Skip to main content

Service as Software

The era in which a services business was structurally capped at 30–50% margins because growth required adding headcount is ending. AI is breaking that link — and the investors who see it are buying Main Street businesses, not building apps.

Software as a Service (SaaS) was the defining business model of the last two decades: build software once, sell it endlessly, and run margins of 70–90% because you never had to add a body every time you added a customer. Service businesses never had that luxury. They ran on human time — answering emails, processing invoices, scheduling appointments, managing disputes — and every new client meant more cost.

Service as software inverts the model. The term was coined by Phil Fersht of HFS Research in 2024 and has since been taken up across venture capital and enterprise strategy — most prominently by Alex Rampell at a16z (who frames it as "software is eating labor," pointing at the $13 trillion US labor market as the next target after the $300 billion SaaS market) and by Madhu Namburi of General Catalyst, who brought it to mainstream business media in a June 2026 CNBC segment: "SaaS was the definition — software as a service. I believe it's going to be service as software. And I think that era has come."

The sharpest formulation comes from Wipro Ventures: if SaaS gives you a calculator, service as software delivers the accountant. The bet is that AI finally breaks the constraint that kept services businesses structurally less valuable than software — and that the companies that rebuild around AI rather than just layering it on top are the ones whose economics change. HFS Research estimates this shift could represent a $1.5 trillion market by 2035; Foundation Capital puts the addressable labor spend at $4.6 trillion.

The Mechanism​

The AI rollup is the operational form the bet takes. General Catalyst, Thrive Capital, Lightspeed, and Andreessen Horowitz are all running versions of it. The playbook: create a holding company, pour in capital, buy up services businesses in a specific vertical — property management, HOA management, corporate travel, accounting, insurance — and then rebuild them around a purpose-built AI platform.

The canonical example is Long Lake, a three-year-old holding company backed by General Catalyst and Alpha Wave. Long Lake has acquired over 30 businesses spanning HOA management, construction, and corporate travel. Its AI platform, Nexus, is purpose-built for the workloads of each industry — specialized workflows, tools, and skills tuned for those specific businesses. Long Lake's CEO, Alex Tobman, says Nexus performs five times better than off-the-shelf models like ChatGPT and Claude for those contexts. That gap is the point. The value is not in the model — it is in who owns the workflow.

Long Lake plans to hold these companies permanently, more like Berkshire Hathaway than a traditional PE fund. Engineers — many from Ramp and Palantir, companies with cultures of putting technical people close to the customer — are embedded inside the businesses, not dropped in from outside.

Why This Matters More Than Enterprise Software​

The timing of the AI rollup wave is not accidental. In the early 2020s, the largest PE firms made a big bet on enterprise software: Vista bought Citrix, Thoma Bravo bought Anaplan and Coupa, Silver Lake bought Qualtrics. The thesis was that recurring software revenue was the safest cash flow in business. Then AI arrived, and that safe software revenue looked less safe — because AI could increasingly substitute for the workflows those tools supported.

Now those same firms are trying to retrofit AI into their software portfolios through OpenAI and Anthropic partnerships. CNBC's framing is blunt: "That can look like a consultant's fix — someone else's AI dropped into someone else's company by investors who don't really own the technology or the workflow." The AI rollup is a different playbook entirely. Don't just buy the software companies AI might disrupt. Buy the services companies AI might transform.

What This Means for a Services Business Owner​

If you own a services business — a property management firm, an accounting practice, a staffing agency, a facilities company, a corporate travel operation — you are now in the crosshairs of both traditional PE buyers and a new class of AI rollup vehicles. The difference in what they pay and what they are looking for is significant.

A traditional PE buyer prices you on your current margin structure and applies a multiple to EBITDA. An AI rollup buyer is pricing you on what your margin structure could look like with AI embedded in the operation — a different number entirely, and a higher one if the story is credible.

The thing that makes the story credible is workflow ownership: AI that is embedded in how the work actually gets done, not bolted on as an optional tool. A team that has built custom workflows for their specific business, documented processes that AI acts on, and demonstrated margin expansion — even incrementally — tells a fundamentally different due diligence story than a team that bought a ChatGPT Enterprise license.

The critique of the consultant's fix applies equally to owners who adopt AI cosmetically. Adding a subscription to a general-purpose AI tool says you have AI access. Rebuilding how a function operates around AI says you own an AI-enabled workflow. Buyers know the difference. So does your P&L.

The Structural Shift in Acquirer Appetite​

The implication is that the pool of potential buyers for your services business is expanding, and the most aggressive buyers in that pool are paying premiums for businesses that have started the transformation. General Catalyst created roughly a dozen AI rollup holding companies in three years and has $1.5B+ targeted at this strategy. Their appetite is for services businesses whose cost structure is about to change — which means businesses that have already begun changing it are more attractive, not less.

The public markets still see boring slow-growth services businesses. The AI rollup sees something different: a cost structure that may be on the verge of changing. If you are building toward an exit, the question is which picture your business tells when a sophisticated buyer looks closely at how decisions get made, how work gets done, and where margin might expand.

That question is exactly what the Agentic Business Blueprint is designed to answer before you go to market — and it is why the AI valuation premium is not hypothetical.

Further Reading​

Source: CNBC, "The AI Rollup Wave Transforming Main Street" (June 8, 2026). Features Madhu Namburi (General Catalyst) and Alex Tobman (Long Lake).