Commercial growth strategies were built for a market that no longer exists. The volume game stopped paying off, the upper hand in payer contracting went with it, and what is left to compete for is the routing decision — where patients go, what they pay, and whether they stay. That decision is increasingly made outside the health system, by the employers who fund the plans.
So a strategy team goes looking for the most visible form of it: employers in its market contracting directly with providers, around the carrier. It checks the largest employers, asks the payer contacts, reads the trade press for a named deal. It finds none, and closes the question. Not happening here.
The scan is looking for the wrong object.
It's happening in pieces
It is watching for a direct contract: signed, named, announceable, the kind that produces a press release. Almost nobody starts there. An employer takes one line off the shelf. Musculoskeletal care goes to a specialty vendor. Fertility carves out to a program built for it. Surgery gets a center-of-excellence rider — a plan feature that sends a procedure to a designated provider, often with travel paid. Each is a plan-design decision, not a relationship decision: made in a benefits committee, implemented at renewal, producing measurable savings the first month it is live.
There is no press release for a carve-out. There is no notice to the health system whose volume just moved. The employer is not being adversarial.
The employer is not thinking about the health system at all.
That is why the survey number looks the way it does. Surveyed employers holding some form of direct arrangement with providers — a bundle, a center of excellence, a steerage program, a full direct contract — went from 18.7% to 27.6% in two years. Another quarter say they are evaluating one inside three years.
It is not evenly distributed. Among employers with 50,000 or more people, adoption runs at 44%. Under 1,000, it runs at 20%. The average conceals a market that is moving fastest exactly where the commercial volume concentrates.
And "some form of" is carrying real weight. The survey counts a center-of-excellence rider and a full network replacement as the same event, because from the employer's side of the table they are the same decision made at different sizes. That definitional spread is not a weakness in the data. It is the finding. A scan built to detect the full arrangement will keep coming back empty in a market where the pieces are already moving.
And the pieces are not small. In one independently evaluated bundled-payment program across self-insured employers, roughly a fifth of eligible surgical patients went through it — nearly a third in bariatrics. A benefits committee added an option and cut the patient's share of the cost.
Every one of those pieces is a decision about where volume goes. What gets missed is what happens once it is made.
This market doesn't reprice. It allocates.
In a contract market, terms get more expensive as the market matures — same position, higher price, and a late entrant can still buy in by paying more. That is the instinct the employer question invites.
This is not a contract market. An employer decides where its valuable surgical volume routes, and today it is deciding in fragments — around the payer, or carved out with the payer — each arrangement priced by an entity setting its own terms. The care still happens at a hospital. But the hospital inherits the fragmentation as instability it cannot see in any contract it holds.
And each fragment is a lane. An employer's spine cases route to one partner. Their musculoskeletal spend carves out to one vendor. A benefits committee does not select three. Once a lane is allocated, there is no paying more for the same position — the position is taken, and what remains is a different, worse purchase.
The position isn't an asset money can buy later.
The lanes that close first are the ones that make a rate increase worth pursuing: spine, joints, bariatrics — the high-cost, schedulable procedures that can be steered rather than simply paid. Both sides of this market read the same margin data.
"We can do this later"
That is where the fair objection comes in, and it usually comes from the CFO: the shift is real, and the function can be built when it matters. Later is available.
In a contract market that would be right. Here, later costs four things.
You are displacing, not introducing. The anchor employer has signed with someone — a competitor, a center-of-excellence network, a navigation vendor. You are now fighting an incumbent with switching costs, a track record, and a broker who owns the relationship.
You go from partner to vendor. Early, you design the arrangement: you pick the service line, set the bundle, write the terms, because the employer has no alternative. Late, five systems that all built direct-to-employer functions answer the same procurement, and competition does what competition does. Early you are chosen. Late you are compared.
You rent access instead of owning it. The employer relationships systems leave unattended do not stay unattended. A system arriving late reaches that employer through a navigator — a vendor that steers employees to providers under contracts it holds — on the navigator's steerage rules, at the navigator's negotiated prices, with the navigator holding the relationship.
