Taking on more clients without more staff

Every practice partner has had the same thought: if we could serve thirty clients per person instead of twenty, everything changes.

The instinct is to look for faster processing. But processing is rarely what caps the number, and practices that buy speed without addressing the real constraints get a smaller improvement than the business case promised.

Here are the five things that actually limit capacity, and what each requires.

One: documents that have not arrived

The largest single consumer of time in most practices, and it is not accounting work at all. Chasing, waiting, following up, and then doing a month's work in three days because everything arrived at once.

What fixes it: not a faster system — a different arrangement. Documents submitted as they occur rather than in a batch, through one channel, with the practice able to see what is outstanding without asking.

Automated chasing helps considerably, because it happens consistently and without the social cost of a person nagging. But the underlying change is behavioural, and it needs to be agreed rather than imposed.

Two: documents that arrive unusable

A photograph of a receipt at an angle. A statement as a screenshot. An invoice inside a forwarded email chain. A spreadsheet where the columns changed halfway down.

Modern document reading handles far more of this than older systems did, which is a genuine improvement. It does not handle everything, and the residue is concentrated in a few clients.

What fixes it: capture at source. A client photographing a receipt at the point of spend produces a better image than the same client photographing a pile of them at month end, and the difference is large.

Three: questions that did not need to reach you

"Has this been paid?" "What did we spend on freight?" "Did you get my invoice?"

Each is two minutes. Across a client book, they are a substantial share of a delivery person's week, and they arrive as interruptions, which costs more than the two minutes.

What fixes it: clients able to see their own position without asking. This is a genuine capacity lever and it is frequently overlooked because each individual question feels trivial.

Four: rework

Information that arrives after the work was done. A missing invoice, a correction, a client remembering something in week three.

What fixes it: partly the first two — if documents arrive as they occur, less arrives late. The rest is a cut-off policy that is actually enforced, which is a commercial decision rather than an operational one.

Five: everything being slightly different

The constraint that is hardest to see and most expensive. Forty clients each configured differently means work cannot move between staff, holiday cover is difficult, and the person who knows a client is a dependency.

What fixes it: standardisation, imposed by the practice rather than negotiated per client. See standardising processes across a client base.

This is the one that most limits how far capacity can rise, because the others can be fixed client by client and this one cannot.

Which of these software solves

Directly: two and three. Document reading handles messy input; client visibility removes routine questions.

Partially: one and four. Automated chasing is more consistent than a person, and earlier document capture reduces late arrivals — but both depend on the client changing behaviour.

Not at all: five. Standardisation is a decision a practice makes and enforces. Software makes it easier to maintain and cannot make it happen.

That distribution explains why implementations disappoint. The two constraints software solves outright are real but not the largest, and the largest is a management problem that no purchase addresses.

The sequence that works

1. Standardise first, at least for new clients. Every client onboarded the old way is a permanent tax.
2. Change how documents arrive for your worst few clients — that is where the effort concentrates.
3. Then automate processing, which now has clean input to work with.
4. Give clients visibility, which removes the interruptions.
5. Rebuild junior roles around exceptions rather than entry.

Practices that do step three first get a modest improvement and conclude the technology is oversold. It is a sequencing problem rather than a product one.

For a modelled view of what the capacity change can look like in numbers, see how an accounting firm adds clients without hiring.

The honest caveat

Capacity is not demand. A practice that can serve fifty per cent more clients has not gained anything until those clients exist.

The firms for which this changes the business are the ones already turning work away, or already unable to hire. For a practice with spare capacity and a thin pipeline, automation improves margin and quality of life and does not change the growth trajectory — which is a legitimate reason to do it, and a different business case.

Common questions

What limits capacity in a bookkeeping practice?

Five things: documents that have not arrived, documents that arrive in an unusable form, routine client questions that interrupt work, rework caused by late information, and every client being configured slightly differently so work cannot move between staff. The accounting itself is rarely the constraint, which is why buying faster processing alone produces less improvement than expected.

Which of those does software actually fix?

Document reading handles messy input directly, and client self-service removes routine questions directly. Automated chasing and earlier capture help with missing and late documents but depend on clients changing behaviour. Standardisation across the client book — the largest constraint — is a management decision that software can maintain but cannot create.

What order should a practice do this in?

Standardise first, at least for new clients, since every client onboarded the old way becomes a permanent tax. Then change how documents arrive for the worst few clients, then automate processing now that the input is clean, then give clients visibility to remove interruptions, and finally rebuild junior roles around exceptions. Automating processing first is the common mistake and produces a modest result.

Does extra capacity translate into growth?

Only where demand already exists. A practice that can serve more clients has gained nothing until those clients arrive, so the firms for which this changes the business are those already turning work away or unable to recruit. For a practice with a thin pipeline the benefit is margin and working conditions rather than growth, which is a legitimate case but a different one.

Related: ai accounting for bookkeeping practices · standardising processes across a client base · when you are the bottleneck