AI maturity curve for accountants/bookkeepers

Artificial intelligence is already changing how accounting and bookkeeping work gets done. If you’re like many practitioners, you’re already using AI tools daily.

Tasks get completed faster. Admin shrinks. Communication sharpens.

On the surface, that looks like progress. But there’s a second effect that deserves more attention.

As AI reduces the time required to complete core work, it creates capacity. But capacity, on its own, doesn’t create value.

If you halve the time spent on a compliance job but charge the same fixed fee, you have empty hours.

If you halve the time and charge by the hour, you’ve cut your own revenue. Neither outcome is progress.

Recent Sage research with AccountingWeb shared at the Finance, Accounting and Bookkeeping Show (FAB), found that while 71% of firms are now using external AI tools, just 7% describe the impact as transformational.

Most firms are using AI. Almost none have changed anything fundamental.

This is the capacity gap: the distance between time saved and value captured.

It’s a central economic challenge of AI adoption in accounting, and it’s forming in firms right now, often invisibly.

AI adoption tends to move through stages, from experimentation to productivity to deeper structural change.

Understanding where your firm sits on that progression matters.

But the more important question is whether you’re closing the capacity gap or widening it.

Here’s what we’ll cover:

The AI maturity model

AI adoption is often described in terms of which tools you use.

In practice, the tools matter far less than what they change about how your firm operates, delivers work, and gets paid.

The stages below offer a way to assess where your firm currently sits. You’ll likely recognise elements of more than one. The important question is which direction you’re moving in.

Stage 1: Awareness and experimentation

At this stage, AI is something individuals are exploring rather than something the firm has adopted.

You or your team are trying ChatGPT, Claude, Copilot or Gemini for emails, notes, or quick summaries.

Usage is informal, uncoordinated, and disconnected from how the firm actually delivers work.

The Sage AI labs at FAB showed there was curiosity and genuine enthusiasm.

But a recurring theme from practitioner discussions is that many people are using AI without a clear sense of how it fits into their operations. It’s happening around the business, not inside it.

The capacity gap is already forming at this stage, even if it’s hard to see.

You save time on tasks, but your firm has no plan for what happens to that time.

And it doesn’t sit empty. It gets quietly absorbed by the expanding scope of your role—supporting clients with technology questions, navigating business decisions, or handling conversations that have nothing to do with accounting.

The time AI saves disappears into work that was never part of the job description and hasn’t been priced for.

Stage 2: Assisted productivity

The next stage focuses on efficiency.

Your firm begins using AI more consistently across day-to-day work—drafting communications faster, summarising meetings, supporting basic analysis, or cutting time on repetitive admin.

The improvements are real and they compound across weeks, months and years.

This is where AI feels like success. Clients are served. Deadlines are met. Your team is producing more with less effort. Nothing appears broken.

That’s what makes this stage dangerous.

It’s easy to get stuck here. Not because you’ve failed to adopt AI, but because assisted productivity feels like enough.

There’s no obvious signal that anything needs to change. But the underlying model of your firm remains untouched. AI is improving how tasks are completed, not how work is structured, delivered, or priced.

The data backs this up.

In Sage research conducted with AccountingWeb, the most common AI use cases reported by firms are tax rules, drafting emails, summarising information, and generating reports.

These are all efficiency gains. None of them change how the firm operates or gets paid.

The same pattern shows up in outcomes. Reducing time spent on manual tasks is the most widely reported benefit. Expanding services sits near the bottom. Firms are saving time. Almost none are using it to offer something new.

The longer your firm operates here without rethinking its workflows or commercial model, the wider the capacity gap becomes.

Time is being saved across the practice, but none of it is being converted into new value.

Instead, hidden work around the edges of the profession consumes much of that capacity—the technology support, the business advice, and the client hand‑holding that you absorb because there is no one else to do it.

AI creates time. The expanding role of the modern accountant or bookkeeper takes it back. And because everything on the surface looks like progress, the problem of hidden hours doesn’t get addressed.

Stage 3: Workflow integration

This is where the shift becomes more substantive.

AI moves from being a personal productivity tool to shaping how work flows through your firm, embedding itself into:

• bookkeeping, reporting, and communication workflows,
• connected to client data and historical records,
• applied to recurring processes,
• and standardised across the team.

This transition can create real friction.

AI performs well in isolation, but without access to actual client data and real workflows, its usefulness degrades quickly.

Tools that aren’t connected to the systems where work happens require constant manual checking, which erodes much of the efficiency gain they were supposed to deliver.

