The real cost of unbillable time in an accounting practice

Most accounting and bookkeeping firms don’t have a productivity problem. They have an unpriced work problem.

Your team is busy all day. They chase bank statements, clarify coding questions, send reminders, clean up client spreadsheets, track down missing payroll information, update task lists, and prepare internal notes before a partner review. Much of that work is necessary. Very little of it is cleanly billable.

For a firm doing between $1 million and $25 million in annual revenue, this leakage commonly lands in the $60,000 to $180,000 range each year. That isn’t because staff are idle or careless. It comes from hundreds of small interruptions that sit between the client, the workflow, and the final deliverable.

A senior bookkeeper spends 20 minutes following up for a bank feed connection. A manager spends 35 minutes explaining why a balance sheet account looks wrong. A partner spends an hour reviewing a close pack that should have flagged obvious exceptions before it reached them. None of these moments looks alarming on its own.

Over a week, they become 15 to 30 hours of work that doesn’t move directly to a client invoice, a faster close, or an advisory conversation.

The best way to reduce unbillable hours is not to ask people to work harder or write longer timesheets. It is to identify the repeatable administrative work around the accounting process, then assign that work to an AI-enabled operating workflow with clear rules, escalation points, and human approval.

That distinction matters. You don’t want an AI tool making unsupervised decisions in a client ledger. You want it to collect inputs, reconcile routine transactions, identify exceptions, draft work for review, and keep the right person moving on the work only they can do.

You can see the operating model behind this on Omni Ops. For accounting and bookkeeping firms, the first opportunity is usually closer to the close process than most owners think.

Where unbillable hours actually go

Partners often focus on the visible workload spike at month-end and year-end. That’s reasonable. The close period is where pressure becomes impossible to ignore.

In many firms, 30% to 50% of staff time is concentrated into roughly four weeks of the year around major reporting deadlines. The team gets through it by extending hours, deferring lower-priority work, and relying on experienced staff to remember what each client needs.

But month-end is only one source of leakage. The larger issue is the work surrounding each client task.

The follow-up loop

A typical monthly bookkeeping workflow may require someone to:

• Check whether bank and card feeds are current
• Request invoices for uncategorised transactions
• Follow up on payroll journals or contractor payments
• Ask the client about material variances
• Confirm that accounts payable and receivable reports are available
• Update an internal task board after each response
• Send a second or third reminder when the first request goes unanswered

The accounting judgment might take 15 minutes. The process to collect enough information for that judgment can consume an hour.

The problem gets worse when follow-ups happen through a mix of email, Teams, text messages, client portals, and conversations that never make it into the job record. Staff lose time reconstructing context. Managers lose time asking where a job stands. Partners lose confidence in delivery dates.

The rework loop

Rework often begins with a preventable gap in the process.

A client uploads partial information. A team member starts coding anyway. Someone else discovers that payroll hasn’t been posted or a loan account has been treated incorrectly. The work returns to the preparer, then back to the client for clarification. The job looks complete several times before it is actually ready for review.

This is where firms quietly give away margin. The first pass may be covered by a fixed monthly fee. The second and third passes are often absorbed.

AI can reduce this kind of work by checking for completeness before the file enters the next stage. It can compare expected source documents with received documents, identify transactions outside agreed rules, and build a structured question list for the client. That doesn’t replace the reviewer. It prevents the reviewer from discovering the same missing information at the end.

The onboarding drag

New client onboarding is another common sinkhole.

Document collection, chart-of-accounts setup, historical clean-up, opening balances, software access, and workflow configuration can stretch for weeks. In a lot of firms, 20% to 30% of new clients delay meaningful billable work by a quarter because their records aren’t ready or the setup process lacks a clear owner.

The commercial cost is larger than the onboarding fee. A poor first 30 days weakens trust before the recurring service even begins. It also consumes experienced staff time that could have supported a stable client portfolio.

The advisory gap

Compliance work can consume the calendar so completely that advisory is treated as something the firm will get to later.

