For accounting firms exploring AI, the technology itself is only part of the story. The more useful question is what changes in the business. Drumm Nevatt & Associates processes more than 1,300 accounting engagements a year. Its AI Investment Fund project aims to reduce repetitive review work so experienced accountants can spend more time thinking about clients, opportunities and advice.
Watch the conversation
Watch the conversation on YouTube
The Short Version
A useful AI project starts with a business problem and a number you want to move. Drumm Nevatt & Associates is developing an AI pre-review assistant to catch routine inconsistencies before manager review. If it works as expected, the bigger benefit will not simply be minutes saved. It will be more capacity for experienced accountants to think about clients, provide advice and do higher-value work.
About Drumm Nevatt & Associates
Drumm Nevatt & Associates is a New Zealand chartered accounting and business advisory practice with offices in Howick, Auckland and Tauranga. Its work spans accounting and tax compliance, business development, cashflow management, trusts and estates, and advice to help business owners understand and improve their financial position.
The firm has also been looking at how technology can remove repetitive work without removing the judgement and client relationships that sit at the centre of good accounting. That makes its AI pre-review project a useful example of what practical AI adoption can look like inside an accounting firm.
What’s Covered
Five Things Worth Taking Away
Start with a business problem, not a desire to use AI.
Measure the outcome. Time saved only matters if the business knows what it will do with that capacity.
For accounting firms, AI may be most useful when it gives skilled people more time for judgement, clients and higher-value work.
Adoption becomes easier when AI solves a visible problem and becomes part of the normal workflow.
Access to secure, usable business data can determine whether a promising AI idea works in practice.
Like many accounting firms, Drumm Nevatt & Associates has work that requires both detailed checking and experienced professional judgement.
A staff member may prepare annual accounts, tax returns, minutes and supporting calculations before the job moves to a manager for review. From there, it may eventually reach a principal before going to the client.
The problem is that relatively small inconsistencies can send otherwise completed work backwards through that process.
A figure in the accounts may not match the tax return. A date may not have been updated. Information in one document may differ from another.
They may be minor corrections, but each one can mean somebody reopening the job, making the change, returning it to the manager and having the work checked again.
“When we’re focussed on the nitty gritty, it’s easier to miss that stuff because you get stuck in the weeds.”
The proposed AI pre-review assistant is intended to sit before manager review. It would check documents for consistency and completeness, flag exceptions and give staff an opportunity to resolve them before the work reaches a senior person.
AI in accounting firms does not have to mean handing financial decisions to a machine. One practical use is to give AI a narrow checking task before a qualified accountant reviews the work.
In Drumm Nevatt’s case, the proposed AI pre-review assistant would compare information across documents and look for inconsistencies or omissions that may otherwise be picked up during manager review.
The manager still reviews the work. The principal still applies professional judgement. What changes is where their attention goes.
Instead of spending as much time finding small discrepancies, managers can focus more on questions such as:
That distinction matters when thinking about AI for accounting firms. The purpose is not necessarily to remove the accountant from the process. It may be to remove some of the repetitive checking competing for the accountant’s attention.
The most useful part of the Drumm Nevatt project is that there is already something to measure.
When Craig submitted the idea, he estimated that the new process might save around 20 minutes per engagement. He then asked members of his team. Their estimates were higher, at around 30 to 45 minutes.
With more than 1,300 engagements being reviewed each year, relatively small changes at job level can add up.
| Expected Saving Per Engagement | Illustrative Annual Capacity Across 1,300 Engagements |
|---|---|
| 20 minutes | Approximately 433 hours |
| 30 minutes | Approximately 650 hours |
| 45 minutes | Approximately 975 hours |
These figures are illustrative estimates based on the expected time saving discussed in the interview. They are not measured results from the completed project.
Craig also sees an opportunity to reduce the delay created when a job is picked up, sent back, corrected and picked up again. He estimates that removing some of that back-and-forth could make certain jobs seven to ten days faster.
That is another hypothesis the team can measure once the new review process is operating.
This is where measuring AI ROI becomes more interesting than simply adding up minutes.
Saving 30 minutes is useful. But the commercial result depends on what happens to those 30 minutes afterwards.
For Drumm Nevatt, Craig sees several possible uses for the additional capacity. The practice may be able to complete more work internally, reduce some outsourcing, improve turnaround times and give senior accountants more room within existing job budgets to think about the client.
That last point may prove to be the most valuable.
“They understand that we’re taking some time to think about their business, not just sliding them a tax return.”
