Convicted or disciplined: How healthcare VCs are split on investing
Has AI changed venture math enough that ownership discipline matters less?
This piece concludes a two-part independent study I oversaw with Kyra Gardner and Andy Wang (both CBS ‘26).
In our last article, 24 founders told us what it actually feels like to fundraise inside the healthcare AI wave. The bar is moving every quarter. Investor consensus rewrites itself faster than founders can ship product. Every pitch comes back to some version of “are you one model release away from being obsolete?”
We closed Part 1 promising to flip the lens. Over the past month, we interviewed healthcare investors about what they are actually screening for, where they think the bar is heading, and which beliefs about the venture model they are quietly throwing out.
We covered a lot of ground. But one topic resurfaced again and again because it lies beneath almost every other conversation happening in venture right now.
Has AI changed venture math enough that investors should rethink ownership discipline?
For decades, ownership targets have been one of the few constants in venture capital. But AI is enabling companies to grow faster, raise at higher valuations, and reach scale in ways that challenge many traditional assumptions. The question facing healthcare investors is whether this moment is different enough to justify rewriting those assumptions or whether healthcare remains healthcare, with all the friction that has historically made ownership discipline essential.
The more we talked to investors, the clearer it became that they are splitting into two camps. One believes the possibility of AI-powered super-compounders justifies stretching on ownership to get into the very best companies. The other believes that when the market gets frothy, discipline matters more than ever. They’re looking at the same market, the same companies, and often the same data and reaching very different conclusions about what it takes to make the venture math work.
The ownership debate
There are a few questions VCs have always been able to answer in their sleep:
What verticals do you invest in?
What size checks do you write?
What ownership percentage do you target?
That last question has a mechanical answer that hadn’t really changed in the last decade. Typically, a fund leading a Series A needs to own ~20%. A Series B investor coming in behind existing shareholders might target 10%. By Series C, 5%. VCs have a target ownership percentage baked into their model because venture math only works if your winners are big enough, and “big enough” is a function of exit size multiplied by ownership percentage. If the exit is large enough and you own enough of it, the winner returns the fund.
But AI-native founders — at least the best ones — can now command valuations that make that calculation harder to hold. In fact, valuations have hit decade highs (beyond ZIRP-era levels).
This is being driven by some AI-native companies reaching $100M in ARR in four to eight quarters, compared to 18 to 24 months for top-quartile traditional SaaS. Leslie Feinzaig at Graham & Walker put it simply: “Capital is chasing high growth, because high growth is possible in a way that it never was before… There is no next round if you don’t have stellar growth.”
So now that valuations are higher and ownership targets are harder to meet, every investor must decide whether to stretch on ownership because they believe the exit math is completely different now… or not.
Investors split into two groups on this. We will call them The Convicted and The Disciplined.
Camp 1: The Convicted
The Convicted have effectively retired the ownership threshold as a binding constraint. Conviction is the constraint. One investor told us that they used to target 20% ownership, but are now “much below that.”
Rather than applying a blanket ownership target, the Convicted approaches each deal on its own terms, asking whether this specific entry valuation, at this specific check size, against this specific exit scenario, still makes the fund math work. If they believe in the opportunity, they will stretch — sometimes very far — to be part of the round.
Rebecca Mitchell at Scrub Capital says she underwrites every investment to ensure “the mix of entry valuation, terminal exit potential and optionality, timeline, and capital leverage could make it a fund returning winner.” If they can write a check size that makes the math work without compromising portfolio diversification, they write it, regardless of ownership size.
Underlying this view is a belief that the distribution of venture outcomes itself has changed. If AI is creating companies capable of reaching dramatically larger outcomes, then ownership targets designed for a previous era may no longer make sense.
“I think unequivocally ownership targets have had to be relaxed. Outcome potential of companies today is an order of magnitude what it was even a handful of years ago. And the power law is more inflected than ever before. The evolution in VC ownership is a reflection of that.” — Leila Rastegar Zegna, Kindred Capital
Perhaps the strongest argument for throwing out ownership targets is that such filters would systematically screen out the fastest-growing, highest-potential companies. When the Ontario Teachers’ Pension Plan invested $300M in SpaceX in 2019 at an estimated valuation of $33-36B, they acquired less than 1% of the company. That math might have looked crazy at the time, but at SpaceX’s current $2.1T market cap, that’s an ~$18B (60x) return for those teachers.
Hunter Walk of Homebrew, a tech investor who has backed healthcare companies like Honor, Headway, and Clarity Pediatrics, treats price as a signal that informs four questions on every deal:
What are the founders optimizing for?