Later rarely arrives.

A partner might see useful patterns in a client’s cash conversion cycle, debtor days, gross margin, or payroll cost. Yet the insight stays in the workpapers because nobody has assembled the story, prepared the discussion points, or scheduled the right conversation.

That matters because advisory work commonly supports rates two to three times higher than routine compliance work. Reducing unbillable administrative hours is not just a cost exercise. It creates room to sell and deliver higher-value work.

What an AI workflow does differently

The useful question isn’t, “Can AI do bookkeeping?”

The better question is, “Which steps should happen before a person needs to apply judgment?”

For most firms, a good AI workflow handles the preparation, coordination, and exception management that surrounds the work. It does not take control of client financial decisions. It produces an organised draft, shows its sources, and routes unresolved items to the right team member.

That is what a properly designed agent does. It follows an agreed process across systems, uses defined thresholds, records what happened, and stops when a human decision is required.

The AI audit for accounting and bookkeeping is designed to map these workflows at the task level. We look at the actual handoffs, not just a list of software your firm already owns.

A practical example, the Month-End Close Agent

The Month-End Close Agent is built for the recurring work that occurs before partner review.

At the start of the close cycle, it checks scheduled data connections and pulls the available bank, accounts payable, accounts receivable, and payroll feeds. It compares the period with prior months and with your defined close checklist. It then separates routine items from exceptions.

For routine items, the agent can propose categorisations based on approved rules and prior treatment. For exceptions, it creates a queue with a clear explanation. For example:

• A bank transaction is over a defined materiality threshold
• A supplier payment differs materially from the recurring pattern
• Debtor days have increased beyond an agreed range
• Payroll clearing does not reconcile
• A suspense account has a new or unexplained balance
• A client document is missing and blocks a specific reconciliation

The agent then drafts the follow-up message using the transaction detail already available. It doesn’t send vague requests like, “Can you review this?” It can ask, “Please confirm the purpose of the $8,450 payment to ABC Equipment on 14 August. We need this to finalise fixed asset treatment for the August close.”

Once the client responds, the workflow attaches the answer to the job, updates the status, and prepares draft journal entries or supporting schedules for review. At the end, it produces a partner-ready close pack with exceptions, completed reconciliations, proposed journals, and key movements.

The accountant still reviews and approves. The partner still owns the client relationship. The difference is that neither person has to assemble the pack from scattered messages and spreadsheets.

In a firm where the close is heavily manual, reclaiming even two to four hours per client manager each week can have a meaningful effect on capacity. Across a team of six to ten delivery staff, that is how 15 to 30 hours comes back into the week.

Fix onboarding before it becomes a service issue

The Client Onboarding Agent tackles a different type of unbillable work. It turns a fragmented setup process into a guided sequence.

A new client receives a structured request based on their entity type, current accounting platform, payroll setup, reporting needs, and historical record condition. The agent tracks what has been received, identifies gaps, and sends reminders against a planned timeline.

As documents arrive, the workflow can sort them into the relevant setup tasks. It can validate basic completeness, create a chart-of-accounts setup draft from your template, and identify items that require a senior accountant’s decision. It can also prepare a clean opening trial balance workpaper based on the records provided and flag opening balance risks.

This is not about removing expertise from onboarding. A new client’s accounting structure still needs informed design. It is about removing the admin burden that causes experienced people to spend their day asking for the same information.

One trades-business owner in our network described the difference clearly. Their accounting firm was capable, but the first two months felt disorganised because every request arrived in a separate email. A guided onboarding workflow makes the firm look more in control because the client can see exactly what is needed and why.

For a useful starting checklist, download the Month-End AI Close Map for Accounting Firms. It helps you list the inputs, handoffs, approvals, and exception points that should be mapped before you automate a close process.

You can also access the worksheet directly here: download the close map.

Create advisory capacity from the numbers you already have

The Advisory Insights Agent is where reclaimed time starts to become new revenue capacity.

