Do small professional services firms really need a formal AI strategy?

The Thomson Reuters 2026 AI in Professional Services Report — which surveyed 1,514 professionals across 27 countries — found that firms with a formal AI strategy are more than 3x as likely to achieve positive ROI as those without one, and 2x more likely to see revenue growth. The strategy doesn't need to be a 50-page document. For a 10-person firm, a one-page document covering which tools you use, for which workflows, with what human review requirements, is sufficient. The data is clear that having written it down matters more than what it says.

What percentage of professional services firms have a written AI strategy?

Only 40% of organizations across professional services sectors now use AI at the organizational level — up from 22% in 2025. But among those that do use AI, the measurement picture is worse: only 18% formally measure AI ROI, and 40% of respondents in the Thomson Reuters 2026 survey said they have no idea whether their organization even tracks it. Most firms that are using AI are doing so without measuring it — which means they can't tell if it's working or adjust what isn't.

What is 'agentic AI' and when will it affect professional services firms?

Agentic AI refers to AI that takes autonomous action across multiple systems — it doesn't just answer questions, it performs tasks. Examples include AI that drafts and sends client follow-ups, pulls data from your billing system and flags anomalies, or processes an intake form and schedules a consultation without a human clicking through each step. The Thomson Reuters 2026 report found 77% of professional services professionals expect agentic AI to be central to their workflow by 2030, but only 15% currently use it. The implication for firm owners: the question is not whether this is coming, but whether your firm's workflows and client agreements are designed to accommodate it when it does.

What should be in a small firm AI strategy?

Based on the Thomson Reuters 2026 findings, the minimum viable strategy covers three areas: (1) Which AI tools you use and for which specific tasks — not 'we use AI' but 'we use Fathom for meeting notes, Copilot for document drafting, and CoCounsel for contract research.' (2) What the human review checkpoint is for each AI-assisted output before it reaches a client. (3) How you'll know if the AI is working — a single metric per workflow, tracked monthly. This doesn't require a consultant or a policy committee. It requires 30 minutes and a shared document your team can see.

How do I measure whether our AI tools are actually working?

Start with one metric per workflow. For meeting summarization: how many hours per week do you spend on post-call notes versus six months ago? For billing capture: compare your average billable hours per week this quarter versus last year. For document drafting: what's the average time from intake to first draft now versus six months ago? The Thomson Reuters 2026 data found that most firms that do measure focus on cost savings and employee usage — but fewer than 30% track client-facing metrics like satisfaction or retention. For a professional services firm, 'did our clients notice?' is a more important metric than 'did we save internal time?'

Written AI Strategy: The 3x ROI Advantage

The question most professional services firm owners are asking about AI in 2026 is the wrong one.

They're asking: which tool should we use?

Thomson Reuters just published the most comprehensive survey of AI in professional services to date — 1,514 professionals across law, tax, accounting, corporate risk, and government sectors in 27 countries. The headline finding answers the real question, and it isn't about tools.

Firms with a formal AI strategy are more than 3x as likely to achieve positive ROI as those without one — and 2x more likely to see revenue growth.

The tool matters less than whether you wrote it down.

Summary

Thomson Reuters surveyed 1,514 professionals across 27 countries and found that firms with a documented AI strategy are more than 3x as likely to achieve positive ROI — and 2x more likely to see revenue growth. Only 18% of firms formally measure AI ROI. The fix isn't complex: a one-page document covering which tools you use, for which workflows, with what human review requirements is sufficient to capture the advantage.

The Strategy Gap Is Now Measurable

Overall AI adoption across professional services has nearly doubled in a year: 40% of firms now use AI at the organizational level, up from 22% in 2025. That's a significant number of firms doing something with AI. But "doing something" and "getting results" are two very different states.

The difference, according to the Thomson Reuters data, is a formal strategy. Not an enterprise AI program. Not a consulting engagement or a steering committee. The 3x ROI advantage comes from having a documented approach — which tools, for which workflows, with what human review requirements.

A firm that says "we use ChatGPT for drafts sometimes" is in a fundamentally different position than a firm that has written down "we use Claude for contract first drafts, attorney reviews before client delivery, and we time each matter to measure impact." The second firm can measure, adjust, and compound. The first firm is hoping.

The Measurement Blind Spot

There's a related finding in the Thomson Reuters data that makes the strategy gap worse: only 18% of organizations formally measure AI ROI. Another 40% of respondents say they have no idea whether their organization even tracks it.

