On 'Epic-as-a-monopoly' framing.
The trade-press habit of describing Epic as a monopoly is wrong in a specific way that matters. Epic has substantial market power. Epic's market power has real consequences for the U.S. health-system technology landscape. Calling that condition "monopoly" generalizes the term out of usefulness, produces bad regulatory framing, and confuses the operator-class read on what Epic actually is and what it actually does.
Monopoly, in the antitrust-and-competition sense the term originally carried, is a market with a single seller, no close substitutes, and structural barriers preventing new entry. Epic is none of these. Epic is the leader in a category (large-health-system EHR) that includes Oracle Cerner, Meditech, Allscripts (now Veradigm), athenahealth, NextGen, eClinicalWorks, and a dozen smaller players, with active customer migration in both directions across the category. Epic's market share in the large-health-system slice is somewhere on the order of 30-40 percent, depending on the cut, with Cerner running 20-25 percent and the long tail covering the rest. That is concentrated. That is not monopoly.
What Epic does have is structural switching cost. A health system that runs Epic for a decade has integrated it with two hundred internal systems, trained ten thousand users, configured the customization layer against years of operational experience, and entangled the financial-and-clinical workflow so deeply that switching out is a five-to-ten-year project costing low-to-mid hundreds of millions. The switching cost is the actual moat. Switching cost is not the same thing as monopoly. A market with high switching cost and multiple competitive vendors, where customers do switch (slowly, expensively, but they do), is a market with sticky-incumbent dynamics, not monopoly dynamics.
The framing matters because the regulatory and operator-class responses to "Epic is a monopoly" are different from the responses to "Epic has structural switching cost in a concentrated market." The first framing points at antitrust enforcement, market-share caps, mandatory unbundling, FTC-or-DOJ enforcement actions. The second framing points at interoperability standards, switching-cost reduction through API-and-data-portability requirements, and the kind of operator-level work the ONC information-blocking rules have been doing since 2020. The second framing is the one that actually helps. The first framing is the one the trade press writes because it is the more press-friendly version.
The part that holds on Epic's actual position is more precise than the trade-press version. Epic is the dominant vendor in a sticky-incumbent market, with substantial market power, real switching-cost moat, and meaningful but slow customer migration. The right policy response is the one that reduces switching cost and improves the data-and-API portability between vendors, not the one that treats Epic as a monolithic anti-competitive actor. The right operator response, for founders building healthcare software in the orbit of Epic and the other EHRs, is to build assuming Epic remains dominant and that the market structure is stable, while building products that operate cleanly inside the interoperability infrastructure that the regulatory framework is steadily improving.
Epic-as-monopoly is the wrong frame. Epic-as-sticky-incumbent-with-substantial-market-power is the right frame. The two frames produce different policy responses, different investment theses, and different part that holds on what to build. The trade press will continue to use the first one because it is shorter and more dramatic. The structural read should use the second one because it is correct.
—TJ