What is the #1 barrier to AI adoption for small businesses?

According to a March 2026 Bookipi survey of 2,121 small business owners, the #1 barrier is lack of expertise — cited by 31.2% of respondents. Cost ranked lower. This is consistent with what professional services firm owners report: the problem is not that AI is unaffordable, but that they don't know how to get started, which tools to use, or how to fit AI into their existing workflows.

Is AI adoption low in professional services because it's too expensive?

No. Cost is not the primary barrier. The Bookipi 2026 survey found that small business owners are more likely to cite lack of expertise, uncertainty about where to start, and unfamiliarity with AI tools than to cite cost as the reason they haven't adopted AI. For professional services firms, free or low-cost tools (Fathom, free tier; Claude.ai basic tier; many built-in Microsoft 365 features) are already available — the barrier is knowing which ones solve which problems.

Where is AI adoption lowest in professional services firms?

AI adoption is lowest in the operational workflows where it would create the most ROI: billing and time capture, client intake and communications, and financial reporting. Bookipi's 2026 data found only 16.4% of small businesses use AI for finance and 6.4% for HR — much lower than adoption rates for drafting and marketing tasks. Professional services firms show the same pattern: AI is common in document drafting and email, rare in billing workflows, client intake automation, and financial analysis.

How do small professional services firms fix the expertise gap?

The fix is sequence, not training. Start with the workflow that already has the highest cost of being wrong — meeting notes and client action items. Use a free tool (Fathom or Otter.ai) on three consecutive client calls. The expertise question answers itself once you see a tool perform a real task in your actual workflow. From there, the next workflow to automate is the one that takes the most time: typically billing entries, client update drafts, or document review prep. Each workflow you complete builds the internal expertise that makes the next one easier.

What percentage of professional services firms use AI daily?

Around 19% of accounting professionals and similar proportions in other professional services sectors use AI tools daily, according to 2026 AICPA/CIMA data. The majority have either tried AI and not made it a habit, or haven't started. The gap is not due to price — it's due to the expertise required to identify the right starting workflow, build the habit, and apply the tool consistently.

Why Small Firms Aren't Using AI (2026 Data)

Published: March 16, 2026 | By: The Crossing Report | 5 min read

Summary

A 2026 survey of 2,121 small business owners found that 31.2% cite lack of expertise — not cost — as their #1 barrier to AI adoption. Professional services firm owners face the same constraint: not "can we afford this?" but "do we know how to start?" Here's the honest diagnosis, and the fix that works.

The Real Barrier

Every conversation about AI adoption in small professional services firms eventually gets to cost. "We'll look at it when pricing comes down." "We don't have the budget for an enterprise tool." "Maybe next year when it's more affordable."

This is the wrong conversation.

A March 2026 Bookipi survey of 2,121 small business owners found that 31.2% cite lack of expertise as their #1 barrier to AI adoption — not cost. When the same group was asked what would most accelerate their AI adoption, "better guidance on how to use it" ranked above "lower prices."

The cost narrative is a more comfortable story to tell yourself than "I don't know how to start." But the data says the comfort is misplaced.

For professional services firm owners — accounting, law, consulting, staffing, marketing agencies — this finding maps directly onto what we observe: the firms that have adopted AI aren't meaningfully better-funded than the firms that haven't. They're better guided. Someone showed them a specific workflow, a specific tool, and a specific starting point.

Where AI Adoption Is Lowest — And Why

The Bookipi data contains a detail worth examining: only 16.4% of small businesses use AI for finance, and 6.4% for HR.

Compare that to adoption rates for drafting and content creation, which are substantially higher. Professional services firms mirror this pattern exactly:

AI for drafting email and documents: widespread, low barrier, easy to start
AI for billing and time capture: uncommon, despite direct revenue impact
AI for client intake and communications: rare, despite being the #1 source of dropped new client inquiries
AI for financial reporting and analysis: very low, despite 50–70% time reduction in tax prep documented in recent accounting studies

The work where AI creates the most measurable ROI — billing, intake, financial workflows — is exactly where expertise demand is highest. To use Fathom to record a meeting is simple. To build an AI workflow that turns meeting notes into structured billing entries requires knowing that the workflow exists, knowing which tools connect, and having enough confidence to run it on a real client matter.

That expertise gap is the actual barrier. Not cost.