Does the cap table reflect good judgment?
Will this pricing make the next round harder?
What does he have to believe to imagine a 20-50-100x return?
He believes these four considerations will matter more to Homebrew’s financial performance than trying to hit an ownership target. Yet he is the first to say that VCs should still err toward discipline.
Camp 2: The Disciplined
“You spend enough time in venture and you learn that concentration and ownership are the two most important things. We really strive to maintain that 10% ownership threshold.” — Derick En’Wezoh, Susa Ventures
The dot-com bubble of the late 1990s produced a generation of internet companies priced for a future that was real, but decades away. The COVID/ZIRP era made the same mistake in a compressed timeframe: a wave of richly valued health tech companies that could not grow into their valuations, followed by down rounds, recapitalizations, and a SPAC graveyard of startups that once commanded premium prices. In 2025, 14 of the 17 unicorns that went public did so at valuations below their last private-round valuations (a reminder that private-market conviction does not always survive contact with public-market scrutiny).
The Disciplined have watched this movie before. They believe that heat in the market is precisely when price discipline matters more, not less.
Leslie does not dispute the scale of the opportunity. “We are definitely in a moment of massive disruption and massive opportunity. Huge outcomes will come from this generation of companies,” she told us. But the heat means many companies are raising at inflated valuations: “It’s a solid core surrounded by a giant bubble.”
The team at NextView has written extensively about this moment. David Beisel points out that changes in entry valuation directly affect exit multiples and thus the return profile of the fund:
“Of course, you’d much rather have richly paid to be in a winning company than to have passed on it altogether. It’s permissible, and even expected, then, to make exceptions and occasionally pay a premium for potentially transformative companies. However, these special cases cannot become the norm without jeopardizing the portfolio’s overall cost-basis…” — David Beisel, NextView
His partner, Rob Go, names the specific failure mode the Disciplined are trying to avoid: paying up to invest in what you think is a super-compounder when the company is really only destined to be a “normal, run-of-the-mill unicorn.” You can do that once or twice in a portfolio, but if you end up overpaying for most deals, there is no next fund.
Because the step-up multiple from seed to Series A has declined from a peak of 4.2x in 2021 to 2.5x today, there is less room between rounds to recover from an expensive entry price. Bryant Barr at Penny Jar Capital (Stephen Curry’s fund) points out that “structural compression means ownership at entry matters more now, not less.”
While median step-ups are shrinking, the median post-money seed valuation remains steady at ~$24M. But at the top of the market, the 95th percentile has nearly tripled from $65.6M to $173.6M in just four years. This rising ceiling with a flat floor means most of the market is holding steady on price, except for the hottest deals where investors are willing to trade ownership for access.
Cameron McLain of Giant Ventures thinks ownership targets are most critical here at seed, where investors have their only opportunity to build a position capable of returning the fund. “At seed, ownership matters to returns. It’s math. But the size of the outcomes is the unbounded variable in the returns formula, so what matters most is being in the right companies.”
To get to that math, some investors are changing other variables in the equation. Steve Kraus at Bessemer still targets 15–20% ownership at the earliest stage, but doing so now means investing earlier and writing bigger checks. So while the ownership bar hasn’t moved, the cost of meeting it has.
Similarly, Derick has found that sourcing structure helps. Susa generally is not competing for opportunities that have hit the market. A large portion of what they invest in is sourced directly or through their network, locking in deals before the broader market even sees them. The discipline is easier to defend when you are not in the auction.
What both camps agree on
The two camps are a useful frame, but most investors don’t live at either extreme. Depending on the deal, the founder, and the heat in the room, many find themselves oscillating between the two. Where they land more often than not, though, is common ground.
No one is bargain hunting
For all their disagreement on discipline, neither camp thinks the answer is to find cheaper deals.
“In VC, there’s no reward for ‘value investing’: success hinges on identifying and backing upside outliers, not securing deals at relatively attractive prices. Playing a game of ‘subprime’ venture capital leads to partnering with ‘subprime’ entrepreneurs, a strategy unlikely to yield the outliers in the power law curve that defines our venture success.” -David Beisel, NextView
The Convicted are willing to pay higher valuations because they believe an even bigger outcome is possible. They know the alternative (passing on the best companies to protect an ownership threshold) would be an untenable loss.
The Disciplined are not hunting for bargains either. To reach their ownership targets, some are sourcing earlier, others are writing bigger checks, but no one is looking for a discount.
