A weekly half-hour, ten years saved

A managing partner at an accounting firm in Sheung Wan recently described their week to us during an audit. They were trying to be helpful. They listed what they did between Monday and Friday. The listed items took thirty-eight billable hours. The actual week had been about fifty-five.

The seventeen-hour gap was not a mystery. The partner could account for it precisely. It was, almost entirely, made up of half-hour tasks. Reviewing a junior’s reconciliation work. Checking a client portal for documents that should have arrived. Sending a follow-up email about an outstanding question. Answering a quick query from a former client about whether the firm could refer them to someone for a separate matter. Reformatting a deliverable that had been built off the wrong template. Updating a spreadsheet of who-owes-what before the partners’ meeting on Friday afternoon.

None of these took thirty minutes. Many of them took five. Together, they consumed seventeen hours.

This is, in our experience auditing professional services firms, the modal pattern. The most expensive operational problem in a partnership is not a single broken workflow. It is a thousand half-hour tasks distributed across the partners and the senior staff, none of which is large enough to justify fixing on its own.

Why the half-hour task is the worst kind

A workflow that takes a partner four hours every Monday gets noticed. The partner blocks the time, complains about it at the management committee, and eventually a junior gets trained to take it on. The four-hour task has its own narrative.

A workflow that takes a partner thirty minutes spread across nine separate small tasks throughout the week is invisible. Each task individually feels like the cost of being a senior person at a service business. Across all the partners, across all the weeks of the year, it adds up to several full-time equivalents of partner-grade work being done by partners. But because no single moment is bad, no one can find a moment to fix it.

This is the cost an audit is designed to surface. We have noticed that when we shadow a partner for a working week and then describe back what they actually did, the partner’s reaction is rarely surprise at any individual task. The reaction is usually arithmetic. They look at the list of half-hour tasks, multiply by fifty weeks, multiply across however many partners are doing roughly the same thing, and arrive at a number of partner-years per decade that surprises them.

That number is what is being saved when an audit results in a successful build.

The pattern, named

The half-hour task that justifies an audit, in our notes, has three properties.

It happens repeatedly. Once a week, once a day, once per matter, once per closing. If it is a one-off, automation is rarely the right answer; the firm should just absorb it.

It involves judgement that a partner can do in seconds and a junior can do in minutes. The reason it has not been delegated is not that delegation is impossible; it is that the delegation overhead has historically been higher than the task itself. The partner does it because explaining it to a junior would take longer.

It produces an output that has a stable shape. A reconciliation report. A formatted deliverable. A status email to the same client at the same time. An updated row in a spreadsheet. The shape is stable; only the contents change.

When all three properties are present, the task is automatable, and the saving is not the half-hour. The saving is the half-hour multiplied by every recurrence over the lifetime of the system, plus the partner’s compounding ability to reinvest that time in work that only partners can do.

What the audit actually finds

Most audits we run end up identifying between three and six of these tasks. The number is fairly consistent regardless of vertical: a six-partner accounting firm in Hong Kong has roughly the same density of half-hour tasks as a litigation boutique of similar size in London. The verticals differ in what the tasks are; they do not differ much in how many there are.

This consistency is the reason the studio is built around an audit-first engagement format. When we walk in, we do not yet know which six tasks the firm has. We do know, with high confidence, that there will be approximately six of them, and that the firm has not surfaced them itself, because the surfacing exercise has never been somebody’s job.

The audit’s actual deliverable is a written list of the half-hour tasks the firm did not realise it had, ranked by recoverable time and ordered by sensible build sequence. About one in five firms reads the list and decides to address it without commissioning further work from us, which is fine. About three in five commission a sprint to address the top one or two items. The remainder commission a longer engagement to address several items in sequence.

The arithmetic, again

If we recover a partner-half-hour every week of the year, that is twenty-six hours per year per partner. At a six-partner firm, that is one hundred and fifty-six partner hours annually. Over a decade, that is roughly one partner-year of effort returned to the firm.

That is the saving from a single half-hour task. If the audit identifies five of them, and three are addressed, the firm has invested in the equivalent of three additional partners over the lifetime of those systems.

The arithmetic does not require the saved time to be spent on more billable work; it usually is, but it does not have to be. A partnership that simply reclaims fifteen hours a week of partner-grade attention has, by any reasonable measure, become a different and better firm.

This is what an audit is for.