Earlier this month, Sequoia Capital published "Services: The New Software" — an investment thesis arguing that the next trillion-dollar company won't sell you a tool, it'll do the work. They named accounting as one of the most attractive disruption targets in professional services.
Sequoia's argument is precise. For every dollar spent on software, six dollars are spent on services. A copilot sells the tool — it makes you faster. An autopilot sells the work — it sells directly to your client. That's where the venture capital is going.
They identify three conditions that make a service market ripe for autopilot disruption: the work is already done outside the building, there's an existing budget line, and the buyer is purchasing an outcome, not a process. Your CAS clients meet all three. They've accepted the work can be done externally — that's why you have a job. There's an existing budget line — it's your invoice. They're buying an outcome — accurate financials, filed returns.
And the line that frames the entire thesis: a company pays $10,000 a year for QuickBooks and $120,000 for an accountant to close the books. The next legendary company will just close the books.
What Sequoia sees
Sequoia's thesis rests on a clean separation: intelligence work and judgment work. Intelligence work is pattern-matching. Rules-based. Learnable. Encoding a CPA's playbooks into an autonomous system is technically straightforward — you've internalized those rules over years. Tax incentive identification, transfer pricing compliance, financial statement analysis against standardized playbooks, exception flagging, workpaper assembly. These are the tasks that justify the intelligence premium you charge.
Judgment work is different. It's the call you make on a Thursday night when your client's business partner just emptied the operating account and your gut tells you this marriage isn't ending well. It's reading the room during a meeting and knowing the real question isn't the one being asked. It's the wringable neck — someone's name on the file, someone who knows the client's family dynamics and business history. You can't encode that. Clients pay for it precisely because it can't be systematized.
The genuine judgment residual is real. But for most practitioners, that's 10 to 15 percent of the actual working week. The other 85 percent is intelligence work we've been calling judgment. And Sequoia's most important sentence: "Today's judgment will become tomorrow's intelligence." As AI systems accumulate data about what good judgment looks like, the frontier shifts. What required professional judgment in 2026 becomes learnable pattern recognition in 2028. The frontier is moving. The judgment zone is contracting. The window for restructuring your practice isn't as wide as it feels.
The four moves that win
Move one: Deliver compliance via AI — don't cede it. The instinct when you hear compliance is getting automated is to retreat to advisory. The right move is to become the one who delivers compliance via AI. You adopt the tools, run the automation, keep the client relationship. If a bookkeeping autopilot can close the books for 20 percent of what a human bookkeeper costs, there are two outcomes. Either the startup captures that client and keeps the margin, or you capture the efficiency and keep the client.
Move two: Price for outcomes, not hours. If AI does the bank rec in 30 seconds, hourly billing means you just billed for 30 seconds. Value-based pricing is no longer optional. A client who pays you $1,500 a month to close the books will leave for an autopilot at $300 a month. A client who pays you $1,500 a month because you caught a cash flow problem before it became a crisis, because you modeled three pricing scenarios that increased their margins by six points, because you called them before they knew there was a problem — that client isn't leaving.
Move three: Build the judgment layer now. If advisory is the defensible ground, you need to be standing on it before the flood hits, not scrambling toward it after. Four areas: financial reporting and insights, cash flow forecasting, profit improvement and pricing, and strategic advisory. These are defined services with a target client, a deliverable, and a price.
Move four: Reduce your exposure timeline. The models are good but not perfect. The startups are funded but not yet at scale. The market is warming but not yet hot. That window is your restructuring period. If you wait until the first client leaves for an autopilot, you're already behind.
The window is narrowing
The question is not if this happens. The question is when. And the answer is sooner than you think, but not overnight. Companies like Rillet, Basis, TaxGPT, and Skalar are funded, building now, backed by the same VCs who published this thesis. Sequoia isn't predicting the future. They're funding it.
And it's not just the startups. When the largest firm in the profession makes AI adoption a condition of partnership status, the signal travels down market fast. The disruptors are building from outside. The Big 4 are mandating from above. The window for mid-market firms to restructure on their own terms is the space between those two forces. And it's narrowing.
Sequoia just handed you the thesis. The execution is yours.