a16z Podcast: So You Wanna Build a Software Company in Healthcare?

Episode Summary

Executive Summary: The episode explains why building software in healthcare is fundamentally harder than in other sectors: workflows are fragmented, adoption is risk-averse, reimbursement and regulation distort incentives, and pure SaaS rarely fits. Jorge Conde and Julie Yu argue successful healthcare companies must deeply understand local workflows, often combine software with services or appliances, choose narrow segments carefully, and expect long sales/adoption cycles before achieving durable scale.

Main Topics: Why healthcare software is different (Priority: 5/5): Healthcare software is undervalued, often sold as a component of a larger service, and must meet much higher standards because errors can affect patient safety and care quality. Workflow complexity and local variation (Priority: 5/5): Providers, payers, and even similar-looking hospitals operate differently, making it difficult to generalize a single product across the market. Kairos case study: call centers and service model shifts (Priority: 5/5): A product intended for hospital call centers revealed massive heterogeneity in call-center scope and duties, forcing a hybrid software-plus-services model. Gnome case study: genomics, cloud resistance, and appliances (Priority: 5/5): A genomics software company discovered clinical labs were unwilling to pay upfront or send data to the cloud, requiring a boxed appliance deployment option. Team building and domain expertise (Priority: 4/5): Customer-facing and product roles need healthcare-specific experience, while engineering can benefit from outside perspectives and modern technical creativity. Timing, product-market fit, and long adoption cycles (Priority: 5/5): Healthcare can be 'too early' for years; adoption is slow, switching costs are high, and companies need significant runway to survive long sales and implementation cycles. Value proposition, pricing, and adjacency strategy (Priority: 4/5): Founders must know who pays, what value is created, and how to expand from one use case to the next without relying on a pure MVP approach.

Key Arguments: Healthcare is harder than other software markets because software is often devalued, embedded in broader services, and subject to patient-safety scrutiny. There is no single standardized workflow in healthcare; local variation across hospitals, labs, and physicians makes one-size-fits-all software ineffective. Pure SaaS is often insufficient in healthcare; companies frequently need services, training, or full-stack support to absorb variability. Reimbursement and payment flows can invalidate seemingly rational ROI models, as with genomic sequencing that was too expensive for many labs upfront. Cloud adoption is constrained by privacy and risk concerns, so product architecture may need on-premise or appliance alternatives. Customer-facing hires should have healthcare context and cultural fluency; engineering hires should be able to invent around legacy systems rather than reproduce old healthcare patterns. Healthcare products often require long periods to gain trust and scale; companies should raise enough capital to survive adoption lags and slow enterprise sales. Founders must identify the specific market segment they serve because 'providers' is too broad to be actionable for product or go-to-market strategy. A credible value proposition matters more than generic cost-cutting claims; companies need to prove meaningful outcomes or workflow improvement with real buyers. The best path to scale is often to start with a narrow wedge, then expand through adjacent use cases and ecosystem integrations.

Data Points: Years to mass-market adoption of medical technologies: 17 years - Cited as the average time for a new medical technology to become mass adopted, illustrating healthcare's slow adoption cycle. Potential annual solution threshold: At least a half a million dollars a year - Suggested as the minimum scale needed for a healthcare business to be financially worthwhile given enterprise sales and deployment costs. Flatiron Health acquisition value: $2 billion - Used as an example of data-driven value creation in healthcare through EMR-based insights. EHR companies' age: About 45 years old - Referenced to show how long market-leading healthcare software companies can persist and how entrenched incumbents are. Company journey to stride: About 20 years - Noted that major EHR companies did not hit their stride until roughly two decades into their existence. Runway expectation: More than 3 years - Implication from the discussion that nothing meaningful happens in healthcare in under three years, so companies need extended runway.

Pivotal Quotes: "A pure software company in healthcare is a really hard thing to do." — Julie Yu: Summarizing why software-only models struggle in healthcare due to variation, services needs, and devalued software economics. "Nothing meaningful happens in healthcare in under three years." — Julie Yu: Describing the long adoption, sales, and implementation cycles that require founders to raise sufficient runway. "I would argue that product market fit doesn't exist in healthcare." — Jorge Conde: An extreme framing that highlights how regulation, pricing, incumbency, and workflow distort normal market dynamics.

Implications: Healthcare founders should plan for deep segmentation, hybrid delivery models, and long time horizons. Success depends on solving real workflow pain for a specific buyer, then expanding carefully with enough capital and operational flexibility.