More time for client conversations could mean better advice and stronger relationships. It may also create opportunities for clients to ask more questions and seek help in areas beyond their annual accounts.
Those commercial effects still need to be measured rather than assumed. But they demonstrate an important point for any business evaluating AI.
The value is not simply the time the technology saves. It is what the business can do with the capacity it creates.
A strong business case means very little if employees do not use the new process.
Drumm Nevatt’s approach is useful because the proposed AI review is not being treated as another optional tool employees need to remember to open.
The workflow itself changes:
Craig believes that will make adoption easier because the AI step simply becomes part of getting a job ready for review.
The team also has experience with robotic process automation. Craig says there was some initial pushback and the rollout was not perfect, but employees have since seen repetitive tasks disappear from their workloads.
That creates a much clearer reason to adopt something new. The employee is not being asked to use AI because AI is fashionable. They can see which part of their working day becomes easier and what they can spend more time doing instead.
Craig also expects AI to affect the skills accounting firms value.
Accounting has traditionally rewarded process discipline and attention to detail. Those qualities will remain important, particularly when someone still needs to check whether an output passes what Craig calls the “sniff test”.
But as software takes on more repetitive process work, he expects other qualities to carry more weight:
That could change the day-to-day job as well.
Craig’s 12-month vision is for senior accountants to have more time to work directly with clients, build their confidence in meetings and become less transactional in how they manage relationships.
For accounting firms considering AI, that may be the more useful way to think about the people question. Instead of asking only which jobs AI could do, ask which parts of a person’s job are worth giving back to them.
There is no single AI use case every accountancy firm should adopt. The strongest opportunities tend to start with a process that already causes measurable delay, repetition or rework.
Depending on the systems involved and the controls required, possible areas to investigate can include:
The starting point should still be the business problem rather than the technology.
An accounting firm that knows a review stage repeatedly creates 30 minutes of rework has something concrete to investigate. “We should be doing more AI” is much harder to turn into a useful project.
A good idea can still run into a basic problem. The AI needs access to the right information.
Craig has experimented with building AI agents himself. The issue he keeps encountering is not necessarily the capability of the AI model. It is access to business information spread across different platforms.
An agent may work in principle but still be unable to access an internal system or the data required to complete the task.
This is particularly important for AI in accounting firms, where financial and client information needs to be handled carefully.
Before moving from an AI experiment to a business process, ask:
A compelling AI use case needs more than a good prompt. The surrounding data, systems, security and workflow need to support it.
The lesson from Drumm Nevatt’s project is not that every accounting firm needs an AI reviewer.
It is that useful AI starts with an existing business problem and a result you can measure.
Before investing in an AI project, four questions are worth asking:
If those questions cannot be answered, it may be too early to invest in the technology.
If they can, the business has something much more useful than an AI idea. It has a testable business case.
Before Investing In AI
A practical AI project should connect the technology to an outcome the business can observe before and after implementation.
A Practical Model For AI Value
About The Project
Drumm Nevatt & Associates was selected through the First Focus New Zealand AI Investment Fund, receiving $20,000 in support to develop and test its AI pre-review idea.
The project is still being implemented, so the time, turnaround and commercial outcomes discussed above remain hypotheses to test. The next step is to compare those expectations with what actually happens after implementation, including adoption, time saved, turnaround and how the additional capacity is used.
About The Guest
Chief Executive, Drumm Nevatt & Associates
Craig formed Drumm Nevatt & Associates in 2016 after building experience in business advisory, including 11 years with PwC and running his own virtual CFO practice.
In this conversation, he discusses how the firm is approaching AI, where it expects to create measurable value and why removing repetitive process work could give accountants more time to work directly with clients.
Common Questions
Have an AI idea but not sure how to connect it to a measurable business outcome?
IT Consulting
IT Strategy
Data Services
IT Infrastructure Projects
IT Procurement
Virtual CIO
Offshore Staff Outsourcing (BPO)
Managed AI & IT Services
IT Support Services
Managed Cloud Services
Managed SharePoint Services
End User Computing
Unified Communications
Connectivity
Managed Security Services
Cybersecurity Risk Assessment
Business Continuity Planning
Cloud Backup and Disaster Recovery
Automation
Digital Transformation
Power BI
Careers
Awards and Certifications
Management Team
Case Studies
Sell Your IT Business
Our Commitment to the Environment
CSAT Charity Donations
Media Kit
Auckland | Tāmaki Makaurau
