesheet measures time spent, while legal work produces something more difficult to capture: mostly protection, leverage, confidence, risk reduction, and the ability to make a decision. Time may contribute to those outcomes, but it does not explain them. A board does not approve a transaction because 300 hours were recorded. A general counsel does not feel safer because a time entry was detailed. The value usually sits somewhere else, in the quality of the judgment and the trust behind it.
The number nobody audits
The least discussed problem with the timesheet is also the one managing partners should care about most: it is one of the softest numbers in professional services.
Some time is captured live. A lot of it is reconstructed later, at the end of the day, at the end of the week, sometimes later still. It is rounded, interpreted, compressed, expanded, and adapted to the conventions of the firm. It absorbs the interrupted call, the half-read email, the thought that occurred away from the desk, the problem solved between meetings, and the moment of judgment no clock ever captured.
Most lawyers are doing what the system asks them to do: turning continuous, fragmented, cognitive work into neat decimal blocks after the fact. The reason? Well, legal work rarely happens in clean units. It happens through interruptions, partial attention, calls, documents, client pressure, partner comments, judgment calls, and constant context switching. The timesheet asks lawyers to turn that mess into a number that looks precise. Some companies are selling software promising to automatize the process, but their usage is still in embryonic phase.
That said, firms use this data to answer questions that matter: which matters are profitable, which clients quietly destroy margin, where leverage works, where capacity is tight, which teams need support, which partners manage well, and which types of work should be pursued or avoided. These are board-level questions for any professional services business. Yet in many firms, they are answered using data reconstructed from memory, rounded into billing increments, interpreted inconsistently across teams, and later adjusted through write-downs or client pressure.
Clients know this too. That is why they ask for detailed time entries and granular narratives, then use the same detail to push back. Entries are queried. Block-billed time is rejected. Hours are written down weeks after they were recorded. Entire audit functions exist on the buyer’s side because everyone knows the timesheet is not considered hard measurement. The firm reconstructs the number; the client contests it. The invoice becomes a negotiation field.
What AI exposes
AI did not create the weakness in the timesheets, but it made the weakness harder to ignore.
For years, document automation, e-discovery platforms, contract-review tools, knowledge systems, and passive time-capture software have been eroding the edges of traditional legal work. Software already drafts standard clauses, populates data rooms, supports first-pass review, searches precedent banks, and reconstructs time that lawyers forgot to record. AI accelerates that shift and pushes it closer to the centre of legal production.
When a model drafts a contract in seconds, reviews hundreds of agreements overnight, or produces the first version of a research memo that used to take a junior lawyer two days, a large part of that work moves outside traditional time measurement. The work most exposed is often the work the timesheet measured best: volume work, repetitive work, research-heavy work, review work, process work. The kind of work that scaled neatly because it sat closest to the factory logic from which the timesheet originally came.
That should make every firm pause. The most measurable work is becoming the most automatable. If the cleanest hours are the first to be compressed, can we find better ways to understand what happened inside the matter?
In our experience, the answer is not to throw away time data. Time still matters. It helps explain effort, capacity, staffing, leverage, and cost. But it should be treated as an internal signal, not as the full truth. In an AI-driven firm, the timesheet should become part of a broader management system: one that captures phases, tasks, complexity, inputs, outputs, reuse, automation, margin, and client value with more discipline.
This is the practical work many firms have postponed. Better capture, better matter taxonomies, better phase coding, better separation between billing data and management data, and a better understanding of what their existing time records can actually tell them. Without that foundation, every conversation about profitability, pricing, capacity, and strategy remains weaker than it looks.
The firms that are moving already know which work is repeatable, which work is profitable, which teams are overloaded, which clients require a different service model, and where AI is genuinely changing the economics of delivery. They are stopping treating the timesheet as a source of truth and start treating it as what it really is: a useful but incomplete sensor.
Do you have issues redesigning the way legal work is measured, and build business intelligence that reflects how matters are really delivered? Contact us, and let’s fix the issue before your clients do it for you.