The Same Bottleneck, Two Industries
Ten years ago, healthcare interoperability had one job: get a patient's record to follow them from one system to the next. Today, after a decade of HL7, FHIR, and federal mandates like TEFCA, the technical capability to move clinical data exists almost everywhere. TEFCA alone has gone from roughly 10 million records exchanged in January 2025 to nearly 500 million by early 2026, with new participants like the Social Security Administration now connected for benefits determination. The bottleneck moved. It's no longer can the data move, it's will the systems holding it agree to let it, and federal officials themselves frame the open question now as data liquidity and governance, not transport.
Regional food systems are running roughly the same decade with a lag. The technical and logistical capability to move surplus production from small and mid-size farms to nearby institutional buyers, hospitals, universities, school districts, has existed for a long time. What's missing is the connective layer: the entity that knows what a given farm has, what a given buyer needs, and is positioned to broker the match at a price and reliability level both sides can plan around.
Strip away the domain language and the structural problem is identical.
A set of stakeholders, each holding partial information about supply or demand, operating without shared infrastructure to make that information visible to the other side. In healthcare, the result is a clinician making a decision without the full record. In food systems, the result is a hospital cafeteria sourcing from a distributor two states away while a comparable product sits unsold a few miles out. Neither failure is a data problem in the conventional sense. Both are coordination problems wearing a data costume.
This matters for how these projects get scoped. A food hub feasibility study that treats itself purely as a logistics and capital question, where to put a warehouse, what equipment to buy, is solving the wrong layer of the problem. The harder and more durable work is building the matching and trust infrastructure underneath it, the same work healthcare interoperability has spent a decade learning is mostly organizational, not technical. TEFCA's own architecture reflects that lesson: it's explicitly built as a network of networks, governed by a common agreement among Qualified Health Information Networks rather than a single mandated data format. Getting competing institutions to share information requires governance and incentive alignment, long before it requires a schema.
For organizations building in either space, the lesson transfers directly: invest in the connective tissue before the infrastructure, not after. The capital project is the easy half.