That is not a forecast. A national surgical center-of-excellence vendor recently described the easiest prospect call of its life: the employer's benefits stack already contained six of the vendor's integration partners, and every one of them had independently named it the preferred specialty-care partner, integrations built. Nobody had worked that account. The routing decision had been made for the employer, by the vendors it already trusted, over four years — before anyone from a health system was in the room. In the vendor's own words: you can't buy that, and you can't announce your way into it.
You negotiate from need. Build the function now and you do it deliberately. Wait, and you build it after the volume visibly walks or in the middle of a payer dispute — and everyone across the table can see which it is.
One more asymmetry runs the other way. Employers built this fragmentation themselves, one point solution at a time, each justified on its own savings case. The complexity is now a cost of its own. The system that shows up early with one coherent answer is positioned to be the consolidation when employers go looking for it. The system that waits is one more fragment in a stack they are already trying to shrink — and the stack is not waiting to be shrunk. The vendors inside it are integrating with each other, quarter by quarter, into a network with one shared answer for specialty care: the consolidation forming with no health system in it.
"Later" is also what the industry said about the last shift of this shape. Fifteen years ago the migration of surgery to outpatient settings was readable in the data — codes moving, payer policy steering to lower-cost sites, higher utilization at lower cost. The systems that read it early accounted for that upside before the capacity existed, while nobody else was competing for the position. Then everyone saw the same opportunity, at the same time, in the same way: the same suburban corridors where employer-sponsored lives are densest, a competing center in every catchment, the same volume chased at rising acquisition cost. The surgery centers became mini health systems, facing the same obstacles the hospitals face — with much less of the upside. The clearest sign is the hospital's oldest subsidy problem arriving at the surgery center's door: the share of centers expecting to pay stipends to keep anesthesia covered went from 28% to 44% in a single year, while the same denials and delays land on them and the schedules fill up like the hospital's. The opportunity was never the surgery center. It was the window before the surgery center was consensus.
The carve-out layer is running the same arc now. Individual vendors, each cheap to contract with a few years ago, are becoming a network — and a network is the consensus phase.
Everyone can see the shift. Almost nobody checks whether the position is still open.
The systems that build the employer function now set the terms. The ones that wait inherit them.
The person who feels it never owns the fix
If the position is that cheap and the cost of waiting that clear, why isn't the function already built? Not for lack of effort. Because the problem is structured to be invisible.
The standard playbook builds capacity and markets it to patients, negotiates contracts and pitches the payer, stands up service lines and waits for volume to follow the network. All of it assumes the person who feels the need is the person who decides. At the employer level, the plan decides, and the patient chooses inside it.
And inside the system, the person who feels the loss never owns the fix. The employee whose knee is shot feels the benefit friction; the CFO owns the plan design that created it. They will never be in the same room. The surgeon feels the referral leaving; access services owns the routing. The org chart puts the wound and the bandage in different departments.
So the loss lands the way the ownership does — in pieces. A little leakage. A little payer-mix drift. A service line running under plan. Volume that walks in fragments never shows up whole in any single report. It cannot be seen whole from outside, either: a self-funded employer's claims sit with the employer and its administrator, and nowhere public. No market report was ever going to show it.
It does not look like a problem. It looks like noise.
There is exactly one seat in healthcare where that split does not exist: the employer. The person who feels the cost of the health plan and the person who holds the lever — plan design, steerage, the direct contract — are the same person. Everywhere else they are divided.
Every commercial failure I've been hired to explain has come down to the same thing: the person who feels the pain and the person who can fix it are two different people, and they don't talk.
Every seat in the C-suite is already reaching for this. Facilities is building the network. Marketing is running growth campaigns. Managed care is negotiating the contracts. Occupational health holds the employer relationships. Somebody bought a dashboard. Every one of those is real, and every one of them is aimed at the same commercial volume.
None of them owns the decision that routes it.
Commercial growth is every department's goal and no single department's job. That is not news to anyone running a health system. What is new is the price of leaving it that way.
Which lines, at which renewal
Which is why the scan asks the wrong question. Not whether anyone in your market has signed a direct contract — but which of your commercially insured service lines an employer in your market could route somewhere else at its next renewal, and whether your standing in those lines is strong enough to survive it.
← All insights