After the monthly numbers are ready, the agent reads the client’s recent results against prior periods, budgets where available, and business-specific thresholds. It surfaces three things worth discussing and drafts partner talking points before the meeting.

For a construction client, that might be an increase in work in progress, a fall in gross margin on new projects, and overdue receivables that are tightening cash. For a professional services client, it might identify an increase in payroll as a percentage of revenue, worsening debtor days, and a shift in revenue concentration.

The partner doesn’t need a generic AI summary. They need a short brief that points to the numbers, identifies the question to ask, and leaves room for professional judgment.

That is why Omni Advisory connects operating data to the conversations that create value. If your team saves five hours in the close process but spends all five on more low-margin corrections, you have only improved efficiency. If part of that capacity becomes structured advisory activity, you have improved the firm’s economics.

Start with one workflow, not a firm-wide project

A common mistake is trying to automate every process at once. That creates too many decisions, too many system dependencies, and too little proof that the work is paying off.

Start with one workflow that has four characteristics:

1. It happens frequently, usually weekly or monthly.
2. It has clear steps and repeatable inputs.
3. It creates a visible bottleneck for staff or clients.
4. It has an accountable reviewer who can approve exceptions.

Month-end close preparation is often the best first candidate. Client onboarding is another strong option if your growth has exposed inconsistent setup work.

Map the workflow as it actually happens. Don’t map the ideal version written in a procedure manual three years ago. Talk to the people who chase documents, prepare reconciliations, review files, and answer client questions.

Look for:

• Manual copying between systems
• Repeated follow-up messages
• Information that arrives without a job reference
• Checks performed only at the end of the process
• Review notes that recur every month
• Work assigned to senior staff because nobody else has the full context
• Decisions that can be converted into rules, thresholds, or exception queues

This is also where many firms discover that their problem is not a lack of software. It is the absence of a connected operating process. The tools exist, but the handoffs between them still depend on memory and email.

If you want to see the broader architecture, Omni shows how agents, voice, operations, and advisory workflows can work together without forcing a full technology replacement.

Measure the value in hours and margin

Before building anything, establish a baseline. Keep it simple.

Track the average weekly hours spent on follow-ups, close preparation, client data chasing, rework after review, and internal status reporting. Separate time spent on technical accounting judgment from time spent gathering, organising, and moving information.

Then calculate the loaded cost of those hours and compare it with the capacity they could create.

For example, if a firm reclaims 20 hours per week across a delivery team, that is about 1,000 hours over a 50-week year. Not every recovered hour becomes revenue. Some will go to better client service, staff development, and necessary buffer capacity.

Even so, if only a portion is redirected to advisory work, overflow capacity, or avoiding another early hire, the economics can be material. The key is to measure the outcome against the original bottleneck. Faster close. Fewer follow-ups. Less rework. More advisory meetings. Reduced partner review time.

Don’t promise that AI will eliminate unbillable hours. It won’t. Professional firms need internal communication, training, quality control, and client relationship time. The goal is to stop spending skilled hours on work that a structured workflow can prepare, track, or escalate.

Get a clear automation plan in 60 minutes

The right first step is not a software demo. It is a practical audit of where your team loses time and what should be automated first.

Book a 60-min Omni Audit and bring one real workflow to the call, ideally a recent month-end close or onboarding file that took longer than it should have.

In 60 minutes, we work through three outputs:

1. The tasks creating the most unbillable drag
2. The workflow an agent can own, prepare, or escalate
3. A practical first implementation path, including the data, approvals, and people involved

There is no slide deck exercise. The aim is to identify a use case that can produce a measurable result within your firm.

You can read more about the approach through our operations insights, but the most useful work happens when we look at your actual delivery process. See Omni for accounting and bookkeeping if you want the accounting-specific view first.

The firms that reduce unbillable hours consistently don’t demand more from already stretched staff. They remove the coordination work that prevents good people from doing the work clients value.

When you’re ready to find the 15 to 30 hours currently trapped in your weekly workflow, Book my Omni Audit.