What the Earnings Calls Are Really Saying About the Next Decade of Health Tech

Disclaimer: The thoughts below are my own and do not reflect those of my employer.

---

If you are interested in joining my generalist healthcare angel syndicate, reach out to trey@onhealthcare.tech or send me a DM. Accredited investors only.

---

Abstract

• This essay synthesizes raw earnings call transcripts and management commentary from large healthcare incumbents and scaled consumer health companies.

• It focuses on what these executives and analysts reveal unintentionally about incentives, constraints, and where capital will actually flow.

• The goal is to help health tech angel investors, current or aspiring, think more clearly about early stage risk, timing, and startup selection.

• The tone is casual but the content assumes deep familiarity with healthcare economics, enterprise selling, and venture dynamics.

Table of Contents

1. Why earnings calls are the most underused diligence artifact in health tech

2. The end of AI as a product and the rise of AI as infrastructure

3. What payers keep telling us, even when they do not mean to

4. Providers are rational actors trapped in bad math

5. Scale breaks everything: lessons from CVS and UnitedHealth

6. Hims and Hers and why verticalization is back

7. Startup archetypes that survive this environment

8. Failure modes that are becoming more common, not less

9. Portfolio construction for angels in a slower, harsher market

10. Closing thoughts on patience, power laws, and choosing your pain

Why earnings calls are the most underused diligence artifact in health tech

Most early stage health tech investing starts in the same places. Founder intros. Demo days. Warm decks. A convincing story about why this time is different. Very little of it starts with listening to what the actual buyers are saying when they are forced to be honest. Earnings calls are not marketing. They are not recruiting pitches. They are not visionary keynotes. They are executives being interrogated by people who are paid to be skeptical, with legal and financial consequences if they get too cute with the truth.

That is why they matter so much. Earnings calls compress incentives. They surface tradeoffs. They reveal where leadership teams are under pressure and where they still have room to experiment. If you want to understand what kinds of startups will get bought, renewed, expanded, or quietly sunset, you could do a lot worse than reading transcripts from the last few quarters of payer, provider, and scaled health tech companies.

What strikes me every time I go back to these transcripts is how consistent they are. Different companies, different business models, same themes. Cost pressure. Labor pain. Administrative drag. Regulatory uncertainty. The language changes slightly, but the math does not. For an angel investor, this consistency is a gift. It means you can pattern match. It means you can separate structural demand from temporary hype.

There is also a humility that comes from this exercise. You realize how little patience the system has for novelty that does not translate into margin or risk reduction. You also realize how slowly even good ideas move once they hit enterprise reality. If you are going to invest early, you need to be comfortable living in that tension.

The end of AI as a product and the rise of AI as infrastructure

A few years ago, every earnings call had a section where management felt obligated to say the word AI. It was usually vague. Something about leveraging advanced analytics or exploring machine learning to enhance the customer experience. That phase is over. What has replaced it is much more interesting and much more dangerous for founders.

Today, when AI shows up in earnings calls, it is almost always attached to a very specific operational outcome. Reduced call handle time. Higher claims auto adjudication rates. Faster prior authorization decisions. Lower administrative cost per member. Nobody is applauding. Nobody is excited. This is AI as plumbing.

For startups, this changes the game. Buyers no longer want tools that demonstrate intelligence. They want systems that disappear into workflows and quietly make something cheaper or faster. They want reliability, auditability, and integration. They want to know what breaks when the model is wrong and who is liable.

For angels, this is a warning sign and an opportunity. The warning is that many AI first pitches will not survive first contact with procurement. The opportunity is that truly useful infrastructure companies will look boring early and compound quietly over time. The trick is being able to tell the difference before the revenue shows up.

What payers keep telling us, even when they do not mean to