This means most firms that are using AI — even the ones seeing genuine productivity gains — have no systematic way to know what's working. They can't tell which workflows are delivering results, which tools are worth the subscription cost, or where the next investment should go.

Among firms that do measure, 77% focus on cost savings and 64% track employee usage. But fewer than 30% track client-facing metrics: client satisfaction, client retention, or new business generated from AI-enabled services. For a professional services firm where the business is relationships, this is a significant gap. "Are our clients happier or more loyal because of how we're using AI?" is a better question than "did we save internal time?" — and most firms aren't asking it.

The Agentic AI Window

Here's the planning horizon that makes the strategy question urgent for firm owners who are still in the "we'll figure this out later" camp.

The same Thomson Reuters survey asked about agentic AI — AI that takes autonomous action across systems, not just answers questions. An AI that drafts and sends the follow-up, reconciles the billing automatically, processes intake and schedules the consultation without someone clicking through each step.

77% of professional services professionals expect agentic AI to be central to their workflow by 2030. Only 15% use it today. Another 53% say their organization is "planning or considering" it.

Four years. That's the window between now and the point when the majority of your competitors believe this technology will be a standard part of how professional services gets delivered.

In the legal sector specifically: 50% of legal professionals now see AI's impact on unauthorized practice of law as a major threat — up from 36% in 2025. Practitioners inside the profession are increasingly worried that AI will blur the line between "legal information" and "legal services" in ways that affect small firm revenue. This concern is coming from lawyers themselves, not from tech press.

The firms that will navigate that transition well are the ones building the foundation now — documented workflows, measured results, clear human review protocols — not the ones that wait until the pressure is obvious.

The 30-Minute AI Strategy for a Small Firm

The Thomson Reuters data doesn't require a 50-page strategy document to act on. For a firm with 5–50 people, the minimum viable AI strategy fits on one page. Here's the structure:

Section 1: Tools and Workflows (10 minutes)

List every AI tool your firm currently uses. For each one, write one sentence describing the specific task it handles. Not "we use it for documents" but "we use it for first-draft NDA review before attorney markup." If you can't write that sentence, the tool isn't being used systematically enough to appear in the strategy.

Section 2: Human Review Requirements (10 minutes)

For each workflow, define the checkpoint. At what point does a licensed professional review the AI output before it reaches a client? "Attorney reviews contract markup before delivery" is a review requirement. "We check it sometimes" is not. The strategy must name the step and who is responsible for it.

Section 3: One Measurement Per Workflow (10 minutes)

Pick one metric for each AI-assisted workflow that you'll track monthly. It doesn't need to be elaborate:

Meeting notes → hours per week on post-call documentation
Document drafting → time from client request to first draft delivery
Contract review → matters closed per attorney per month
Billing capture → average billable hours per timekeeper per week

At the end of 90 days, you'll have data. That data becomes the foundation of every subsequent AI investment decision your firm makes.

Why This Matters More Than the Tool Choice

The tool landscape in 2026 is relatively competitive. For any professional services workflow you're trying to automate or accelerate, there are multiple viable options at comparable prices. CoCounsel and August both handle contract review. Fathom and Otter both capture meeting notes. Claude and GPT-5.4 both draft documents well.

The competitive advantage is not in selecting the theoretically superior model. It's in building the workflow that makes any tool produce consistent, measurable results. And the only way to build a repeatable workflow is to write it down.

The 3x ROI finding isn't an anomaly. It's consistent with every study that has tried to identify what separates the firms in the top adoption quartile from the majority: they structured their AI use. They know what they're using, why, and whether it's working.

Your competitors who are pulling ahead on AI didn't necessarily choose better tools. They wrote it down first.

The Action Item

Block 30 minutes this week and draft your firm's one-page AI strategy. Use the structure above. Share it with your team so they can see which workflows have defined review requirements and which don't.

If you discover you're using AI tools without measurement, add the measurement before you renew the subscription. If you discover you have no human review checkpoint on a client-facing workflow, add it before the next client delivery.

The Thomson Reuters data is clear: the strategy document is not bureaucracy. It's the thing that separates the firms compounding their AI advantage from the ones that are hoping it works.

Sources: Thomson Reuters Institute, "2026 AI in Professional Services Report" (February 2026, 1,514 respondents, 27 countries). Additional analysis: AI-to-ROI Substack, NY Daily Record legal profession highlights.

The Firms Winning at AI Have One Thing You Probably Don't: A Written Strategy