Accounting's pipeline problem is structural and well documented: fewer accounting graduates, fewer CPA candidates, and a large cohort of practitioners at or near retirement. Firms feel it as permanently open seats and busy seasons that run on overtime. The consequence is blunt. Capacity has to come from workflow, because it is not coming from the labour market.
The research, applied to a firm
McKinsey's automation research estimated that about 30 percent of activities in most occupations are automatable with technology already demonstrated, and the work of a firm in busy season is unusually dense with that kind of activity: chasing documents, answering status calls, moving data between systems, assembling the same package for the tenth client. Deloitte's 2022 survey of 479 executives found organisations that scaled intelligent automation beyond pilots reported an average cost reduction of 32 percent. In a firm the same work returns hours, and hours are the constraint.
Automate first, in this order
Client document collection: requested, reminded, received, filed and checked for completeness without a staff member chasing. Status updates: clients told where their return or engagement stands before they call to ask. Scheduling and reminders for reviews and signings. Engagement letters and onboarding paperwork issued and tracked. Data moved between the tax, bookkeeping and practice systems without re-keying. Each one is high volume, structured and countable, which is where Stanford's Index finds the measured gains.
What stays with the CPA
Judgment, review and sign-off. The engine prepares and moves; the professional decides. A firm that lets automation opine on a position has misunderstood the tool. A firm that uses it to stop chasing documents has understood it exactly.
The upside is the advisory shift
Advisory and client accounting services carry better margins and stickier relationships than compliance work, and selling them takes partner time that document chasing and status calls consume. Automating the internal grind is what funds the move up-market every firm's strategy already promises. Measure it: hours per engagement, partner hours in delivery versus advisory, document turnaround, and the busy-season overtime line.
Sources
- McKinsey Global Institute, “A Future That Works” (2017) — ~30% of activities in most occupations are automatable with currently demonstrated technology
- Deloitte, Automation with Intelligence (2022 global survey, 479 executives, 35 countries) (2022) — 32% average cost reduction achieved by organizations that scaled intelligent automation beyond pilots
- Stanford HAI, AI Index Report 2026 (2026) — 14–26% measured productivity gains from AI in customer support and software development