The path to selling healthcare innovation has never been more complex, or more critical to get right. This week in New York, we convened a panel of healthcare operators, buyers, and founders to explore how innovation can survive and thrive within the most scrutinized sales environment in tech. Moderated by journalist Kristen V. Brown, the conversation brought together voices from across the ecosystem:

• Dr. Chris Gilligan, CMO, CQO & SVP, Robert Wood Johnson University Hospital
• Mike Goodman, Co-Founder and Chief Growth Officer, Oshi Health
• Dr. Anand Joshi, SVP, Chief Procurement and Supply Chain Officer, NewYork-Presbyterian Hospital
• Dr. Dan Knecht, Chief Medical Officer, EmblemHealth

With 40+ founders, executives, and investors in the room, the panel delivered hard truths about what it takes to sell into health systems, payers, and the broader healthcare market.

Market Pressures: The New Baseline

The financial reality is stark. Cuts to Medicare Advantage and Medicaid, rising employer costs, and exploding pharmacy spend from categories like GLP-1s are straining every stakeholder in the system. Meanwhile, a sicker, aging population demands more care from fixed resources.

For startups, this means the bar for proof has never been higher. Claims of value creation aren’t enough—founders must demonstrate ROI beyond reasonable doubt, and not just once. Evidence needs to be validated across different patient cohorts, payer contracts, and health system contexts. A proof point that works at one academic medical center may not translate to a community hospital with different patient acuity and payer mix.

The takeaway: Build your evidence strategy for repeatability, not just a single win.

Health Systems: Drowning in Pitches, Starving for Proof

Health systems can evaluate 100+ vendor pitches per quarter. In that noise, differentiation comes down to two things: seamless integration and undeniable ROI.

Epic’s deterioration index is the gold standard. This early warning system, built natively into Epic’s EHR, flags patients at risk of clinical decline. The result: a 20% reduction in mortality and thousands of lives saved across health systems. It succeeded because it embedded directly into clinical workflows: no new logins, no separate dashboards, no cognitive burden on already-stretched clinicians.

Contrast that with early sepsis alert systems. Promising on paper, many triggered excessive false alarms that eroded clinical trust and wasted scarce nursing resources. The technology wasn’t wrong, the implementation was. Without direct collaboration with frontline clinicians and respect for existing workflows, even strong solutions fail to scale.

The evaluation process compounds the challenge. Clinicians, IT teams, finance, and administration must all align before a purchase moves forward. And even after a successful pilot, solutions often underperform when rolled out across hospitals with different patient populations, staffing models, and reimbursement structures.

The takeaway: Non-Epic native solutions face an uphill battle. If your product requires IT maintenance, adds clicks to a clinician’s workflow, or lives outside the EHR, you’re already behind.

The Payer Perspective: The Most Misunderstood Buyer

Payers are often painted as the villains slowing down innovation. The reality is more nuanced. In a fee-for-service world, payers are the only stakeholder consistently pushing back against overutilization. Their job is to prevent unnecessary spending without compromising patient access—a balancing act that makes them deeply skeptical of unproven solutions.

Payment models have evolved. The traditional per-member-per-month (PMPM) structure is giving way to shared savings arrangements and guaranteed backstop agreements, where vendors assume financial risk for outcomes. This shift means founders can’t just promise value—they need to guarantee it, often with their own capital on the line.

Clinical rigor is non-negotiable. Non-FDA approved compounds, even if cheaper, are rejected outright. Cost pressures from rising hospitalizations and expensive specialty therapies have made payers even more selective. To earn a spot on a payer’s roadmap, startups must show measurable clinical outcomes, demonstrated cost savings, and a clear path to patient safety.

Dr. Knecht emphasized that payers aren’t trying to block innovation. They’re trying to separate signal from noise in a market flooded with unsubstantiated claims.

The takeaway: If you’re not willing to take on risk-based contracts, you’re not ready to sell to payers.

Case Study: How Oshi Health Cracked the Code

Mike Goodman’s journey building Oshi Health illustrates what success actually looks like. From day one, Oshi committed to building actuarial models line by line, mapping ICD-10 and CPT codes directly to patient outcomes. The company partnered with UnitedHealthcare on a clinical trial that validated more than $10,000 in all-cause cost savings per patient.

