Four Accounting Firms That Rebuilt on AI — What They Actually Changed

Published: March 17, 2026 | By: The Crossing Report | 6 min read

---

Summary

CPA Trendlines published "Four Real-World CAS Transformations" in March 2026, documenting named accounting firms that rebuilt their client advisory service practices on AI — with real numbers, not hypothetical projections. This is the "what does it actually look like?" answer that two years of AI hype never quite gave us.

---

The Gap Between Advice and Evidence

For the past two years, the message to accounting firm owners has been consistent: AI will transform your practice, CAS is the future, advisory work is where the growth is.

What's been missing is the specifics. Which firms? What exactly did they change? How long before results showed up in revenue?

CPA Trendlines closed that gap in March 2026. "Four Real-World CAS Transformations" names the firms, names the results, and describes the actual decisions — not the narrative arc of change, but the specific choices that preceded growth.

Here's what the case studies show.

---

Next Dimension Accounting: 200% Growth Without Adding Staff

Next Dimension Accounting, an Australian firm, grew 200% over two years without adding headcount. The mechanism wasn't hustle or lucky client acquisition — it was removing the constraint that limits growth at almost every small accounting firm: staff capacity.

Most 5-20 person accounting firms hit a ceiling. You can't take on more clients because your staff is fully booked on existing ones. Hiring is the standard answer — but hiring costs $70,000+ annually per position, carries risk, and takes months to produce a net-positive contribution. The result is that most small firms grow slowly, decline client work they could handle, or overextend their team.

Next Dimension's approach: automate the work that fills the hours, freeing staff time for higher-value advisory engagement. Bookkeeping automation, document classification, client follow-up workflows — the tasks that collectively consume 40-60% of a small firm's staff time without generating proportional client value.

The 200% growth figure reflects what happens when you remove the capacity ceiling without adding headcount: you can say yes to more clients, more advisory work, and more complex engagements — with the same team.

What the owner actually changed: Not the staff, not the client base, not the service menu. The infrastructure supporting service delivery. The workflow underneath the work.

---

Armanino: 25,000 Transactions in Minutes

Armanino, a California-based firm, deployed an AI-powered 13-Week cash flow model that processes 25,000+ client transactions in minutes for real-time financial visibility.

This is worth sitting with for a moment. Twenty-five thousand transactions, analyzed for cash flow implications, in minutes. Before AI, that same analysis — pulling transaction data from multiple sources, classifying it, running the model, generating a projection — took analysts days. It was the kind of work that justified a large team, that required senior oversight, that you couldn't do for a small client because the economics didn't work.

Now a small firm can offer a version of it.

The 13-Week cash flow model is a specific advisory deliverable: a 13-week forward projection of cash position, updated with actual transaction data as it comes in. It answers the question most business owners carry in the back of their head every morning: am I going to run out of money? When is payroll a problem? Can I make this investment?

When that analysis is automated — when AI ingests the transactions, runs the model, surfaces the exceptions — the firm's contribution shifts to interpretation and recommendation. The client pays for the judgment call: what does this cash position mean for the decision they're about to make?

That judgment call is what advisory retainers are actually for. AI just made it possible to deliver it to clients who couldn't afford it before.

What Armanino actually changed: The unit economics of advisory work. By automating the data preparation, they made it possible to offer sophisticated cash flow analysis to clients who previously couldn't justify the cost.

---

The Pattern Across All Four Firms

The CPA Trendlines case studies aren't four isolated success stories. They're four examples of the same underlying shift.

In each case, the firm:

1. Identified the staff time that wasn't client-facing — bookkeeping, data entry, reconciliation, report preparation. The work that fills the hours between client conversations.
2. Automated it directly — not by buying a general AI tool and hoping for the best, but by mapping specific workflows to specific automation tools.
3. Reinvested the recovered hours into advisory engagement — more client contact, more proactive analysis, more of the interpretation work that clients value and can't get from software alone.
4. Repriced accordingly — not by raising hourly rates, but by shifting from compliance deliverables to advisory outcomes. The client pays for cash flow visibility and decision support, not for the hours spent building the model.

The revenue and growth numbers follow from that sequence. They're not the result of selling more of the same thing faster. They're the result of a different thing — advisory work that generates more value per engagement, with AI handling the data foundation underneath it.

---

What a 5-20 Person Firm Can Replicate This Week

You don't need Armanino's enterprise stack to start the same process.

The first decision is which staff time to reclaim. Map one week of your team's work. What tasks are repetitive, data-intensive, and low-judgment? Transaction classification, reconciliation, standard report generation, follow-up emails from meeting notes. That's your automation target list.

The first tool depends on where you are now:

• If your clients are on QuickBooks: QuickBooks AI handles transaction classification and cash flow basics. Activate it and run one client through it this month.
• If you're doing FP&A or cash flow planning: Datarails or Jirav gives you the automated model layer Armanino built, at a price point that works for a firm with 5-20 employees.
• If you're doing bookkeeping as a service: Botkeeper or a similar automation-first platform handles the classification layer.

The second step is changing what you deliver. Not immediately, not with a price increase next month — but with one client, one engagement, framed around the outcome rather than the deliverable. "We'll manage your cash flow visibility and flag when you need to make decisions" is a different conversation than "we'll deliver your monthly reconciliation and financial statements." The first is advisory. The second is compliance.

The firms in the CPA Trendlines study didn't transform overnight. They picked one workflow, automated it, reinvested the hours, and measured the result. The 200% growth at Next Dimension happened over two years, not two months.

You're two years behind them if you start now. You're further behind if you wait.

---

The Actual Question

Most accounting firm owners who read about Next Dimension or Armanino ask some version of: "That's a different kind of firm than mine."

They're right — it is. Different scale, different market, different resource base.

The question isn't whether those firms are identical to yours. The question is whether the underlying move — automate the data work, reinvest in advisory, reprice around outcomes — is available to you.

It is. The tools are there. The clients who pay advisory rates for advisory outcomes are there. The gap is the decision to start.

CPA Trendlines documented what starting looks like. The firms that acted two years ago are now reporting 200% growth. The firms that read about it today are deciding whether to start now or wait for more proof.

---

Source: CPA Trendlines, "Four Real-World CAS Transformations," March 15, 2026.

The Crossing Report covers AI adoption for professional services firm owners. Published every Monday.