84% of Accounting Firms Use AI. Only 22% Have a Strategy. Here's the Gap.
84% of Accounting Firms Use AI. Only 22% Have a Strategy. Here's the Gap.
There's a number that should stop every CPA firm owner cold.
Not the 84% — the part where most of your peers are already using AI tools. That's not surprising anymore. The number is 22%.
According to Progress Software's AI in Accounting Report, published May 12, 2026, only 22% of accounting firms have a defined AI strategy. While 84% of accountants use AI tools, only roughly one in five firms has decided what those tools are actually for, who owns them, and how the outputs connect.
The gap between those two numbers — 84% and 22% — is not a technology problem. It's a management problem. And it's costing firms far more than they realize.
The Numbers That Should Make Every CPA Firm Pause
Let's look at the full picture from the Progress Software report:
- 84% of accountants use AI tools
- 93% of accounting firms now offer advisory services (up from 83% — a 10-point jump in one year)
- Only 22% have a defined AI strategy
- 75% say too many workflow steps slow them down
- Average of 8 different AI tools per firm
The 93% advisory stat is the one that doesn't get enough attention. In 12 months, 10% more of the profession moved into advisory work. Clients are demanding it. Compliance is getting commoditized. The market is telling accounting firms to move up the value chain — and most are responding.
But here's what that stat means in practice: firms are taking on advisory work using AI tools they haven't formally evaluated, in workflows they haven't documented, without clear rules about when AI output needs human review before it reaches a client. That's not a strategic position. That's exposure.
The 8-tool average tells the same story from the other direction. When a firm uses eight different AI tools and 75% of the team says there are too many workflow steps, those two facts are not a coincidence. They're the direct symptom of tool adoption without strategy.
Each tool was added to solve a specific problem. None of them were designed to talk to each other. The firm now has eight independent tools, eight sets of outputs, and no map of how any of them connect — or who's responsible for the handoffs between them.
That is the strategy gap. The crossing that hasn't been made yet.
Why Tool Adoption Without Strategy Produces Fragmented Workflows
Here's how most firms got to eight tools: a staff accountant found an AI drafting tool that saved time on client emails, so it stayed. An office manager tried an AI invoicing tool during a busy month, and it became permanent. Someone watched a demo for a tax research AI and signed up for a trial that never ended. A partner read about an AI document review tool and added it to the mix.
None of these decisions were wrong. Each tool probably delivers real value for the specific problem it was adopted to solve. The problem is the architecture — or the lack of one.
The Difference Between "Using AI" and "Having an AI Strategy"
Using AI means your team has access to tools that automate or accelerate specific tasks. It's a capability. It happens one tool at a time, often at the staff level, often without formal approval or documentation.
Having an AI strategy means your firm has decided which tools are used for which workflows, who owns each workflow, how the AI outputs connect, and where a human must review before a client sees anything. It's a system. It happens at the firm level, with ownership and accountability.
The difference shows up in the 75% statistic. Three out of four accountants say too many workflow steps are slowing them down — not because they have too many tools, but because no one has connected the tools into a coherent flow. Every handoff between tools is a manual step that someone on your team is doing by hand, usually without a written protocol.
That's not an AI failure. That's a management gap that AI has made more visible.
What a Defined AI Strategy Looks Like for a 5–40 Person CPA Firm
This is where most "AI strategy" advice loses accounting firm owners. It conjures images of 40-page policy documents, IT departments, and implementation consultants billing $400 an hour. That's not what you need.
A defined AI strategy for a 5–40 person CPA firm is a one-page workflow map with three components:
1. Tool inventory — Which tools is your firm using, for which workflows, and who owns each?
This doesn't need to be exhaustive on day one. Start with the tools your team is already using. Map each one to a specific workflow: "We use [Tool X] for [this task], owned by [this role]." That's it. If you have eight tools and can't answer this for each one, that's exactly the gap you're closing.
2. ROI tracking — For each tool in your inventory, what does success look like?
Time saved per week is the simplest metric. Error rate and output quality per workflow are more sophisticated. You don't need all three at the start — but you need at least one per tool. If you can't measure it, you can't manage it. Thomson Reuters Institute found in 2026 that 40% of professional services firms have adopted AI tools, but only 18% measure ROI. That's the same gap in different data.
3. Decision rules — When does AI output require human review before client delivery?
This is the risk management component. For a tax return: which sections can AI draft without review? For a client advisory memo: what's the review protocol? For a research summary: who checks the citations? The goal isn't to review everything — that defeats the efficiency gain. The goal is to know, in advance, where human judgment is required.
The 3-Workflow Minimum to Start
You don't need to map all eight tools at once. You need three.
