The Abundance Playbook: AI and the Services Economy
Oct 12, 2026 · Vilksa Partners
A new playbook for the largest market on earth
The next generation of great companies will not sell AI to service businesses. They will run service businesses with AI at the core, and grow them faster than anyone thought possible.
Every major wave of value creation has come from pulling one layer of a business apart from another. Hardware companies learned to design in one place and manufacture in another. Software companies learned to deliver their product as a subscription instead of a box. Each split created a new kind of company and a new kind of investor.
The split happening now is bigger. Across the roughly $25 trillion global services economy (accounting, insurance, healthcare administration, logistics, property services and hundreds of niches in between), judgment and expertise have always been bundled with the people who deliver the work. AI is unbundling them. For the first time, the thinking part of a service business can scale like software while the human part focuses on relationships, trust and care.
Each era unbundled something
| Era | What got separated | What it created |
|---|---|---|
| 2000s | Design and IP from physical manufacturing | Asset-light hardware giants |
| 2010s | Software from on-premise installation | SaaS and the subscription economy |
| 2020s onward | Intelligence from service delivery | AI-native service companies that own the work, not just the tools |
The pattern is the same each time: the layer that can scale is pulled away from the layer that cannot, and margins move toward the scalable layer. What is different now is who is building it. The blueprint is coming as much from investors and operators who buy companies as from founders who sell software.
Three ways to build it
The smartest capital is not stopping at selling copilots into legacy firms. It is buying the firms, rebuilding them around intelligence and compounding from there. Three models stand out.
1. Model-led acquisitions
A team builds a best-in-class model for one service vertical, then buys established firms in that vertical, often from founders ready to retire, at modest multiples of revenue. Instead of grinding through enterprise sales, they own distribution from day one.
The experienced professionals in those firms become the engine of the model. Their judgment trains it, checks it and improves it. Better data makes a better model, which wins more work, which produces more data. A low-margin practice starts to carry software-like economics.
2. Platform buyouts
A vertical software company that already serves an industry takes a bigger swing and acquires large operators in that industry outright. Metropolis, which began as software for parking operators and then acquired SP+, is the clearest example. Owning the operations lets the platform deploy across hundreds of sites at once rather than one sales cycle at a time.
Once the technology is everywhere, a single data layer opens up pricing, utilization and customer insight that no individual operator could ever see on its own.
3. Tech-enabled rollups
Investors acquire fragmented, operations-heavy businesses (senior living, home services, specialty clinics) much like a traditional private equity firm would. The difference is what happens after close. Voice agents handle scheduling and inbound calls, workflow agents clear the back office, and coordination tools keep clinicians and families in sync.
Each new acquisition adds data and scale to a shared platform, so the hundredth location runs better than the first.
Beyond the copilot
Copilots make individuals faster; owning the work makes whole companies different. Tools like GitHub Copilot and Harvey can lift individual productivity substantially, but the firm buying them still bills, staffs and grows the same way it did before.
The larger prize sits in industries where the intelligence layer (underwriting, risk assessment, document review, coordination) can be cleanly separated from the human layer (relationships, judgment calls, care). Title insurance, insurance claims administration, tax and accounting, and healthcare revenue cycle all fit that description.
| Layer | Typical gross margin | Typical net margin | What it does |
|---|---|---|---|
| Intelligence | 60%+ | 30%+ | Proprietary models, compounding data, near-zero marginal cost |
| Software | 35% to 60% | 10% to 30% | Workflow automation and standardized process |
| Physical operations | Under 35% | Under 10% | Local presence and human relationships |
The winning companies will own all three layers, and let the top layer lift the economics of the bottom two.
Where most of private equity is getting it wrong
Most of private equity is treating AI as the next cost-cutting lever, and that is a mistake. The default playbook goes like this: buy the company, deploy automation, reduce headcount, expand margins, sell. It produces a better EBITDA line for a few years and a weaker business underneath it.
The problem is that cost cutting has a floor. Once the obvious roles are gone, there is nothing left to cut, and the company has shed the very people who understood its customers best. Morale drops, the best performers leave first, and growth stalls just as the next buyer starts asking where the upside is.
Viewing AI as a way to do the same work with fewer people misses what the technology actually changes. It does not just make work cheaper. It makes far more work possible.
Why we believe AI is a positive-sum game
We believe AI is a positive-sum game: it grows the pie for the company, its people and its customers at the same time. Used well, it is not a way to do the same with less. It is a way for everyone in the business to do more.
The loop only keeps turning if the last step happens. When the gains flow back to the people doing the work, they lean in, and the cycle accelerates.
Here is what that looks like in practice:
- Every person can do more. When AI takes on the repetitive work, a claims adjuster can handle more files, a bookkeeper can serve more clients and a care coordinator can spend more time with patients. Capacity expands without anyone burning out.
- The company can sell more. More capacity means saying yes to work that used to be turned away, entering new markets and launching services that were never economical before. Growth replaces cuts as the main source of value.
- Employees get paid more. When each person creates more value, the company can afford to pay them more, and should. We share the upside through raises, bonuses and equity, so the people closest to the work have a stake in making it better.
- Customers get better service. Faster turnaround, fewer errors and more attention from people who are no longer buried in paperwork.
- The work itself gets better. The tasks AI absorbs are usually the ones people like least. What remains is judgment, problem-solving and relationships, the parts of the job that drew people to the field in the first place.
This is not charity. It is the better investment thesis. A company that grows revenue per employee and keeps its best people compounds for decades; a company that only cuts runs out of road. The returns from growth are larger, more durable and far more valuable to the next owner than a one-time margin bump.
The bottom line
In the last decade, the winning move was to build software and sell it to businesses. In this one, it is to build intelligence and use it to run, and grow, the businesses themselves.
The firms that win will not be the ones that cut the deepest. They will be the ones that make every person in the company more capable, share what that creates, and let the flywheel compound. The services economy is not a cost to be squeezed. It is a $25 trillion opportunity to be expanded, and there is more than enough of it for everyone at the table.