Count the tabs open on your screen right now

Here is a small experiment we have run with about forty lawyers this year. Mid afternoon, mid matter, we ask them to count what is open on their screen.

The answer is almost never fewer than eight. A research database in one tab. The MCA portal in another, probably logged out. A data room. Word, with the agreement. Outlook, where the client's last three questions are sitting. A shared drive folder. WhatsApp Web, where the actual decision was made. And somewhere in the pile, an AI tool that someone in the firm bought last year.

Every one of those is a competent piece of software. Together they produce a working day where a lawyer spends a startling proportion of their time being a router: moving a document from one place to another, re-explaining context to a system that has no memory of the last thing they asked, and holding in their head the connections that none of the tools can see.

We think that last part is the interesting one, and it is almost never discussed.

The lawyer is the integration layer. Not metaphorically. Structurally. The only place where the matter exists as a single coherent thing, with its parties and deadlines and document set and the reason a clause was conceded in round four, is inside one person's memory. Every system in that list holds a fragment. None of them holds the matter.

This has two consequences that firms feel constantly and rarely name.

The first is that adding good software can make the day more fragmented, not less. A tool can make an individual task dramatically faster while still adding another place where context has to be carried, another system that needs to be checked, and another boundary between pieces of the same matter. The software has made the task more efficient. The work, taken as a whole, may not be.

The second is more serious. When the matter lives in a person, it leaves with them. An associate rotates off and something is lost that no handover note captures. A partner goes on leave and three decisions have to be reconstructed. This is normally described as a knowledge management problem, which makes it sound administrative. It is closer to a design problem. We have organised legal software around documents and tasks, because those are easy to build for, while the actual unit of legal work is the matter, which nobody built for.

None of this is solved by a better model. It is a question of what the work is organised around.

So, genuinely: count your tabs. Then ask what that number is measuring. In our experience it is not a distraction. It is how much of the firm is being held together by people rather than by anything you could point to.