Strip out nine companies and digital health funding fell in 2025.
Golden Hour invests at the other end of that barbell, at seed and Series A, where clinical workflow meets the ledger.
The recovery is nine companies.
Digital health took $14.2B across 482 deals in 2025. It reads like a recovery. It is not one. Twenty six megadeals absorbed 42% of the capital, and in the first half of 2026 that rose to 45%, with 8% of deals taking 46% of the dollars. Thirty five percent of 2025 rounds were flat or down.
Underneath the headline, the sector rotated hard. Alternative care, the virtual-first and telehealth bucket, fell from 42% of healthtech investment in 2021 to 9% in 2025. Provider operations went the other way, from 19% to roughly 44%.
The telehealth generation did not fail on technology. It failed on acquisition economics with no payer underneath. Thirty Madison went from a $1B mark to $500M. SteadyMD sold for less than it raised. Hinge Health went public below its last private round, and that was one of the good outcomes.
We invest where the clinical event and the financial event are the same event.
Three things we believe.
Hospital at home is the only category in American healthcare with five years of locked federal payment parity.
The Acute Hospital Care at Home waivers were extended through September 30, 2030. Three hundred seventy three hospitals across 140 systems in 37 states already operate under them. Every other reimbursement thesis in digital health has a 2027 or 2028 expiry stapled to it. This one does not, and almost nobody has repriced for that.
On January 1, 2026, utilization management became a two-sided market.
Under the WISeR model, CMS now pays technology vendors a share of demonstrated savings to review and deny prior authorization requests in traditional Medicare, across six states and twelve service categories. Whatever your politics on that, the provider-side response is a demand curve created by federal rule rather than by a sales cycle: authorization defense, appeals automation, and documentation for medical necessity.
The end of enhanced ACA subsidies is the largest untraded event in consumer health.
Effectuated enrollment is projected to fall from 22.3 million to between 16.5 and 17.5 million. Average net premiums are up 58%. The average deductible moved from $2,759 to $3,786, the steepest single-year increase this market has recorded, and Bronze plans overtook Silver for the first time. A record-deductible population is mechanically a cash-pay population. That, not consumer preference, is why 29% of digital health companies now sell direct.
Five places we look first.
Two passes we make every time.
Companies whose thesis is that they are AI-enabled. Rock Health retired AI-enabled as a tracked deal category in the first quarter of 2026, because half of all deals qualified and it had stopped distinguishing anything. If AI is the answer to why now, we do not have a conversation.
Anything that requires a fixed hospital count to expand. There are not more hospitals coming.
Send us the part you are worried about.
Not the part you have polished. If you are building where clinical workflow meets financial workflow, we want the ugly slide.
Say hello