Capacity Planning for Accounting Firms: A Method That Fits on One Page

Capacity planning has a reputation for complexity it does not deserve. The core calculation is two multiplications and a subtraction. Required hours are your clients multiplied by average hours per client. Available hours are your staff multiplied by realistically available hours each. Subtract one from the other and you have your gap.

What makes capacity planning fail is not the arithmetic. It is using optimistic inputs, running it once a year at firm level, and producing a number nobody then acts on.

Start with contracted hours. Remove holidays and public holidays. Remove sickness at whatever rate your firm genuinely experiences. Remove training, internal meetings, appraisals, and firm administration. Remove business development for anyone who does it. Remove the general overhead of being an employee: systems, email that is not client work, the unavoidable interruptions.

What remains is available hours, and for most firms it is substantially below the headline figure. Using the headline is the most common reason a capacity plan says everything is fine while the team works weekends.

Managers and partners need particular care. A manager's available client-work hours are usually far lower than a preparer's, because supervision, review and client relationships consume the day. Modeling them as full producers is how firms end up with a reviewer bottleneck they did not predict.

Ask a partner how long a standard job takes and you will get the time it takes when everything goes well. Actual hours include the chasing, the queries, the rework and the client who reorganized their records into a format nobody can use.

If you record time against engagements, use it. If you do not, this is the strongest practical argument for starting: without hours against clients, every capacity figure and every profitability figure in the firm is an estimate.

Segment rather than averaging across everything. A single blended hours-per-client number across simple and complex clients produces a plan that is wrong in both directions at once.

This is the step that turns the calculation into something useful. Firm-wide annual capacity is almost always fine, and it is almost always fine in a way that conceals the real problem.

Per role, you usually find preparation capacity is adequate and review capacity is not. That changes the response completely: hiring another preparer makes the bottleneck worse, not better, because it feeds more work into the constrained stage.

Per quarter, you find the shortage is concentrated. A firm can be comfortable across the year and heavily oversubscribed for six weeks. The fix for that is smoothing demand, not adding permanent headcount you cannot keep busy in July.

There are four honest responses, and most firms use a combination.

Smooth the demand. Move internal dates earlier for clients whose records are available sooner, so work spreads across the quarter instead of landing in the final three weeks. This is the cheapest option and the most underused.

Add temporary capacity. Seasonal preparers, contractors or offshore support, matched to the specific role that is short. Worth noting that under per-user software pricing this option carries a software cost, which is one of the ways a pricing model quietly shapes operational decisions.

Change the book. Decline, reprice or restructure the clients consuming disproportionate hours. Firms are consistently reluctant here and consistently better off afterwards.

Accept overtime, deliberately, with a cost attached and a limit. That is a legitimate choice when it is a choice. It is only a problem when it is the default that nobody decided on.

A spreadsheet done each October is better than nothing, and it is out of date by December. The version that actually changes decisions is the one that updates as work is scheduled and hours are recorded, so you can see in week two that a person is committed beyond their available hours.

That requires the work to live in a system where it is scheduled and where time is recorded against it. Firms running on a shared calendar and a whiteboard can do the annual arithmetic, but they cannot see the position today, which is when the intervention is still cheap.