Morning. The cleanest trade in the market right now is not software versus hardware. It is proof versus promise.
The companies getting rewarded are the ones that can show who pays, who signs, and what happens when the nice demo runs into a budget line. A little less glamour. A lot more survivability.
Today is mostly about businesses that turned the hard part into the business model. That makes the math less elegant and the moat more annoying, which is often how useful companies begin.
Menu of the day
Markets: the tape is grading the bill
The article: a transplant network that outgrew the device
Shot: policy, payments, and product habits
Startup Lesson: budget the aftercare
What to watch: where control becomes the pitch
Market pulse
Quick note: Friday’s close is the latest full session before this goes out.
Markets: U.S. risk appetite is still intact, but it is more selective than celebratory. Growth can still work, yet the market is less forgiving when capex, energy use, or operating expense looks open-ended. That keeps the AI trade hot and the financing language colder.
Names: TransMedics is the cleanest version of that shift. The business is growing, but the market is watching the cost of scaling the logistics layer almost as closely as the revenue line. Anthropic’s foreign-access restriction also matters here, because access itself is becoming a regulated surface, not just a product feature.
The article
HealthTech / Infrastructure
TransMedics and the price of keeping organs alive
Quick reminder: organs do not care about the pitch deck. They care about time, temperature, transport, and whether the thing showing up at the hospital is still usable.
TransMedics reported first-quarter 2026 revenue of $173.9 million, up 21% year over year, and reiterated full-year revenue guidance of $727 million to $757 million. That is still a strong growth profile. The catch is that it is also a more expensive profile. Gross margin came in at 58%, down from 61% a year earlier, and operating expenses rose as the company kept building out the network around its OCS platform and National OCS Program.
That combination is what the market is trying to price. The headline still says medtech. The real business looks more like a regulated logistics and coordination layer with a device attached to it. The device matters. The network matters more. The aircraft, clinical teams, procurement process, and monitoring stack are not a side quest. They are the point.
That is why this kind of company can look expensive on the way up and still be making the right move. TransMedics is not just trying to sell hardware into an existing workflow. It is trying to expand the number of organs that can actually make it to transplant, which is a much bigger proposition. If you can improve utilization, you are not stealing share from a competitor. You are enlarging the market, and the pricing argument gets a lot more interesting.
The problem for public markets is that they do not like paying for the future in the shape of current margin pressure. They will accept that the bill exists. They just want to know who owns it, how long it lasts, and whether the spending creates a defensible control point or just a larger expense line.
That is where TransMedics gets useful as a read-through for the broader market. The easy version of the story is “AI and infrastructure spend are still fine.” The harder, better version is “the market is willing to fund systems that control a workflow, but it wants the control surface to be legible.” In this case, the workflow is transplant logistics. In other cases it is procurement, payments, compliance, or model access. Different category, same mood.
So yes, this is a medtech story. It is also a reminder that the best businesses in 2026 are increasingly the ones that make an ugly operational problem boring enough to bill for. Not exactly a small detail.
Shot
Model access gets political
Anthropic’s newest models were pulled back for foreign nationals after a U.S. export-control directive. The important part is not the model name. It is the precedent. Access to frontier AI is starting to look less like a product toggle and more like a regulated privilege. That changes how global teams, research orgs, and investors think about dependency.
Payments still break at the border
The cleanest payments systems in the world still stumble once money crosses a border. UPI and PIX are great domestically, but the friction reappears fast outside the home market. Tokenization keeps getting sold as the elegant fix. That may be true eventually. For now it mostly means the market is still searching for a cross-border stack that does not make everybody miserable.
Consumer instincts are usually right
Mark Pincus’s “Proven, Better, New” framing is a useful reminder that product work usually dies in the gap between novelty and usefulness. Copy what already works, make it better, then add the new thing only if it earns its place. It is a nice antidote to founders who think ambition alone counts as distribution.
Startup Lesson
Budget the aftercare
If your product changes workflow, compliance, logistics, or access, the buyer will ask a boring question faster than they ask about delight: who owns the outcome when this breaks?
That is the test. The startup that can answer it cleanly usually has a better shot than the startup that only sells aspiration. Control is what gets budgeted. Surprise is what gets delayed.
Doggy Bag
Fun fact: transplant logistics is one of the few businesses where geography can create value instead of just distance.
The number: 22 aircraft. That is TransMedics’ owned fleet as of March 31, 2026. No one buys that kind of thing for vibes.
The quote: The market wants a receipt, not a mood.
Reco: Write the one-page buyer map for your product. User, approver, budget owner, and the person who gets blamed when it fails.
What to watch
Whether more AI companies start talking about approval paths, access control, and employee eligibility instead of just model quality.
Whether investors keep rewarding businesses that own a workflow, even when the workflow comes with heavier capex and more operational drag.
Whether cross-border payments and tokenized settlement move from theory into actual procurement language.