Professional Services Firms Are Using AI to Bill More Hours, Not Fewer
Law firms and consultancies deploying AI agents report large time savings on research and document review. Here is what they automate first.
Which non-billable operations the leading professional services firms automate first, the specific time and revenue recovery numbers they report, and how a boutique firm applies the same pattern without enterprise infrastructure.
AI-driven billable hour recovery in professional services refers to the practice of deploying AI agents to automate non-billable operational tasks (administrative work, document management, research, scheduling) so that professional time is redirected toward client-facing, revenue-generating work. The result is higher utilization rates without adding headcount or raising prices.
Law firms, accounting practices, and consulting firms face an odd paradox. AI threatens to automate the very work they bill for, yet the firms deploying AI agents are billing more hours, not fewer. The reason is structural: they are automating the non-billable operations that consume a large share of professional time — the exact share varies by discipline and by how a firm defines billable, so take it from your own timekeeping rather than from a benchmark. When administrative overhead shrinks, utilization rates climb. Utilization is what drives professional services revenue.
Statistics Canada puts AI use in professional, scientific and technical services at 32.4% of businesses in Q2 2026, second-highest of any sector (Statistics Canada). The firms adopting fastest are not replacing billable workers. They are recovering billable capacity that was buried under admin, research, and document management.
The Utilization Rate Crisis
Average billable utilization across professional services dropped to 68.9% in 2024, down from 73.2% in 2021, while EBITDA margins fell to 9.8% — the lowest in five years (SPI Research).
Average billable utilization across professional services fell to 68.9% in 2024, down from 73.2% in 2021, across 403 firms (SPI Research, Professional Services Maturity Benchmark). Revenue growth slowed to 4.6% year over year in 2024, from 7.8% in 2023, and EBITDA fell to 9.8%, down from 15.4% in 2023.
The math is straightforward. Worked example (modelled, not measured): a firm billing $200 per hour across ten timekeeping seats, each with 2,080 available hours a year, generates roughly $2.87 million annually at 68.9% utilization. Push utilization to 75% (the optimal threshold per SPI benchmarks) and that same firm generates $3.12 million. That is $250,000 in recovered revenue with zero new hires, zero new clients, zero price increases.
Where does the other 31% of time go?
Where the Time Goes
Small law firms consistently report spending too much of the week on administrative work rather than practising law (Legal Current). The non-client-facing load — legal research, court filings, managerial tasks — is where the hours go.
Consulting firms fare slightly better at 67-70% utilization, but still lose a third of capacity to non-billable operations (Runn, vendor blog). Agency employees average 25 billable hours per week alongside 13 hours of non-billable tasks (Promethean Research via TimeRewards, vendor blog).
This is not a talent problem. These are qualified professionals spending a third to half of their week on tasks that generate zero revenue. The fix is operational, not motivational.
What the Leading Firms Automate
McKinsey deployed Lilli, its proprietary AI platform, across its 45,000-person workforce. 72% are active users generating 500,000+ prompts per month. Consultants report up to 30% time savings in searching and synthesizing knowledge. Deck-building sessions save 90-120 minutes each (McKinsey).
In legal, Bridgewater reports 95%+ time savings on large-scale agreement reviews using Harvey AI, cutting vendor contract review from two days to two hours (Harvey AI).
Not sure where AI fits in your operations?
Take the Free AI Readiness Scorecard →KPMG launched Workbench in mid-2025 with 50 live AI agents, built on Microsoft Azure AI Foundry. It integrates directly into KPMG Clara for audit, Digital Gateway for tax, and Velocity for advisory (KPMG).
EY reports that 96% of businesses say they are transforming their tax operating models, and that 71% of organization leaders would rather hire a less experienced candidate with AI skills than a more experienced one without (EY).
The pattern across all four: AI is not replacing professional judgment. It is eliminating the operational friction between assignments, the document hunting, the formatting, the scheduling, the reconciliation.
Worked as a model, not a measurement. A boutique consulting practice billing at $200 per hour runs at 69% utilization. Partners identify three non-billable time sinks: weekly client reporting (12 hours/week across the team), discovery call preparation (3 hours/week), and time tracking reconciliation (4 hours/week). Total: 19 hours per week of professional time spent on pattern-based operational tasks.
Modelled outcome, not a measured one. If targeted agents absorbed roughly three-quarters of those 19 hours, the practice would recover about 14 hours per week of billable capacity, moving utilization from 69% to roughly 74%. At $200/hour that is on the order of $145,000 a year — from three focused automations rather than a firm-wide platform overhaul. Actual recovery depends on how much of each workflow is genuinely rule-based.
The Small Firm Advantage
McKinsey spent months building a custom platform for 45,000 employees (McKinsey, via Future of Consulting). A boutique consulting practice or a small law firm does not need that. It needs 3-5 coordinated AI agents handling specific operations: client intake processing, document preparation, meeting summaries and action items, time tracking reconciliation, and follow-up scheduling.
This is where smaller firms have an advantage. They can deploy targeted agents against their two or three biggest time sinks in weeks, not quarters. No enterprise procurement process. No IT governance committee. No 18-month implementation roadmap.
92% of legal professionals now use at least one AI tool daily (Wolters Kluwer). The adoption curve has already bent. The question is no longer whether to adopt but what to automate first.
59% of professional services leaders report difficulty predicting project resource needs in advance, which is the visibility problem underneath poor utilization (Kantata, vendor blog). Firms that measure their own non-billable time typically find more of it than they expected — the recoverable amount is your hours times your rate, not an industry constant.
Where to Start
The firms seeing results share a pattern: they started with a structured assessment of which operations consumed the most non-billable time, then deployed AI agents against those specific workflows. Not the most complex process. Not the flashiest use case. The workflow with the clearest input-output pattern and the highest cost.
Most AI projects fail to deliver value. Professional services firms and creative agencies face the same pattern. The ones that succeed begin by understanding where time and money are lost before writing a single line of automation.
- Worked example: a firm with ten timekeeping seats billing $200 per hour that moves utilization from 68.9% to 75% adds about $250,000 in annual revenue at the same headcount and rates.
- The leading firms (McKinsey, Harvey AI users, KPMG, EY) all automate the same category of work: document hunting, research synthesis, reporting, and scheduling — not billable judgment work, but the operational friction that consumes a large share of professional time.
- Small firms have a structural advantage in AI deployment speed: 3-5 targeted agents against specific time sinks can be deployed in weeks without enterprise procurement, while delivering the same utilization recovery pattern the large firms report.
A focused AI readiness assessment maps your current operations, identifies the highest-ROI automation targets, and produces a sequenced implementation plan. The difference between a firm that recovers $250,000 in billable capacity and one that spends $50,000 on a chatbot nobody uses comes down to this step.
The pattern is even clearer in verticalized firms. Law firms lose billable capacity to admin drag, CPA firms lose it to month-end bottlenecks and talent shortages, and many legal practices discover their first leak is actually broken intake, not research.
Contact us to find out where your firm's non-billable hours are going.