I have spent much of my career arguing that the business model of professional firms—especially hourly billing, timesheets, and the obsession with efficiency—is fundamentally incompatible with a knowledge economy. AI may finally force the issue.
That is why I enjoyed Rob Hamilton and Jeff Seibert's Zero Entry. The title describes a future in which essentially no accounting entries are created or touched by a human accountant. But the more interesting question is what happens after the entries disappear. Their answer is that accountants move from producing the books to standing behind them—from doer to "trust layer," providing judgment, interpretation, accountability, and ultimately taking responsibility for outcomes.
The authors are particularly strong on the economic consequences of AI. They understand that hourly billing is not merely an inconvenient pricing mechanism; it embeds a worldview in which value depends upon effort. As AI drives the effort of routine accounting work dramatically downward, charging for the inputs used to create an outcome becomes increasingly absurd (it always was, as it’s rooted in The Labor Theory of Value). Their warning is memorable: a future firm operating a future ledger on a past pricing model eventually collapses under the contradiction.
They really had me when they invoked Jevons Paradox. Increased efficiency does not necessarily reduce aggregate consumption. It can make something sufficiently cheaper and more accessible that demand explodes. AI may therefore produce more demand for certain forms of accounting expertise, not less—just as ATMs didn't simply eliminate the need for banking tellers but rather expanded the number of branches and tellers. That's a much more sophisticated argument than the usual "AI will destroy/save accounting jobs" binary.
But I part company with the authors when they suggest a firm that once served 200 clients might serve 600 with "deeper relationships." No, I don’t think so. Technology can scale information and transactions; relationships don't scale nearly so obediently. If accountants really aspire to be advisors, Dunbar [of Dunbar number fame, i.e., we can handle about 75-150 relationships] eventually gets a vote.
The discussion of outcome-based pricing is another highlight. Digits itself has put skin in the game: a client is considered automated when at least 95 percent of transactions are processed without human intervention, and if the platform doesn't achieve the specified outcome, the firm doesn't pay. That is genuine pricing innovation because it shifts risk from buyer to seller and aligns price with an agreed result rather than seats, credits, tokens, hours, or other inputs.
[AI companies seem to be turning to these latter strategies, which is cost-plus pricing (and hourly billing) in drag. Haven’t they learned anything from usage-based pricing, such as hourly billing, and what a lousy customer experience it is? And that cost doesn't determine price; value does. Price justify costs incurred]. We already have credits, they are called dollars, and token pricing is absurd, like paying the for number of grams (or calories) in your meal.
I am less persuaded by the book's claim that subscription pricing makes true partnerships "structurally impossible." That may be an interesting critique of SaaS pricing, but it doesn't survive contact with professional firms. Concierge and direct-primary-care physicians, among others, are flourishing under subscription models, which demonstrate that subscription relationships can deepen over time precisely because both parties escape the transactional meter and math of the moment for the more critical indicator of lifetime customer value.
The authors are also perceptive about intellectual capital, though they call it intellectual property, which is a much narrower concept. They warn that expertise residing solely inside senior people's heads walks out the door when those people do. I would quibble with their terminology: this isn't merely intellectual property. It is intellectual capital, much of it tacit. And AI doesn't eliminate the need to cultivate and capture that knowledge, which comprises a firm’s invisible balance sheet. If anything, it makes practices such as After Action Reviews more important, because firms need mechanisms for converting individual experience into institutional learning and knowledge.
One of the book's best warnings concerns what the authors call "truth drift": when AI intelligence and the underlying ledger live in different places, the AI's representation of events can gradually diverge from the system of record. Their advice—don't merely ask what today's software can do; ask what its architecture makes possible tomorrow—is excellent.
I also appreciated that this isn't 200 pages of disguised marketing collateral for Digits. The authors have a product to sell, obviously, but the book asks legitimate questions about the future of the profession: What becomes scarce when production becomes abundant? What happens to pricing when the cost to serve collapses? What happens when clients have access to much of the same intelligence as their accountants? And what exactly is the accountant responsible for when machines increasingly perform the accounting?
My largest reservation is that Zero Entry sometimes assumes that removing transactional work automatically elevates accountants into advisory work. It doesn't. Strategic advice, insight, creativity, and judgment, aren't the residue left behind after automation. They are distinct competencies that firms must deliberately develop. Calling yourself a "trusted advisor" has never made you one.
Still, Zero Entry is one of the better attempts I've read to think beyond AI as merely another efficiency tool for accounting firms. Its real subject isn't AI. It's business-model transformation. And on that, the authors and I agree emphatically: putting twenty-first-century technology inside a twentieth-century business model isn't transformation.
It's just a faster horse.
This entire review has been hidden because of spoilers.