The Compliance Extinction Event

I’ve been (doom?) scrolling LinkedIn this week and the pattern is impossible to ignore. Post after post from reputed experts in accounting and finance, all making variations of the same claim: AI is going to transform your practice. Adapt or die. The future is here.

And nearly every one of these posts is itself clearly written by AI. Staccato sentences. Surface-level ideas with no exploration. No evidence that the author has actually sat with the problem they’re describing, or inserted any original thinking into the argument. They’ve had a fragment of an idea, fed it into a language model, and published whatever came out the other end. And we’re supposed to engage with this as expertise.

It’s disturbing, but it’s also revealing. Because what you’re actually seeing is an industry that has been through five years of broken promises, and the people who are supposed to be leading the conversation have stopped doing the work of thinking. They’re recycling the same message — “add advisory, embrace AI, transform your practice” — using the very tools they’re warning about, without any of the nuance or specificity that would make the advice actually useful.

This is what happens in a market that’s battle-scarred. Not skeptical. Battle-scarred. Webinars titled “The Ultimate Advisory Toolkit” that turn out to be three generic tips and a sales pitch. Platforms that tell you to “add advisory” while building automation that competes directly with the work you bill for. Bench — a venture-backed bookkeeping startup with slick marketing and bold promises — collapsing overnight in late 2024 and leaving twelve thousand businesses scrambling. After enough of that, the rational response is to stop trusting any voice telling you to change.

And that’s the real danger. Not AI. Not offshore competition. Not platform capture. Those are forces, and they’re real, and I’ll get into each of them below.

This newsletter is not going to make promises. Every Thursday, I’m going to share frameworks, observations, and the occasional tool recommendation from ten years of running a vCFO practice and from acquiring, scaling, and selling a bookkeeping firm. Things I’ve actually done. Not things I’ve read about and repackaged.

I should be upfront about one thing: I’m also building something called Baifokal, which is an advisory delivery system for bookkeepers, fractional CFOs, and CPA firms. You’ll see links at the bottom if you’re curious. But this newsletter exists to be useful whether you ever look at it or not — the goal is to lift the conversation in this industry, and if that leads some of you to explore what we’re building, that’s great, but it’s not the price of admission.

The business model that most bookkeepers, fractional CFOs, and CPA firms have built their practices on is being dismantled. Not by one force, but by several forces compounding at the same time. And the pace is faster than most people inside it are willing to acknowledge.

I say this from inside it. My wife Kristina and I have been running a virtual CFO practice for ten years — we’re still running it today. In 2021, during COVID, we acquired a small bookkeeping firm. Over the next two and a half years, we more than doubled the staff, the clients, the revenue, the profit, and the enterprise value. We sold it in late 2023.

We did everything right by every measure the industry uses — accurate books, responsive service, competitive pricing — and the whole time, I watched the business model slowly collapse under market forces that had nothing to do with the quality of our work.

If you’re making most of your revenue from compliance work — and by that I mean the monthly and periodic recording, reconciling, and reporting of past financial data — then what I’m about to describe will probably feel familiar, regardless of whether you’re a solo fractional CFO managing fifteen clients, a firm owner with a team of bookkeepers, or a CPA practice partner wondering how to extend your annual tax relationships into something more than a once-a-year touchpoint.

The threat isn’t coming from one direction. It’s coming from four forces at once, and they compound each other.

The first is platform capture. This is the one I’ve watched most closely, because it’s the most cynical. QuickBooks and Xero built their dominance on the back of the practitioner channel. When Intuit launched QuickBooks in 1992, they went straight to end-users and bypassed accountants entirely. Practitioners pushed back. So Intuit pivoted — launched ProAdvisor in 1996, invested in certification, built professional features, repaired the relationship deliberately. By 2008 they had 94% market share and over fifty thousand enrolled practitioners recommending QuickBooks to every client. You became their distribution network. That was the goal.