This isn’t 2021, nor is it 2000
“This is the most important technology shift of our lifetimes. This will be bigger than the internet, bigger than cloud, bigger than mobile. The lion’s share of enterprise value was created four to five years into those cycles. And we’re right there with generative AI right now.” — Derick En’Wezoh, Susa Ventures
Across the board, investors believe the surge in valuations is wholly different from prior bubbles.
The dot-com bubble produced companies with weak underlying financial models. Startups were priced on the conviction that the internet would transform everything, which it eventually did, just not on the timeline founders and investors believed.
The COVID/ZIRP era made a different mistake: low-cost capital chasing a pandemic-shaped demand surge led investors to confuse temporary tailwinds for structural change. As Steve Kraus at Bessemer put it, “in our industry, investors mistook telehealth as being permanent.”
But today’s heat is from a combination of transformative technology and an actual shift in buying behavior across health systems, payers, and life sciences.
“The definition of bubble is when the market price of an asset or industry far exceeds its intrinsic value. Are there mispriced assets right now? Yes, but the potential of this transformative tech is so large that I don’t think we have eclipsed the intrinsic value part of the definition and we may not even be close yet.” — Steve Kraus, Bessemer
Multi-year contracts are the new durability test
The single signal both camps cited most often as separating durable AI businesses from the rest was the multi-year enterprise contract. Health tech startups are landing their first enterprise partnerships faster than ever before. But as more founders show up to pitches with impressive partner logos, the question has shifted from whether you have any customers to what kind of customers you have.
“One of the clearest things we have heard from payers and health systems is that they are buying with an exit ramp in mind. They are describing short pilots, one-year terms, optionality if a platform vendor ships something comparable. It means durability must be earned in the contract structure, and founders who close multi-year commitments with near-term, attributable outcomes are demonstrating exactly that.” — Margaret Malone, Flare Capital Partners
This builds on the live-ARR bar founders described to us in Part 1. Founders told us live ARR replaced theoretical ARR as the bar for “this product actually works.” Investors are now saying multi-year duration is the next rung, the clearest signal that the product is actually durable.
Founder ambition is the new ceiling
The pattern that surprised us most was how often investors named founder ambition, not market size or technology, as the binding constraint on outcomes. “There is some point in the conversation where I wonder if they’re not thinking big enough,” Derick told us. He learned this when building Viz.ai, a category leader in AI and a billion-plus-dollar company (disclosure: I’m an investor): “In some ways, we weren’t thinking big enough. There’s a world where if we had thought and dreamt a lot bigger, this would be a 10x bigger company.”
Chirag Shah at Define Ventures described the same dynamic on the talent side, where the bar has moved toward people who can think AI-natively “versus just saying, ‘let me look at a process and make it more efficient.’”
The unanswered question
The question neither camp can yet answer is whether AI's promise of unprecedented scale can apply to healthcare.
The Convicted are making an implicit bet. They believe AI is creating a generation of companies whose growth and eventual scale will look fundamentally different from what healthcare investors have historically seen. Take, for example, the investors who backed OpenEvidence at $12 billion (~120x revenue) while Doximity is valued at ~$3.7 billion (5.7x revenue). These investors are betting that a new era of healthcare AI companies can grow faster and become something much larger than their predecessors.
The Disciplined are a little more uncomfortable with that approach. They believe that while AI may accelerate growth, it does not eliminate the structural realities of healthcare: long sales cycles, complex regulation, incumbent power, and, perhaps most importantly, finite market sizes. If those constraints remain, ownership still matters because the eventual outcomes remain bounded.
But all investors are vulnerable to another force: competition. If enough investors start chasing the same handful of perceived super-compounders, valuations eventually rise to a point where future returns deteriorate. Capital floods into increasingly marginal opportunities, expectations outrun reality, and even quality companies struggle to justify their prices.
Eventually, venture math will settle the argument.
Thank You
This piece draws on interviews with healthcare investors conducted in May and June 2026 as part of a Columbia Business School independent study. Our deepest thanks to the investors who shared their time, candor, and hard-won perspective:
Parth Desai, Flare Capital Partners
Derick En’Wezoh, Susa Ventures
Leslie Feinzaig, Graham & Walker
Steve Kraus, Bessemer Venture Partners
Margaret Malone, Flare Capital Partners
Cameron McLain, Giant Ventures
Rebecca Mitchell, Scrub Capital
Chirag Shah, Define Ventures
Alyssa Tsenter, Flare Capital Partners
Leila Rastegar Zegna, Kindred Capital
This newsletter is entirely free! How you can support this work:
📘 Order my book, Massively Better Healthcare!
⭐️ Subscribe to the Heart of Healthcare podcast.
💻 Download a copy of my digital health startup database.
🫶 Share this email with a digital health friend.
