I Watched This Coming for Twenty Years
CBIZ is gone. Not dead — swallowed. Grant Thornton Advisors, backed by $5.2 billion in New Mountain Capital financing, has agreed to acquire CBIZ at $55 per share in cash. The publicly traded shares disappear. The independent corporate identity disappears. What remains is a wholly owned subsidiary inside a private-equity-backed professional-services conglomerate. The post going around LinkedIn has a tombstone with the inscription: "Acquired. Integrated. Forgotten."
It's melodramatic. But it is not wrong.
I have a different relationship to this news than most people posting about it. I did not read about CBIZ in a deal announcement. I worked there. Before that, at Marcum. Before that, at Rachlin Cohen & Holtz. My professional timeline is, almost exactly, a miniature history of the consolidation machine that just reached its logical conclusion.
I started in IT at Rachlin — a regional firm, relationship-driven, the kind of place where the partners actually knew the clients. Technology was infrastructure. Invisible unless it broke. Then Rachlin merged into Marcum. I stayed through the transition, moved from IT into infrastructure, and eventually into digital forensics. That progression tells the whole story in miniature. The firm was becoming larger, more sophisticated, more centralized, more capable of selling technology-adjacent services. IT stopped being a back-office function and started becoming something the firm could monetize.
Then Marcum merged with CBIZ. National scale. Publicly traded. Real institutional machinery. And now CBIZ becomes a subsidiary of a vehicle backed by private equity.
Regional partnership. National firm. Public corporate platform. Private-equity-backed conglomerate.
I spent over a decade inside that organism as it evolved through the middle stages. What I watched was not accounting consolidation in the abstract. It was a live experiment in what happens when institutional capital discovers that professional services look a lot like a product.
Here is what those deals actually look like from the inside.
When two firms merge, someone has to make all the incompatible systems behave like one company. Different networks, different identity systems, different practice-management software, different file structures, different backup schemes, different partner hierarchies, different collections of undocumented exceptions that nobody wrote down because the person who built them left three years ago. That was my job. I was sent around the country to do it. And here is what I learned: mergers are not systems-engineering exercises. They are political exercises with systems consequences.
The buyer decides which platform is the enterprise standard. Not on the basis of which system is actually better. On the basis of which company did the acquiring. The acquiring firm's platform wins — full stop — because admitting otherwise would reverse the psychological hierarchy of the deal. So engineers spend a year migrating everyone onto an objectively inferior system because the integration plan is really an organizational power map dressed up as an IT strategy. Every time. Without exception.
I watched it happen at Rachlin. I watched it happen at Marcum. And whatever follows the CBIZ acquisition will go through the same ritual. The acquiring organization will insist everyone migrate onto its platform. The people who understand why that is a bad idea will say so clearly and be ignored. Then everyone will spend 18 months suffering through it, and the institutional memory of every previous migration will quietly disappear.
This is the part nobody captures in the deal announcement.
Now here is the part that should make people in this industry uncomfortable.
The deal makes sense. It makes sense precisely because of where AI is going. Private equity sees the accounting industry and finds: recurring revenue, mandatory or quasi-mandatory services, aging partners who need liquidity, thousands of fragmented firms, duplicated back offices, labor-intensive workflows, client relationships that are difficult to recreate, and enormous potential for automation and centralization.
The thesis is not simply "let us buy accounting revenue." It is closer to: "let us acquire the client relationships, licenses, distribution, and recurring work before automation dramatically changes the economics."
You acquire 20 firms with separate administrative teams, software contracts, research subscriptions, document-processing operations, and internal IT functions. You consolidate the infrastructure, standardize the data, and deploy AI across the entire platform. Without increasing revenue, the operating leverage becomes enormous. That is the trade. The CBIZ transaction positions that trade at $5.2 billion scale.
But the larger problem is not who owns the firms. It is what AI does to the model that has supported them for 50 years.