Now that goal has been achieved — 600,000 ProAdvisors globally — the strategic calculus has shifted. They don’t need the practitioner channel the same way. QuickBooks Live competes directly with bookkeepers for clients. Xero’s JAX automates reconciliation work that firms bill for monthly. You’re not a partner in this arrangement. You’re a distribution channel they’ve outgrown.

The second is AI — but not in the way most LinkedIn posts are framing it. The simplistic message is “AI is coming for your job.” The reality is more specific. AI is making DIY bookkeeping feel accessible to business owners who would have hired someone three years ago. It’s lowering the threshold for “good enough.” Every time that threshold drops, the addressable market for pure compliance work shrinks. It’s not that AI replaces bookkeepers overnight. It’s that it shifts client perception of what compliance work is worth. And that perception shift, once it happens, doesn’t reverse.

The third is offshore arbitrage. Competent bookkeepers in the Philippines and India doing comparable compliance work for fifteen to twenty-five dollars an hour, against your seventy-five to a hundred and fifty. This isn’t a quality argument. In many cases the output is largely indistinguishable. For firm owners carrying fixed overhead, this compression is existential. For solo fractional CFOs, it’s a slower squeeze. For CPA practices, it raises the question of whether building an internal bookkeeping team even makes economic sense anymore.

And the fourth is commoditization — the cumulative result of the first three. Clients no longer see compliance work as something they value. You see it in the client who treats your monthly reconciliation as background noise. The unopened email with the P&L attached. The meeting where they say “looks good, thanks” without asking a single question about what the numbers mean.

These four forces feed each other. Platform capture accelerates AI adoption. AI accelerates the perception of commoditization. Offshore competition pushes pricing down. And you cannot fight market-level forces with individual effort. The platforms have billions in capital. AI is improving on a curve that doesn’t care about your pricing model. Offshore labor arbitrage is structural. Client perception, once shifted, is almost impossible to unshift.

The only viable response is evolution. Move upstream to advisory before the wave fully arrives. Not by becoming a different person — but by changing where you spend your time, and how you structure what you deliver.

Next week, I’ll share a concrete framework for deciding what to automate first and what to keep doing by hand — and why getting that sequence wrong is the trap that catches most people attempting this transition.

When I say automation in this newsletter, I’m not talking about handing your client work to ChatGPT and hoping for the best.

Most of what needs automating in compliance work is deterministic — it follows rules. Reconciliation matching, transaction categorization, variance calculations, report formatting. These are logic tasks. The output is the same every time because the rules don’t change. This is formula-based, algorithm-driven automation.

AI — the probabilistic kind, the large language models — has a role too, but it’s a different role with different guardrails. AI is useful for communication and narrative: turning accurate numbers into plain-English commentary that a client will actually read. But it should never be trusted with the numbers themselves, because probabilistic systems don’t guarantee the same output twice.

This distinction matters enormously. Your professional reputation is built on accuracy. Deterministic automation protects that. Probabilistic AI, used without proper architecture, threatens it. We’ll dig much deeper into this in a future issue.

I dictate a lot of my content — LinkedIn posts, client notes, even parts of this newsletter. Wispr Flow is a voice-to-text tool that runs locally on your machine and works across any application. It’s fast, it’s accurate, and it’s changed how I capture ideas because I’m no longer trying to type at the speed I think.

Full disclosure: I don’t get paid for this. If you sign up using the link below, you’ll get one free month. If you end up sticking with it, I’ll get a free month too. That’s the extent of it.

Try it: Get one free month of Wispr Flow →

The Automation Triage Framework — a concrete method for deciding what to automate first, what to automate later, and what to keep doing by hand because it’s actually the high-value work you want to be doing more of. It includes a prioritization formula that tells you exactly where to start and what each hour of automation is worth.

If you’re in the middle of this transition right now, I’d like to hear where you’re stuck. Hit reply. I read every response, and the conversations that come out of these emails are often more useful than the essays themselves.

Join the Waitlist  •  Schedule a Demo

Know someone navigating the compliance-to-advisory transition? Forward this email — or better yet, send them to baifokal.beehiiv.com to subscribe.