For over a decade, I told the same story about Unity Environmental University. We must transform from a caterpillar to a butterfly.
It was a good story. It was also the only one the sector was willing to hear. Higher education loves metamorphosis because metamorphosis is safe. It promises that the ugly part is temporary, that the pain has a schedule, and that something beautiful is waiting at the end of it. Boards find that comforting. Employees find it survivable. Donors find it fundable.
And the biology held up longer than most metaphors do. Inside the chrysalis, the caterpillar does not politely rearrange itself. It dissolves. Almost everything it was becomes liquid, and a small cluster of cells that were there the whole time rebuilds the animal out of that liquid.
That is a fair description of what we did. We sold the residential campus. We froze tuition. We changed our governance and operating model. We built a ninety-credit bachelor’s degree while the sector defended one hundred twenty (sometimes more). We took enrollment up fifteen times over and operating cash reserves up twenty times over. We did not optimize the old animal. We liquefied it.
So I am not disowning the metaphor. I am retiring it, with respect, because it has one flaw I can no longer work around.
A butterfly is a final draft.
That is the flaw in how this entire sector thinks about change. We treat transformation as an event with a finish line. Endure the disruption. Unveil the new structure. Declare victory. Then spend the next twenty years protecting the thing we just built from the next disruption. The transformed institution quietly becomes another institution that cannot transform.
The real question for the next ten years is not what we are becoming. It is what kind of animal can keep becoming things.
The trap on the other side of the chrysalis
Transformation is not a virtue. It is a bill.
It comes due when an organization has refused to adapt for long enough that only demolition will work. Every transformation I respect in this sector costs somebody something permanent. A career. A reputation. A payroll. Years of institutional peace. I have paid parts of that bill personally, and I am not romantic about it.
The safest question any organization asks under pressure is whether it is optimizing. I have watched that question hide the ground moving underneath an entire function while everyone stayed busy. Optimization is what you do inside a design. It cannot save you from the design.
Which is why the celebration of transformation makes me uneasy. We hand out awards for the rebuild and never ask why the rebuild was necessary. An organization that must transform is an organization that already failed to adapt. The transformation is the receipt. I have made the sector version of this argument before, that we spent decades expanding access without ever redesigning how any of it operates.
And here is the part almost nobody plans for. The moment the new model works, the reflex to defend it arrives right on schedule. The people who fought hardest for the change become its most protective owners, because they can still feel what it cost. That protection is how a new model calcifies into an old one. I have watched leaders do genuinely brave structural work and then guard it with the exact instincts that made the work necessary in the first place.
So the target is not a better transformation. The target is an institution that does not require one.
The animal we almost became
Before I describe what we chose, I want to describe what we walked toward, because I nearly built it and most systems that call themselves systems already have.
We almost became a hydra.
There are two hydras, and both are warnings. The one from myth has many heads and one body. Cut a head off and two grow back, which sounds like resilience until you look at the actual arrangement. Every head competes for the same stomach. Every head believes it is the brain.
Nothing gets swallowed that all of them do not agree to swallow. That is what happens when an institution adds units without redesigning the organism. Nine leaders, nine cabinets, nine strategic plans, one balance sheet. The redundancy that looks like strength is duplicated overhead with a governance problem stacked on top. Hercules did not win that fight by cutting faster. He won by closing the necks. Organizations with many heads do not die at the head. They die at the connection.
The second hydra is real, and it is closer to home.
The freshwater hydra appears not to age under favorable conditions. It renews its own cells continuously, and it spends much of its life attached to a surface, waiting for what it needs to come within reach. It is longevity without reach.
Higher education spent sixty years confusing the ability to persist with the ability to pursue. The current delivered eighteen-year-olds, and the strategy worked, so nobody noticed it was not a strategy. Then the current changed, and an animal built to wait started calling that a demographic cliff. It was not a cliff. It was a diet. And the economics underneath that diet are far more damning than the demography, because an institution that cannot survive without subsidized tuition is not financially sustainable. It is financially dependent.
I know how close we came to this, because our own instinct the first time we stood up a new enterprise was to give it its own everything. Its own leadership. Its own support staff. Its own version of every process. That instinct feels like empowerment. It is head growth. We stopped, and we started asking a different question about every function in the place. Not whether it should be centralized or decentralized. Whether it needed a brain of its own at all. Answering that with any rigor cost us nearly three years of data integration work that no one outside the institution found interesting, and it is the only reason anything since has been possible.
Thinking with your tentacles
An octopus carries roughly two-thirds of its neurons outside its head, distributed down its tentacles, which means each tentacle holds real processing power and can solve a local problem without checking upstream. It is often described, imperfectly but usefully, as nine brains.
Take three things from that animal and leave the rest of the aquarium alone.
Distributed intelligence. The head sets intent, holds the strategy, and reads the environment. The arms execute with genuine authority and do not petition for permission to do what they were built to do. The head does not track the position of every sucker and would be a worse animal if it tried. I have made the individual version of this argument about leading people who know more than you do.
Set the why and the what. Let them own the how. The organizational version is the same discipline at scale. In our model, the functions are the nervous system. The work underneath them sits centralized or distributed based on what that work actually requires, decided function by function rather than by a philosophy applied to the whole enterprise.
That distinction sounds small, and it is the entire thing. Most reorganizations pick a side. Centralize everything, and you get a fast head with numb arms. Decentralize everything, and you get a hydra. This design refuses the question.