The traditional accounting firm runs on a pyramid. At the bottom: associates gathering documents, reconciling accounts, populating workpapers, preparing forms, reviewing contracts against templates, formatting reports. In the middle: senior associates and managers reviewing that work, resolving exceptions, researching technical questions, supervising staff. At the top: partners exercising judgment, signing opinions, owning client relationships, accepting liability. The pyramid works because you can place large numbers of moderately trained people at the bottom and a small number of highly experienced people at the top. AI is extraordinarily well suited to most of the bottom and a meaningful chunk of the middle.
One experienced person with a competent AI system may soon perform work that previously required several associates and a senior. That breaks the pyramid.
And once the pyramid breaks, the firms inherit a second problem nobody is talking about clearly: where do tomorrow's managers and partners come from?
People historically developed judgment by spending years doing the repetitive work. They saw hundreds of returns, reconciliations, controls, exceptions, and workpapers. Much of it was miserable. But it built the pattern recognition that made them useful at the top. AI can eliminate the drudgery while simultaneously eliminating the apprenticeship environment. You cannot automate every foundational task and then expect a 28-year-old to mysteriously possess the judgment that used to emerge from 10,000 hours of reviewing the underlying material. That is the structural problem the industry does not have a good answer to yet.
What survives is not the most intellectually complicated work. It is the work attached to trust, accountability, and legal authority.
Audit opinions still require licensed professionals who accept responsibility for the conclusion. That is fundamentally different from paying someone to populate 40 spreadsheet tabs. Regulators are watching AI's role in audit work carefully — their focus is quality, evidence, supervision, and professional skepticism, not replacement. Clients will keep paying for someone who can say: "We examined this, we stand behind it, and we accept the consequences."
The durable functions: regulatory sign-off, complex judgment, contested valuations, investigations, forensic work, negotiation with tax authorities, litigation support, transaction structuring, high-stakes advisory, executive relationships, accountability when something goes wrong.
That last item is the one people overlook. Accountability is the product. The deliverable is not the workpaper. It is the professional who stands behind the conclusion when the IRS calls or the lawsuit arrives. My forensics work lived on that side of the line. A forensic matter is not "find the numbers." It is provenance, evidentiary integrity, human behavior, incomplete records, competing narratives, chain of custody — and a conclusion that will be challenged by lawyers in a room. AI dramatically accelerates the mechanics. But someone credible still has to own the methodology and defend the result.
There is one more thing worth saying, and it is the part the consolidators may be underestimating. AI gives small firms leverage too. A five-person specialist practice may soon possess research, document review, workflow automation, and analytical capabilities that once required 50 people. So the market could polarize: enormous platforms winning through scale, data, brand, capital, and regulatory infrastructure — and small elite firms winning through specialization, personal relationships, and extreme automation.
The middle gets squeezed. Ordinary midsized firms will have neither the intimacy of a boutique nor the economics of a platform. That may be the next round of consolidation — not PE buying big firms, but big firms and small specialists dividing the wreckage of every firm that was too large to be personal and too small to be a platform.
And the irony is not lost on me. When I was at Rachlin, the network supported the accounting firm. In the next version, the computational platform is the accounting firm — and the licensed professionals provide supervision, judgment, legitimacy, and relationships around it. The function I spent years treating as infrastructure is becoming the means of production.
The CBIZ tombstone says "Acquired. Integrated. Forgotten."
I have watched the same thing happen to Rachlin's name. To Marcum's independence. Now to CBIZ's public identity. The clients, the revenue, and many of the people remain. What disappears is the institutional memory — the reason certain decisions were made, the workarounds that exist because of something that happened in 2009, the relationships that are held inside the heads of people who were there for all of it.
Those people eventually leave. And the next team inherits a set of systems nobody fully understands, maintained by processes nobody fully remembers, serving clients who vaguely recall when things were different. I have lived through that transition three times. The fourth wave is not a merger. It is AI automating the labor that held the whole structure together. That one is going to be harder to integrate.
Rich Washburn is a technologist and strategist working at the intersection of AI, infrastructure, and capital. He is Managing Partner and Chief AI Officer at Eliakim Capital and CIO of Data Power Supply.
Comments