Accounting Technology

79% of Heavy AI Users Expect Higher Profits, but Only 23% Bill the Time Saved

AI is saving accountants hours, and committed adopters expect sharply higher 2027 profits. Yet only 23% convert saved time into billable work — the profession's ROI gap widens.

The Problem with AI : Where’s the ROI?
Exhibit · The Problem with AI : Where’s the ROI? — Firmbee / pixabay

Adjustments

  1. 79% of committed AI adopters expect higher 2027 net profits vs. 24% of AI holdouts — a 55-point gap
  2. 44% of committed adopters save at least five hours weekly, but only 23% turn that time into billable work
  3. 85% of surveyed accountants use AI in some form; only 28% are committed adopters
  4. McKinsey: 80% report individual productivity gains, only 37% attribute EBIT impact to AI
  5. 62% expect to raise technology spending for busy season; only 20% expect to add staff

Seventy-nine percent of accountants who have made AI a routine part of their practice expect higher net profits in the 2027 busy season. Only 23% of that same group say they are converting the time AI saves them into additional billable work.

That disconnect sits at the center of the latest CPA Trendlands Outlook 2027 research. AI is delivering on its promise to save accountants time; whether it is making them more money is another question entirely.

Who is actually using AI?

AI is no longer a novelty for most respondents: 85% report using it in some form. But use means very different things at different firms.

The adoption breaks down as follows:

  • 39% use AI for a few tasks only
  • 17% are experimenting
  • 15% are holding off altogether
  • 25% use AI regularly across the firm
  • 3% say AI is embedded in most daily work

Together, the last two groups — the committed adopters — make up 28% of respondents. The productivity gap is stark: 44% of committed adopters save at least five hours per week, compared with 5% of selective users. Experimenters report no savings on that scale at all.

Across all respondents, benefits remain modest. About 33% report little or no net weekly time savings, 30% save two hours or less, and 3% say AI actually adds more time than it saves. Only 15% of the full sample save at least five hours weekly.

What does this change for the economics of a practice?

The CPA Trendlands findings point to a profession moving through two distinct stages of AI transformation. The first — well underway — involves getting work done faster and better. The second, still much less developed, involves changing the economics of the practice.

Paul Raye of Think Accountants identifies two obstacles: "Change in procedures and time to learn." Lesley R. Lewis of Lewis & Co. CPAs says she needs time "to figure out ways that AI can help our firm better and be able to do some of the tasks that are so time consuming."

The case studies show what redesign looks like in practice. At BDO USA, senior tax manager Yvonne Ofori built an AI agent for repetitive tax quality-checking work — searching PDFs, tracing figures, drafting notes. BDO reports savings of one to two hours per return and estimates a broader rollout could create substantial additional capacity, though the estimate does not represent realized revenue or verified profit.

"I didn't start with the goal of building a chatbot," Ofori says. "I started with a review problem." Her control-point framing matters for any firm contemplating similar tools: "It organizes and drafts. The reviewer validates and decides."

At Armanino, an AI tax preparation application from Accrual produces initial draft returns in a reported median of 15 minutes, with preparation-hour reductions of up to 60% on certain complex multi-entity returns. Those figures cover drafting and preparation only, not the full completion-and-review cycle, and Armanino has published no matching firmwide profit figure.

The profit expectations are hard to ignore

The gap between committed adopters and holdouts on expected 2027 net profits is 55 percentage points: 79% versus 24%. Selective users sit at 56%; experimenters at 59%. The survey notes these are expectations, not returns already generated, and do not establish causation.

The pattern holds at small firms. Among respondents at firms of one to nine people, 80% of committed adopters expect higher profits versus 21% of holdouts — though the underlying groups are small, at 25 and 28 respondents.

The broader outlook is favorable across the board: 77% of respondents expect higher billing rates, 73% expect higher revenue and 57% expect higher net profit. Fee increases alone could produce some of those gains even at firms that have made little automation progress.

Outside research finds the same gap

Financial Cents, surveying 486 North American accounting and bookkeeping professionals, finds that 56% of firms at its highest AI adoption level report clear, measurable ROI. That falls to 24% one level down and 4% among early-stage firms. Even among AI-active firms, only 20% report clear ROI; 52% call the payoff promising but difficult to quantify.

Karbon similarly reports greater time savings among advanced users, particularly where firms invest in training. Xero's Australian research links higher margins to intensive AI use, advisory services and pricing practices: higher-margin firms save about 11 hours weekly through AI versus eight hours overall.

McKinsey's 2026 global AI survey crystallizes the pattern: 80% of respondents report individual productivity improvements, while only 37% attribute any positive EBIT (earnings before interest and taxes) impact to AI.

Where the money is going

In the CPA Trendlands survey, 62% of accountants expect to increase technology spending for the coming busy season; only 20% expect to increase staffing. The comparison does not prove headcount replacement — recruitment difficulties, outsourcing and labor-cost control may all contribute — but it shows where firms will commit resources.

The economics depend on the entire production process, including work the technology creates as well as eliminates. Nine percent of respondents already say AI's costs outweigh its benefits. Under fixed-fee arrangements, fewer hours can improve margins without added billings; under hourly billing, fewer hours may reduce collections unless pricing or scope changes.

The emerging question for firms is not simply whether AI saves time. It is whether those savings survive implementation, verification and subscription costs — and whether the resulting capacity produces higher profits, better service, or lower bills for clients.

via research.net (Original)

Share this article:

More from Priya Raman

Priya Raman

Show full bio

Senior reporter covering technology and product news in CPA News is a professional daily at CPA News.

12 articles

Linked entries

  1. 74% of AI-Using Clients Would Fact-Check Their Accountant's Work
  2. Accountants Forecast a Better 2027 Busy Season — Barely Seven Months After a Bad One
  3. AI Is Killing the Big Four's Advice-Selling Business, KPMG Says
  4. IAASB issues first AI audit quality guide with 2027 rollout
  5. AI Reliability Tests Need 20 to 30 Cases, Not One-Off Checks

▲ Previous article