Pick the three highest-frequency workflows in your firm — the tasks your team does most often. Map each one: which tool is used, what the output is, who owns the workflow, and where human review is required before client delivery.
Three workflows mapped correctly is a strategy. Eight tools with no map is not.
Build the three-workflow map first. The rest of the tools will follow the same template once the pattern is established.
The 93% Advisory Signal: What Your Firm Needs to Survive the Transition
The jump from 83% to 93% of firms offering advisory services in one year is not an accident. Compliance work — tax prep, bookkeeping, standard auditing — is being compressed on both time and price by AI tools that can do it faster and cheaper. Firms that want to protect their revenue are moving into advisory: CFO services, financial planning, strategic analysis, client coaching.
That's the right move. But it creates a specific problem that the 22% strategy gap makes urgent.
Advisory work is relationship-intensive and judgment-intensive. When you deliver an advisory recommendation to a client, they're trusting your expertise — your interpretation, your judgment, your accountability. If AI is helping you form that recommendation (and it should be), you need to know exactly where AI contributed and where your judgment took over.
Without a defined AI strategy, you don't have that documentation. You have eight tools and no map.
The liability risk is real. The reputational risk is real. And the competitive risk is real: firms with defined AI workflows for advisory work will be able to price it more confidently, deliver it more consistently, and scale it more efficiently than firms that are winging it.
The firms currently in the 22% are already building that advantage. The other 78% are feeling the consequences of the gap — which is exactly what 75% of accountants are describing when they say workflow complexity is slowing them down.
One Action: Build Your AI Workflow Map in 30 Minutes
Don't wait until you have a full strategy to start. Start with the map.
Here's the template. For each of your top three workflows:
| Workflow | Tool Used | Time Saved/Week | Human Review Step | Owner |
|---|---|---|---|---|
| Client email drafts | [Tool] | ~2 hrs | Partner review for advisory clients | Staff A |
| Tax research | [Tool] | ~3 hrs | Senior accountant verification of citations | Staff B |
| Invoice generation | [Tool] | ~1 hr | Office manager approval before send | Staff C |
That's the entire structure. Fill it in for your three highest-frequency workflows this week. You don't need software, a consultant, or a policy document. You need a table and 30 minutes.
Once you've done this for three workflows, you have a working AI strategy. You know which tools are being used, what value they're delivering, and where human oversight is required. That's the 22% — and that's the crossing.
This week: Open a Google Doc or a spreadsheet. Create three rows — one per workflow. Fill in all five columns. Share it with your team. That document is your AI strategy. It replaces "we use some AI tools" with "here's our workflow architecture." The difference is enormous.
For more on measuring what AI is actually delivering to your firm, see Measuring AI ROI in Professional Services. If you're evaluating AI platforms specifically for accounting workflows, Accrual vs Basis: AI Accounting Platforms for Small CPA Firms has the comparison you need.
Source: Progress Software AI in Accounting Report, May 12, 2026. Thomson Reuters Institute 2026 AI Adoption Study.
The Crossing Report covers practical AI intelligence for professional services firm owners every week. Subscribe at crossing.one.
Related Reading
Frequently Asked Questions
What percentage of accounting firms have a defined AI strategy?
Only 22% of accounting firms had a defined AI strategy as of May 2026, according to Progress Software's AI in Accounting Report. This despite 84% of accountants already using AI tools — a 62-point gap between adoption and strategy.
Why do most accounting firms use AI without a strategy?
Most firms adopted AI tools reactively — one tool for invoicing, another for research, another for drafting — without a plan for how they connect. The result is fragmentation: 75% of accountants say too many workflow steps slow them down, and the average firm now uses 8 different AI tools.
What does an AI strategy for a small CPA firm actually include?
At minimum: a tool inventory (which tools are used for which workflows), a ROI tracking method (time saved, error rate, output quality per workflow), and decision rules for when AI output requires human review before client delivery.
How many AI tools does the average accounting firm use?
The average accounting firm uses 8 AI tools, according to Progress Software's 2026 report. That number reflects reactive adoption rather than intentional strategy — adding tools to solve individual problems without building a connected workflow architecture.
What is the connection between AI strategy and advisory services in accounting?
93% of accounting firms now offer advisory services (up from 83% in 2025), according to Progress Software. Firms moving into advisory must document how AI assists that work — for liability protection, pricing justification, and competitive differentiation. A defined AI strategy is the foundation for sustainable advisory expansion.
Get the weekly briefing
AI adoption intelligence for accounting, law, and consulting firms. Free to start.
Related Reading
This is the kind of intelligence premium subscribers get every week.
Deep analysis, cross-sector patterns, and the frameworks that help professional services firms make the crossing.