It also resolves the argument that eats most shared services models alive. Centralized expertise does not require centralized decision-making. A specialist can belong to one institutional function while being accountable for the outcome of one specific enterprise. Same person, same function, different accountability. That is how a nervous system stays connected without forcing every decision back through the head.
A body without a skeleton. An octopus can pass through any opening larger than its beak. Everything else conforms. Fixed cost structure, physical footprint, credit hour orthodoxy, academic calendar; all of that is skeleton, and skeleton is what determines which openings you can enter. We did not eliminate ours. We stopped letting it decide where we were allowed to go.
The only hard part
If everything is soft, something has to be hard, or you do not have an animal. You have a puddle. Our beak is small, and it does not negotiate on four things.
Governance model and Academic quality that would survive any legitimate review we did not schedule ourselves.
Transparent pricing, meaning the number a learner is quoted is the number a learner pays.
Measurable learner outcomes, published whether or not they flatter us.
One source of institutional data, which every enterprise reports into whether it wants to or not.
Everything else can move. Market, delivery, staffing, brand, price point, partner, calendar, location, degree level, credential, name on the door. Any venture that fits through those four can become a tentacle (We call them SEBUs). Any venture that does not is not an exception, a pilot, or a special case with its own reporting line. It simply is not us.
The model only becomes credible at the point where you can see where the flexibility ends. Most institutions have this backwards. Their academic structure is rigid and their standards are negotiable.
Education is a layer, not a destination
This is the part that has nothing to do with colleges.
Education is not a place people go. It is a layer that belongs inside industries, independent but relevant, rather than adjacent to them. The moment you accept that, the entire ecosystem redraws itself, and the family of brands stops looking like a marketing decision and starts looking like the only sane response to the fact that learners in different industries do not want the same animal.
Employers understand this faster than institutions do, once you put it plainly. What they call a talent shortage is a design flaw, and nobody is arriving to fix it for them.
Watch how a single loop closes. A working professional enrolls through an employer that needs a specific capability. The employer’s own operation becomes the applied setting for the coursework, which means the curriculum is being validated in the field while it is being taught.
The learner finishes and stays with that employer, now more valuable and more likely to send the next colleague. The best of those graduates come back as practitioner instructors, which lowers our cost of instruction and raises its credibility at the same time. What the employer learns about its own capability gaps becomes the next program we build.
The learner is also the workforce. The employer is also the customer. The graduate is also the instructor. The operation is also the classroom. The content is also the product.
Pipelines behave nothing like this. A pipeline has an opening you must keep filling from the outside, at rising cost, against everyone else fishing the same water. A loop can generate a meaningful share of its own demand, though only if it is designed to and only if somebody owns the closing of it. Loops do not close by accident, and an unclosed loop is just a pipeline with better vocabulary.
We are not finished building this. The data ecosystem is real and still under construction. Our learner-facing tools are working and unfinished. Una was the first Agentforce-undergirded agentic AI deployed at a university at this scale in this country, and I will still tell you that Una is early. I intend to keep describing our tools that way until they are not, which is not remotely the same as conceding ground to people who dismiss it all. Announcing a completed animal is how you end up a butterfly again.
What the model makes possible
Mission may attract capital. Structure determines whether capital can accomplish anything. That is the sentence I wish someone had said to me fifteen years ago.
So here is what the structure does.
The defensible asset is not the brand, and it was never the real estate. It is the marginal cost of the next tentacle. In a hydra, entering a new market means growing a new head, with its own leadership, its own overhead, and its own opinion about strategy. Here it means extending a tentacle off a nervous system that already exists and has already been paid for. One of those compounds. The other gets more expensive and harder to govern with every addition.
The technology stack is leverage, not the asset. Tools get rented, outgrown, and replaced. The design and the data are owned. We wrote that into how we govern Project Stratus. Systems support the operating model. They do not define it. The moment a platform starts deciding your organizational chart, you are organized around vendor logic and paying for the privilege.
Governance is an asset class that almost nobody prices. A board that works instead of opines determines your speed, and speed determines everything else. Most institutions in this sector cannot move because their governance was built to prevent movement, and that is a permanent discount on every other asset they hold.
And the model is adaptive by construction rather than by intention. I did not build against a forecast. The forecasts I was handed when I started described a world that never showed up. I built something that can add a tentacle, or retire one, without a crisis, a consultant, or a chrysalis.
Why the Octopus?
The butterfly taught us how to survive the destruction of an old form. That mattered. It got us here.
But an institution cannot enter a chrysalis every time the world moves. Eventually it runs out of time, or money, or the courage to dissolve itself again.
The next institution will not win because it completed the most dramatic transformation. It will win because it never needs one.
The butterfly got us here…. What comes next must never be finished…
One more thing, because someone will ask what this scales into. Everything above describes a single animal. The harder version of the question is what happens when that nervous system runs underneath institutions that are not ours. Each one keeps its name, its mascot, its traditions, and the texture that made someone choose it. What they share is the machinery no student should ever have to think about. I laid out that architecture separately in what a private university system actually requires. Sovereign in everything students experience. Shared in everything they do not.
That is the Octopus Model, and that is what our
Enterprise Model aims to become!


Melik, the last section is the one that matters most: the nervous system running underneath institutions that aren't yours. Sovereign data infrastructure and governed agents sit upstream of the transaction systems, so institutions keep their name, their mascot, and their data while sharing the machinery no student should ever have to think about. Your beak and our foundation layer are the same commitment... one source of institutional data, no exceptions. Unity is further down this road than anyone I've seen. Love to reconnect.