Thanks to Diana Buccella of Laurel for sharing this article she wrote with us.
Without visibility into how work actually gets done, every AI investment is a bet, not a decision.
That was the thread running through our recent webinar with Eric Zaarour, Co-founder at Laurel, and Tom Barry, Managing Partner at GHJ and North America Chair at HLB International. The conversation kept returning to the same place: you cannot improve what you cannot see.
The Paradox Hiding in Plain Sight
Firms have never been more profitable, and they have also never been more confused about why.
Over the last five years, the traditional KPIs used to measure performance, charge hours, utilization rates, hours worked per person, have been trending down. At the same time, revenue is growing and clients are being served. The metrics that were supposed to explain firm health are telling a different story than the P&L.
A professional still works eight hours a day, but with better tooling, more gets done within those same hours, and the input, measured in hours, no longer reflects the output being delivered. Capacity increases without any corresponding increase in recorded time, and the traditional metrics have no way of capturing that shift.
The instinct many firm leaders have is to push for more hours. If we get five more billable hours out of each person, profits go up, and mathematically that is true. But it is also the wrong question. The right question is what is actually happening inside those hours, and whether leadership has the visibility to know.
The Gap Between Assumption and Reality
Most firms are operating on estimates about what their work costs to deliver, and the gap between those estimates and reality tends to be larger than anyone expects when they finally see it.
The pattern is consistent across firms: what leadership assumes takes six hours routinely takes ten, and that gap is where margin compression lives. It is almost always more significant than firms predict when they finally have the data to see it clearly.
Part of what makes this hard to diagnose is that the drift in efficiency is not uniform. High performers complete work at a different pace than junior associates, and firms that build margin models assuming average performance across a service line are quietly underestimating their costs. The data, when it finally surfaces, tends to show that those assumptions were optimistic across the board.
When timekeeping is inaccurate, everything built on top of it is inaccurate too. Without objective visibility into what is actually happening at the task level, recovery rates are guesses, realization is guesswork, and any conversation about AI-driven efficiency has no baseline to measure against.
Why AI Makes This More Urgent, Not Less
AI accelerates the work that was already captured, but the work that was never recorded stays invisible regardless of what tools a firm deploys.
When a firm uses AI to speed up drafting, research, or document review, it is making captured work faster. It is not surfacing the work that was never recorded in the first place. And if that unrecorded work contains the highest-leverage activities, the partner judgment calls, the nuanced client conversations, the reasoning clients are actually paying for, then the AI strategy has no way of knowing where to focus.
Firms need to be intentional about understanding what they are doing, because time is the largest asset the firm holds, and timekeeping, done right, is the foundation for every downstream decision that matters: pricing, staffing, AI investment.
Think about how you treat your health. When a doctor gives you data showing where something is wrong, you act on it because the data makes the path forward clear. The problem in professional services has never been a lack of willingness to act. It has been a lack of ability to observe, and the firms that close that gap will be the ones making decisions grounded in evidence while everyone else is making educated guesses.
The Question Worth Taking Back to Your Firm
AI is not arriving. It is already inside your firm, changing how work gets done faster than most leadership teams can track, and the firms that navigate this well will not be the ones that deploy the most tools.
They will be the ones that understand their own work clearly enough to know which tools to deploy, which activities to protect, and which decisions actually require the kind of human judgment that cannot be automated. That clarity requires visibility, and visibility requires capturing work as it actually happens rather than as people remember it.
If you don’t know what your work actually costs to deliver today, you don’t have a foundation to price it, staff it, or improve it tomorrow.