The Pattern Across Firm Types

The expertise barrier shows up differently by firm type, but the root cause is the same:

Law firms: AI adoption is broadly documented (70% of legal professionals report using AI in some form), but it's concentrated in research assistance and document drafting. Using AI for client intake qualification, matter-level billing capture, or deadline management requires knowing how to connect the tools — and most small firm owners haven't been shown how.

Accounting firms: The AICPA/CIMA 2026 survey found only 19% of accounting professionals use AI daily. Of those not using AI daily, the most common reason given was not cost but "not knowing how to integrate it into existing workflows." The workflows with the highest return — tax prep automation, variance analysis, engagement letter drafting from meeting notes — require a setup step that most owners haven't taken.

Consulting firms: AI adoption is high for research synthesis and presentation drafting. Low for the work that builds repeatability: standardizing prompts, documenting client workflow templates, building client reporting formats that AI can fill in monthly. The firms doing that work aren't more sophisticated; they started with a documented workflow and iterated.

Staffing firms: AI adoption is concentrated in job description writing and basic candidate sourcing queries. The workflows that create competitive advantage — AI-assisted screening criteria, automated candidate briefings from call transcripts, client reporting from placement data — require more setup than "open ChatGPT and paste the job description." That setup step is the barrier.

Marketing agencies: Widely adopted for content drafting, less so for the operational layer: client reporting automation, briefing standardization, campaign analysis from actual data. The high-ROI workflows take more initial investment to define.

The Fix: Sequence, Not Training

The expertise gap is real, but it doesn't require a training program to close.

The firms that have closed it followed a specific pattern:

1. They started with the workflow that already had the highest cost of being wrong.

Not the most impressive AI use case — the most consequential one. For most professional services firms, that's meeting notes and action items. A missed client commitment from a call costs client relationships. An AI meeting tool (Fathom, Otter.ai) solves that in 30 minutes. No integration required. No training. One recorded call, one summary, one moment where the tool proves its value.

2. They let the first workflow create the second.

Meeting notes become the raw material for billing entries. Billing entries teach you what AI can do with structured data. Once you've seen AI turn a meeting summary into a billing entry draft, the jump to "AI drafts my client follow-up email from the same summary" is obvious. The expertise compounds.

3. They defined the human review point before running anything client-facing.

The firms with sustainable AI adoption have a clear rule: AI generates the draft, a human reviews before it goes anywhere. This isn't a legal disclaimer — it's the operational design. Without a defined review step, the first time an AI output has an error (and it will), the response is "we're pausing AI." With a defined review step, an error is a process improvement, not a crisis.

The Practical Starting Point

The expertise needed to start is less than you think. You don't need to understand how the models work. You don't need to know what "agentic AI" means or whether GPT-5.4 outperforms Claude Sonnet 4.6.

You need to know the answer to one question: which task on your team takes the most time per week and follows a predictable pattern?

For most professional services firms, the answer is one of:

Summarizing what happened in client meetings
Drafting routine client communications
Entering time for small tasks that get forgotten
Reviewing standard-form documents (contracts, NDAs, engagement letters)
Preparing monthly client reports from data that doesn't change format

Pick the one that applies to your firm. Find the tool that does that task. Run it on one real example this week.

The expertise gap closes itself once you've seen the tool work in your workflow. Every subsequent decision — which tool next, how much to automate, what to review — gets easier because you have a reference point.

The barrier is real. It's also a 30-minute problem to start solving.

Your Action Item This Week

By firm type — one specific thing to do before Friday:

Law firm: Open fathom.video, create a free account, and record your next client call. Read the summary before you write the follow-up email. Compare the time.

Accounting firm: Log into your Microsoft 365 or Google Workspace account and check what AI features are already active. You may have tools you haven't turned on. Run one client's monthly data through whatever AI analysis is available. Note what it produces.

Consulting firm: Take your last three client deliverables and upload them to Claude.ai. Ask: "What are the five patterns that appear most consistently across these documents?" You're testing AI, and simultaneously discovering your own intellectual property.

Staffing firm: Take the job description for your current most active search. Run it through a free AI tool (ChatGPT, Claude.ai). Ask: "What are the five non-negotiable qualifications for this role, and what's the most common reason a candidate fails at the interview stage?" Compare its output to your current screening criteria.

Marketing agency: Find your most recent monthly client performance report. Type the data into Claude or ChatGPT and ask it to write the executive summary. Compare what it produces to what you wrote. Note where it's right and where it missed context only you have. That gap is where your value lives.

None of these require a budget. All of them move you past the expertise barrier that's actually holding your firm back.

The #1 Reason Small Firms Aren't Using AI Has Nothing to Do With Money