The EHR vendors will eat every health-tech startup that the IPO window validates.

The structural pattern across the U.S. health-tech category since 2015 has been consistent. A health-tech startup identifies a clinical or operational workflow gap that the major EHR vendors are not addressing. The startup ships a product. The product gains traction across some number of health systems, demonstrating that the workflow gap is real and that buyers are willing to pay for the fix. The traction validates the category. Some years later, when the category is large enough to justify the platform-tier engineering investment, the EHR vendors ship the equivalent feature into their platform. The startup's market position erodes from the EHR-customer base, which is most of the addressable U.S. market. The startup pivots into a smaller niche, sells to one of the EHR vendors, or quietly winds down.

This is not a new pattern. The pattern has held across multiple categories (provider scheduling, patient communication, revenue-cycle management, prior authorization, clinical documentation, telehealth, post-discharge engagement, several others) for the last decade. The pattern compounds because each repetition validates the structural advantage the EHR vendors have, and the category-of-2024-2025 startups that are validating new IPO-window categories are structurally on track to be eaten by the same dynamic that ate the prior wave.

This case-study walks the structural advantage, the IPO-validation pattern, the categories of startup that survive the pattern, and the structural read on what to build that does not get eaten.

The structural advantage

The EHR vendors have three structural advantages against the health-tech startup category. The first is the customer-and-data graph: the EHR vendor's product is already deployed at the customer, integrated with the clinical workflow, holding the patient data, and running with the existing user-and-permission infrastructure. Shipping a new feature into the EHR ships it to every customer using the EHR with zero customer-acquisition cost on the new feature.

The second is the workflow-integration depth: the EHR is the operational system the clinicians actually use, and any feature that runs alongside the EHR rather than inside it carries integration friction the in-EHR feature does not. The friction shows up in clinician adoption, in the IT-and-procurement preference for in-EHR features, and in the contract structure that prefers bundled features over standalone vendor relationships.

The third is the validation-and-compliance posture: the EHR has already cleared the security, compliance, and audit reviews the health system requires, and the in-EHR feature inherits that clearance. The startup vendor selling alongside the EHR has to clear the equivalent reviews independently, which is multi-quarter work per customer.

The three advantages compound. The customer-and-data graph means the EHR can ship features faster. The workflow-integration depth means the features stick when shipped. The validation-and-compliance posture means the deployment is operationally cheaper for the buyer. The startup competing against the EHR for the same workflow has to overcome all three, and most do not.

The IPO-validation pattern

The pattern by which the EHR vendors decide which features to ship is roughly predictable. A startup category needs to demonstrate enough enterprise traction to validate that the category is meaningful at the platform-vendor scale. Traction milestones that signal validation include: cumulative customer logo count exceeding 50-100 major health systems; aggregate annual revenue at the category-leading startup exceeding $100-150 million; an IPO or acquisition transaction at a meaningful multiple. When one or more of these milestones lands, the EHR vendors take the category seriously and the engineering investment to ship the in-EHR equivalent gets prioritized.

The lag from validation to in-EHR shipping is typically 18-36 months. The startup category that validates in 2024-2025 should expect EHR-equivalent shipping in 2026-2028, with the EHR-side economics being meaningfully better than the startup-side economics for most customers. The startups that captured the most traction during the 2022-2024 enterprise-AI wave are the ones in the validation window now and are facing the EHR-shipping cycle through 2026-2027.

The IPO-validation pattern is the part of the picture the founder-class building in the category needs to read carefully. The IPO is not the safety milestone the founder-class often treats it as; the IPO is the validation event that triggers the EHR-ship cycle that consumes the category over the following 24-36 months.

Which startups survive the pattern

Three categories of startup typically survive the EHR-eats-the-feature pattern.

The first is startups with a data-acquisition-or-data-leverage moat the EHR cannot replicate. If the startup has built proprietary data infrastructure (specialty disease registries, longitudinal cross-system data, real-world-evidence cohorts, claims-and-clinical fusion data) that the EHR cannot acquire by shipping a feature, the startup retains a defensible position. The data moat is the strongest of the three.

The second is startups with deep vertical specificity that the EHR's horizontal-platform engineering cannot match. Specialty practice areas where the workflow depth requires multi-year clinical-domain investment (oncology pathway management, dermatology imaging, cardiology rhythm analysis, OB-GYN obstetric-care pathways) tend to support startups that the horizontal EHR does not displace because the EHR cannot operate at the depth the specialty requires.

The third is startups that integrate with multiple EHRs simultaneously and provide the cross-EHR data-and-workflow layer that no single EHR can ship. The aggregator position, where the startup is the integration tier between the major EHRs and the customer's broader operational stack, has structural defensibility because it is not in the interest of any single EHR to ship the cross-EHR-integration capability.

The startups that do not have one of these three positions are structurally vulnerable to the EHR-eats-the-feature pattern. The category leader without a data moat, a vertical-depth moat, or an integration-tier moat is the startup that gets eaten when the EHR ships the equivalent.

What the operator class should take from this

For founders building in the U.S. health-tech category in 2025-2026, the operator-class advice is to build for one of the three defensible positions or to build with a clear-eyed plan for the EHR-ship cycle that will eventually arrive at the category. Pretending the EHR-ship cycle does not exist is the founder-class mistake the prior wave repeatedly made.

For investors evaluating the category, the structural read is similar. The IPO-validation milestones are not the end of the de-risking; they are the start of the EHR-eat cycle. Pricing the round against an enterprise-software comparison set without accounting for the EHR-eat trajectory is the investment-class mistake the prior wave repeatedly made.

The Epic and Oracle Cerner platforms are not malicious; they are operating their structural advantage exactly as the platform-tier vendors in any category operate it. The pattern is general: when a platform vendor with a deep customer-and-data graph identifies a feature category that has been validated by a startup wave, the platform ships the feature and the startup wave consolidates into the platform. The U.S. health-tech category has been running this pattern for a decade and the trajectory through 2026-2028 is going to repeat it for the current IPO-window cohort.

Build for the data moat, the vertical depth, or the integration tier. Build with the EHR-ship cycle in your forecast. The startups that read the pattern correctly survive it. The startups that did not have generally not.

—TJ