Capital Rotation Inside the Health Value Chain

It’s pretty difficult to navigate headlines and conversations in health without following the money and understanding the magnitude of decisions and sometimes the inertia. If you had a chance to read our thesis, you would know Penultimate Capital is intent on understanding the healthcare value chain of tomorrow and how we invest in it today.

A few things continue to stick. 6-3 (not to be confused with 6-7 😄). $63 billion of capital spend sits inside health IT budgets. In most industries, software is the lever. In healthcare, labor still carries the weight, which means the real leverage isn’t always IT at all, it’s in the provider services layer where the spend actually lives.

Hold the 📞 because this mismatch presents an immense opportunity in the system. Reflecting on my investments in the tech-enabled wave of care, think virtual primary care, async and remote tools. That entire category captured ~42% of all healthtech dollars in 2021. Today it is ~9%, barely $1B so far in 2025.

The drop is structural. Payments in this segment are complicated and margins are thin. During this period, many founders who pitched me struggled to articulate how their revenue model actually worked. Basic questions often went unanswered. What was true care margin. How did it expand over time. How were human costs controlled. Layer on the hard limits of clinician hiring, state licensing and payer enrollments (kudos to Harbera for attacking this head on). Most of the underlying tech became table stakes, and the category simply could not scale beyond its labor constraints, but the jury is still out.

Capital rotated. It followed the operational pain.

Since 2023, provider ops has taken over with 44% of all healthtech investment. Scheduling, outreach, analytics, RCM, prior auth, patient access. These workflows sit on top of more than $700B in provider services that have never been touched by real intelligence. This is where the cost structure lives, which is why capital naturally shifted here once alt care momentarily topped out. 👀

There’s also a safety dimension behind this shift. Although medical devices are not directly a part of tech enabled services, the hardware side of the market offers a useful heuristic. As called out by SVB, nearly 50% of all AI enabled device recalls happen within the first year of clearance, which is twice the rate of devices overall. Hospitals have been primed and grown weary, which is why they have gravitated toward lower-risk AI tools first. Scheduling assistants. RCM automation. Front door workflows that add efficiency without touching clinical decision making.

And this is where a Gartner Hype Cycle lens helps perspective. When you look at the curve through the frame of capital rotation and the services value chain, a few simple patterns start to emerge. Early ideas cluster where expectations run ahead of evidence. The workflows that actually scale sit further along. The curve mirrors where capital has moved, effectively toward the parts of the chain that have convinced stakeholders to buy.

Early tech, like ambient intelligence and digital twins still sits in the hype zone, where expectations might run ahead of critical adoption and proof. The real momentum that captures budgets is in mid curve workflows such as workforce management and process automation, which match the rotation into provider ops because they deliver clear ROI and low risk. Together, the pattern mirrors the broader capital shift in that frontier tech makes the headlines, while mature operational and infrastructure layers capture the deeper spend.

On the flip side, the procurement process for intelligence related tools has compressed, but mostly in the parts of the value chain with low clinical risk. Health systems that once took years to approve new software are now deploying AI-driven tools in a matter of months when the use case delivers clear ROI, measurable margin impact and no real exposure to clinical decision making. According to Menlo Ventures, health systems went from 8 months to 6.6 months (~18% reduction) and outpatient providers from 6 months to 4.7 months (~22% reduction). The speed is real, but it is targeted.

That acceleration happens when the workflow sits firmly in the operational part of the value chain, not the clinical one. And that’s the bridge to the larger opportunity.

For the first time, it feels like intelligence can genuinely penetrate the provider services layer. Instead of selling into an IT silo, founders can augment labor and for the futurists out there, in some targeted cases even replace it.

I think a new generation of talent will build companies that convert human services into intelligent systems. These founders will go deep to understand manual workflows, conduct robust need finding, understand and reshape margin structures. They will rewire operating models. They will collapse cost centers that have lived untouched for decades.

Follow the labor.

That is where the next decade of healthcare value will be created.

Julian Eison
Founding Managing Parter

Sources