Why does growth bring more admin rather than more profit?

What does growth actually do to a firm's margin?

A firm of thirty becomes a firm of sixty. Fee income roughly doubles. Profit per partner does not, and in many firms it goes backwards for a year or two before it recovers. The partners know the revenue line is healthier than it has ever been, and they also know the week feels worse than it did at thirty people.

Between those two facts sits a layer of work nobody sold and nobody prices: the work of running a bigger firm. Every piece of it was added for a sensible reason. It simply grew at a different rate from the fees, and nobody was watching that rate.

Why does admin grow faster than fee income?

Fee income scales with the volume of work. Admin scales with the number of connections around that work, and connections multiply faster than volume.

Take one matter. When a single fee earner runs it from enquiry to invoice, there are no handoffs. Split the same matter across a partner, an associate, a paralegal and a billing clerk and you have created three handoffs, each needing a briefing, a status check and a chase when it stalls. The legal work has not changed. The coordination around it has appeared from nowhere.

Now multiply the clients. Twice as many clients means twice as many take-on checks, twice as many renewal conversations, twice as many sets of reporting preferences. Each client brings its own small variations, and variations are where standard processes quietly stop being standard.

Then add the firm's own need to see itself. At twenty people a managing partner knows what is happening by walking across the floor. At eighty they need a report, and someone has to produce it, which means someone else has to feed it. Board and management reporting is pure overhead in the accounting sense: it exists only because the firm got big enough to need it.

Where does the extra work actually show up?

It is rarely one big new task. It is four or five old tasks that each got heavier.

• Taking on work. More instructions means more identity checks, conflict checks, engagement letters and file openings. Client onboarding at low volume is a form. At high volume it is a queue with a person guarding it.
• Winning work. Growth usually means bidding for larger or more formal work, so proposal and bid production moves from a two page letter to a document with a method statement, team CVs and a policy appendix.
• Telling clients what is happening. A bigger book means more client reporting cycles, each assembled by hand from systems that do not quite agree with each other.
• Proving the firm is run properly. Larger firms attract more scrutiny, from regulators, from insurers and from clients' own procurement teams. Compliance evidence that used to be gathered once a year becomes a standing job.

None of these is new work. They are the same four jobs the firm did at half the size, now running at higher frequency and higher formality at the same time. That combination is what makes them expensive.

Why doesn't hiring support staff fix it?

It is the obvious move and it half works. A good operations hire takes real weight off fee earners in the first few months. Then two things happen.

First, the judgement does not transfer. A support person can assemble a report, chase the missing figures and format the appendix, but they cannot decide what the client should be told about a difficult variation. So the fee earner still reviews, still edits, and the final approval sits where it always sat. The bottleneck moves one step back and stays the same size, which is the pattern described in the partner bottleneck.

Second, the hire adds another handoff. Work now passes between two people where it passed between none, and briefing that handoff is itself admin. The firm has converted a senior time problem into a permanent salary plus a coordination cost. That trade is sometimes right, and automation versus hiring sets out when it is. It is rarely right as the only answer.

Why do more systems often make it worse?

Growing firms buy software, and every purchase is justified. The practice management system, the CRM, the document store, the time recording app, the compliance portal a large client insisted on. Each holds a slice of the truth.

What nobody budgets for is reconciliation. Someone has to make the figure in the billing system match the figure in the report, and the matter status in the CRM match the one in the case file. Within two years the firm has a job that is really just keeping systems in agreement with each other. The tools did not create the process, they fragmented it, which is the argument in a tool is not a process.

What actually breaks the link?

The link breaks when the admin attached to one unit of work falls. Pick the unit your firm actually counts: a matter, an engagement, an instruction, a placement, a scheme, a policy. Then ask how many unbillable hours each one carries from enquiry to final invoice.

That number decides whether growth pays. If it stays flat, doubling the work doubles the overhead and the margin stands still. If it falls, growth starts converting into profit, because each new matter arrives carrying less weight than the last one did.

Making it fall means rebuilding the job rather than speeding up the typing. The record has to come out of the work itself: the report drafted from the data that already exists, the file note drafted from the meeting that already happened, the compliance pack assembled from evidence captured in the course of doing the job. A named person still reviews and approves anything that reaches a client. What disappears is the reassembling, the rekeying and the chasing.

This is also why firms should resist fixing everything at once. One rebuilt process that holds at double the volume teaches the firm more than a programme touching ten processes lightly. One process before any programme makes that case properly.

Where should a growing firm start looking?

Start with frequency, not with irritation. The process everyone complains about is often monthly. The process eating the margin is often the one that runs on every single job and is small enough that nobody mentions it.

Count the hours properly first. The guide to calculating unbillable hours sets out the method, and the unbillable hours calculator turns it into an annual figure using your own charge-out rates. Then rank the candidates by how often they run and how senior the people running them are, using the process priority scorer.

One last check before you commit. Model what happens at next year's volume, not this year's. A process that creaks today will fail at a third more work. Capacity planning when work gets faster covers where the constraint moves next. For the short version, the audit will name the process to tackle first.