Strong results, right? But the initial go-to-market strategy, a referral-based partnership model, failed. The problem wasn’t the clinical outcomes. It was payer mix misalignment. Patients being referred didn’t match the insurance contracts Oshi had in place, creating friction that killed momentum.

Oshi pivoted. Instead of operating separately, the company embedded gastroenterologists directly within health system clinics, billed through the health system’s Tax Identification Number (TIN), and structured shared economics that aligned incentives. This model created the operational and financial alignment necessary to scale.

The lesson: proving clinical value is table stakes. Scaling requires adapting your business model to match the operational realities and economic incentives of each partner.

The takeaway: Your first GTM strategy is a hypothesis – and it might be wrong. Build flexibility into your model from the start.

Avoiding Dead Ends

The panel didn’t hold back on issues that consistently impede a solution breaking through with a purchasing committee:

Integration Barriers

Non-Epic native solutions create ongoing maintenance burdens for IT teams already stretched thin. Unless your technology lives inside the EHR and requires zero additional support, adoption will likely stall.

Reimbursement Black Holes

Prescription digital therapeutics continue to face a triple threat: FDA review delays, physician adoption resistance, and lack of payer reimbursement. Several panelists mentioned companies in this space that burned through $50M+ without achieving sustainable revenue.

GLP-1 “pill mills” relying on non-FDA approved compounded drugs were called out as both potentially unsafe and commercially unsustainable. The regulatory risk alone could scare away serious investors.

The Niche Solution Trap

Point solutions that solve a single, narrow problem rarely scale in healthcare. Health systems don’t want another vendor relationship, they want platforms that adapt across multiple conditions, departments, or workflows and deliver measurable value in more than one place.

The takeaway: Healthcare purchasing committees have seen these patterns fail repeatedly. Avoid the graveyard: non-native integrations, reimbursement uncertainty, and single-problem solutions rarely survive contact with the market.

Looking Ahead: Telehealth and AI

Academic medical centers already conduct complex interventions in person and maintain virtual follow-ups, extending their geographic reach far beyond traditional catchment areas. Hospital-at-home programs exist and work—but operational adoption lags because payment models remain unclear.

The panel drew a sharp distinction between expanding virtual visits within health systems and outsourcing core patient interactions to third parties. The former is accelerating. The latter remains misaligned with hospital mission and clinical accountability.

Future opportunities like agentic voice bots entering the home show promise, but without clearer reimbursement pathways, these innovations remain pilots in search of business models.

The takeaway: Focus on tools that extend a health system’s existing clinical capabilities, not platforms that try to replace them.

AI: Traction in the Boring Stuff

The consensus among panelists was that healthcare will keep humans firmly in the loop. Technologies that try to remove clinician judgment from the equation will fail, and those that augment clinician capacity and reduce administrative burden will succeed.

What’s working:

• Radiology flagging tools that highlight abnormalities
• Documentation and dictation platforms that reduce charting time
• Back-office automation for prior authorizations and claims processing

What remains unproven:

• Ambient listening tools show value only for heavy users; light users report minimal benefit or workflow disruption
• Generative AI for chart summarization and patient scheduling remains mostly experimental
• Utilization management AI faces deep skepticism about accuracy and liability

The takeaway: If your AI requires clinical trust to work, you need clinical validation to sell it. And until the pricing models align with demonstrated value across all user types, adoption will remain uneven.

Closing Thoughts: Healthcare is About People, Not Technology

Dr. Anand Joshi closed the panel with a reminder that cuts through the hype: Healthcare is not in the business of innovation. It’s in the business of people.

Employees, clinicians, and patients sit at the center. Technology only matters if it supports them. For founders, that means building trust through rigorous evidence, continuously proving value across different contexts, and creating alignment across a complex ecosystem of payers, providers, and patients.

At 53 Stations, we back founders who balance vision with discipline: those who adapt quickly, respect the system’s human fabric, and prove their impact at every step.  If you’re building something in this space, or thinking about it, we’d love to hear from you!