As practitioners at FAB noted repeatedly, AI becomes significantly more useful when it operates within context rather than alongside it.

The challenge here isn’t finding better tools. It’s redesigning how work moves through your firm so that AI becomes part of the infrastructure, not an add-on.

This is also where the gap between firms can widen.

Practices that integrate AI into their workflows build scalable, consistent ways of operating.

Those that remain at the level of individual usage see uneven results, no firm-wide benefit, and a capacity gap that continues to grow unchecked.

Stage 4: Commercial transformation

At this stage, the impact of AI stops being operational and becomes economic.

AI reduces the time required to deliver core compliance work.

Tasks that once took hours can be completed significantly faster, or in some cases automatically. In theory, that creates real, measurable capacity across your firm.

In practice, that capacity often isn’t freely available.

Consider what plays out in practice. You automate significant portions of a recurring compliance job.

Work that used to take eight hours now takes three.

Bill the client by the hour? Revenue drops by more than half.

If the client pays a fixed fee, your firm has gained five hours of capacity—but only if you have somewhere productive to put them.

If neither your service model nor your pricing changes, you’re doing the same work in less time and capturing none of the value AI created. More efficient and less profitable at the same time.

This is the structural reality of AI adoption in a profession that has priced work on time and complexity for decades. AI collapses both.

It makes complex work simpler and fast work faster. AI compresses the basis on which many firms charge, and firms that adopt it most enthusiastically often compress it fastest.

To close the capacity gap, you should do a few things differently. Use freed capacity to take on more clients without increasing headcount.

Redirect time from processing to interpretation—less effort preparing data, more time helping clients understand it.

Expand into advisory and planning work that compliance deadlines previously squeezed out.

Firms reshape how services are packaged and priced, shifting away from time‑based billing towards value‑based models that reflect outcomes rather than hours.

But this is harder than it sounds.

Advisory work requires a different skill set, a different client conversation, and a different commercial model. You might not be set up for that pivot.

The capacity gap isn’t just a firm-level problem. It’s a profession-level pricing challenge, and no individual firm can solve it simply by deciding to offer more strategic services.

Firms that don’t close the gap will continue to deliver the same services in less time, and the efficiency gains will quietly compress their margins.

They’ll have adopted AI successfully by every visible measure. They just won’t have captured the value it created.

Where is your firm today?

Many firms currently sit somewhere between assisted productivity and workflow integration.

AI is being used, and in some cases, it’s delivering clear efficiency gains.

But in many practices, individuals still drive usage, while firms fail to embed it firm‑wide, connect it to the platforms where client work happens, or link it to how services are structured or priced.

The capacity gap is already forming.

Time is being saved in pockets. No one has decided what to do with it. And because productivity gains feel like progress, there’s no urgency to ask the harder question: where is that time going, and is it generating any return?

AI and MTD: a single structural squeeze

Making Tax Digital for Income Tax increases the frequency and continuity of work across the year.

Instead of annual compliance cycles, you move towards ongoing interaction with clients, with quarterly updates creating a rolling obligation that doesn’t pause between submissions.

AI changes how that work can be delivered.

It creates the capacity needed to manage more regular reporting and client support.

But it also sharpens the pricing question. When firms prepare quarterly submissions faster and partially automate routine bookkeeping, the time required to serve each client falls. But the responsibility doesn’t.

Together, they create a situation where firms are doing more work, more often, in less time, with no mechanism to capture the additional value. MTD increases the demand for continuous work.

AI compresses the economics of delivering it.

And beneath both, your role continues to expand into territory that neither system accounts for—the hours spent on client support, business guidance, and relationship management that don’t appear in any workflow or any invoice.

The capacity gap, the continuous responsibility challenge, and the invisible expansion of the practitioner’s role are not three separate problems.

They are the same structural squeeze viewed from different angles. Address one without the others and the pressure shifts rather than resolves.

Final thoughts

AI adoption is often framed as a question of which tools to use.

In practice, the tools are the easy part. The harder question is what happens to the time AI creates, and whether anyone is capturing the value.

The profession is heading towards a split.

On one side are firms that use AI to create capacity and then find ways to convert that capacity into better services, stronger client relationships, and more sustainable commercial models.

 On the other are firms that use AI to do the same work faster, charge the same fees, and see their margins erode so gradually it only becomes visible once the pressure is structural.

Most firms won’t realise which side they’re on until the numbers tell them.

Firms that benefit most financially from AI won’t be the fastest adopters. They’ll be the ones closing the capacity gap.

---

---