Start with the hours. A firm that wants more advisory work needs senior people available to provide it. Deeper client relationships require partner attention. Growth becomes easier to manage when recurring work follows a consistent process instead of depending on whoever is at the desk.
All three ambitions draw on capacity that is already committed. Review cycles, production checking and recurring deadlines take priority because the work has to go out. Advisory development remains on the plan until someone can make time for it.
Write down where the firm’s hours should move before comparing tools. A proposal then has a practical test: which recurring work will change, whose time will be released, and what will that person do with it? A good demonstration cannot answer those questions for the firm.
The capacity you have is already spoken for
In most firms, recurring compliance work absorbs the same capacity needed for advisory growth. Lodgement cycles, document handling, assembly and delivery fill the calendar. When the deadline tightens, senior people return to production and checking because the routine work still relies on them.
The firm can be committed to advisory work and still struggle to create it. The operating week allocates the hours before the strategy does.
Why the timing matters
The profession’s structural shortage is well documented: shortage-listed occupations, a degree pipeline that halved before starting to stabilise, and a forecast shortfall of thousands of accountants by the end of the decade, as set out in the labour evidence I have covered elsewhere. Recruitment remains important, but hiring alone is unlikely to produce all the capacity firms want for advisory work. Some of it has to come from changing the routine load inside the firm.
At the same time, AI use is spreading faster than operating practice. Thomson Reuters’ Future of Professionals research found 80% of professionals expect AI to have a high or transformational impact on their work within five years. That is an expectation rather than a result. CPA Australia’s Business Technology Report 2025 found 89% of surveyed Australian businesses used AI in 2025, while 16% described it as widely used across their business. The CPA figure covers Australian businesses, not accounting firms specifically. Together, the figures describe broad use with much thinner business-wide adoption.
Recurring professional work sits at the centre of both issues. It consumes scarce capacity and contains many of the repeatable steps that firms are starting to systemise. That makes it the sensible place to test whether AI can change the operating week rather than add another subscription to it.
Why I use a two-year planning window
I use about two years as a planning horizon. The evidence does not create a deadline. Two years is simply long enough to choose a workflow carefully, build it, test it and collect operating results, while remaining close enough for the plan to influence today’s decisions.
Client expectations around turnaround and cost are likely to keep moving as automation becomes more common. A firm can prepare by deciding which work to change, why that work comes first and where any released capacity will go. Those choices are within management’s control even though the pace of the wider market is not.
The period can easily disappear into scattered trials. A tool is tested, a pilot attracts interest, client work takes over and nobody carries the experiment into the operating model. Two years later, the capacity position is unchanged and the subscription list is longer. One governed workflow running every week provides better evidence than a collection of demonstrations.
The deliberate version
Begin with a roadmap based on the firm’s workflows. Record the volume, effort, elapsed time, consequence and readiness of the work. Some processes will be ready to systemise, others will need repair first, and a few should remain unchanged. The evidence creates an order for management to consider.
Build the selected workflow across the systems already in use and test it with the people who run the process. A bounded first component makes the evidence easier to see: what completed, what failed, where a person stepped in and whether the control record is complete.
Then operate it. Runs need monitoring, exceptions need owners and material outputs still need sign-off. The capacity result should also be tracked. Released hours have to appear in advisory work, client attention or another chosen use; otherwise the workflow may be functioning without advancing the firm’s plan.
What it means for the people
This framing gives the team a more useful discussion about roles. Routine production may reduce while review, exception judgement and interpretation become a larger share of the work. That shift is more senior and more client-facing, and it is closer to the work many professionals trained to do.
There is a recruitment consequence as well. Firms are competing for a shrinking pool of professionals. A role centred on judgement and clients is likely to be a stronger offer than one dominated by reconciliation. Redesigning the routine load can therefore support both the client plan and the employment proposition, provided the new role is real rather than promised.
Back to the question
The firm cannot set the labour market or the pace of technology. It can decide which work to change, where its people’s time should move and what operating evidence will show that the shift occurred.
If you are working through that decision, begin with one week’s recurring work and account for where the senior hours went. The capacity plan will be easier to see from there.
Damien Zeller is the Principal of VonBolt Technology Group.
