What should a managing partner measure every month?

Why does the monthly pack tell you so little?

Most firms between 20 and 200 people produce a monthly pack that is almost entirely financial. Fees billed. Work in progress. Lock-up and debtor days. Utilisation by team. Pipeline by stage. It is a careful document and it is usually accurate, and it still leaves the managing partner guessing about the only question that matters in the meeting: why did the month go the way it did?

Financial measures are outcomes. They tell you that recovery slipped, several weeks after anybody could have done something about it. They do not tell you that three reports sat with one partner for eleven days, or that the proposal team rebuilt the same methodology section four times, or that every fee earner in the property group now loses part of Friday to a spreadsheet nobody asked for.

Those are operational facts, and they are the ones you can change. A short set of operational measures, reviewed monthly, turns the pack from a record into a decision.

Which measures actually earn their place?

Six is about the limit. Beyond that nobody reads them, and a measure nobody reads is worse than none, because it creates the impression of control.

• Cycle time on the thing the client waits for. Pick the deliverable your clients judge you on: the report, the set of accounts, the advice letter, the shortlist, the valuation. Measure calendar days from trigger to delivery, because calendar days are what the client experiences.
• Rework rate. What proportion of those deliverables went round the review loop more than once? You are not chasing perfection. You are watching for a number that stays high, which means the loop is doing work the first draft should have done.
• Unbillable hours per fee earner per week. Not the gap between target and actual utilisation, which flatters everybody. The genuine count of hours spent on work that is necessary, internal and unsaleable.
• Where those hours go, by process. The same total, split four or five ways: write-ups, bids, client updates, compliance evidence, internal admin. This is the measure that names the problem.
• Concentration. How much of the month's output passed across the desk of one of your three busiest people. If that share is rising, capacity is falling whatever the headcount says.
• Time to first substantive reply on a new enquiry. Not an acknowledgement. The first response that engages with what the enquirer asked.

How do you collect these without buying anything?

All six can be assembled from what the firm already holds. A measurement exercise that needs a procurement decision will not survive its first busy month.

Cycle time comes from two dates in the matter or job record. If the second is not held, record it. Rework comes from version counts in the document store, or from reviewers ticking a box once per deliverable for a month. Concentration comes from the same records, counted by approver.

The time measures are different, because timesheets will not give them to you honestly. Recorded time is a billing instrument, and unbillable hours are exactly the ones people stop recording. The reliable method is a short, named, one-week count across grades and teams. The guide to calculating unbillable hours sets out how to run that without it becoming an audit, and the unbillable hours calculator turns the result into an annual figure at your own charge-out rate. Repeat it quarterly, not monthly.

What does it look like when something is wrong?

Patterns, not single readings. A cycle time that is long and consistent usually means too many handoffs; the work is moving, just through too many hands. A cycle time that is short on most jobs and occasionally enormous means a queue, and the queue is almost always in front of a person rather than a stage. That is the shape described in the partner bottleneck, and the cure is not asking that person to go faster.

A stubborn rework rate means the firm's standard lives in people's heads rather than in its templates. Every draft is a guess at what the reviewer wants, and the review becomes the real authoring. We set out what that costs in the cost of rework.

Slow first replies are about qualification, not laziness: nobody wants to answer until they know whether it is worth answering, so it waits for the person who can judge. See enquiry qualification. And a long cycle time on a report a client is waiting on carries a cost beyond the write-off, which is the subject of the cost of a late report.

Why measure by process rather than by person?

Because person-level measures change behaviour and process-level measures change the work. Publish unbillable hours by individual and people will record fewer of them. You will have improved the number and nothing else. Publish them by process and you have an agenda item nobody needs to defend: the firm loses a known quantity of senior time to bid production, and that is a fact about how bids are produced.

It also keeps the measures off the appraisal table. The moment a measure is used to judge a person it stops being evidence and becomes a target. Keep the six on the operations side of the house.

What do you do when a measure moves?

Pick one process and rebuild it properly. The measures are a pointing device, not a programme. Their job is to make the choice obvious: which single piece of work, if it took minutes instead of hours, gives the firm back the most senior capacity. The process priority scorer ranks candidates once you have two or three, and how to choose a process to automate covers the trade-offs.

Then watch the same six. Cycle time and rework move within weeks if the rebuild was real. Unbillable hours move at the next quarterly count. Concentration moves last, because it takes time for people to believe they no longer need to route everything through the same desk. That sequence convinces a partnership in a way no projection does. Measuring return goes further into what to compare against what.

If you have no measures today and want a starting read rather than a quarter of counting, the audit estimates the unbillable hours your fee earners lose each week, what they cost at your own charge-out rate, and which process to take first.