Healthcare does not have a shortage of technology.
There are more patient portals, navigation platforms, scheduling systems, engagement tools, artificial intelligence products, virtual care companies, and electronic health record capabilities than any healthcare leader could reasonably keep track of. Every year brings another wave of technology promising to make care more connected, personalized, accessible, efficient, intelligent, or human.
And yet the experience of receiving care remains remarkably incoherent.
The familiar explanation is that healthcare is unusually complex. It is regulated, fragmented, expensive, operationally difficult, and dependent on technology that cannot easily be replaced. All of this is true. Healthcare may be one of the most complex systems we have ever created, responsible for serving people across the complete spectrum of human life, from routine checkups to the most vulnerable moments we will ever experience.
But complexity has also become an alibi.
It has become a way for healthcare institutions to describe conditions they have gradually accepted as beyond their control. The electronic health record works a certain way. The vendor does not support a particular capability. One department owns scheduling while another owns the patient portal. The clinical operation cannot accommodate what the digital team wants to build. The organization responds to each constraint individually until, over time, those constraints become the product strategy.
No one explicitly chooses the resulting experience. It emerges from procurement decisions, legacy systems, internal politics, operational boundaries, and compromises made across different parts of the organization.
The patient nevertheless experiences it as one system.
They experience it when they are asked to choose a medical specialty before they understand what kind of care they need. They experience it when the scheduling tool shows no appointments and offers no meaningful next step. They experience it when a chatbot answers the question it can recognize instead of the one they are actually asking. They experience it when they have to repeat the same information to multiple people because one part of the system has no knowledge of what happened in another.
The organization may not have intentionally designed any of this. But refusing to choose is still a choice.
The deeper problem in healthcare is not that institutions lack technology. It is that many have surrendered the responsibility to decide what their technology should make possible.
When procurement becomes product strategy
Healthcare institutions rarely set out to let vendors define their care model. It happens more subtly than that.
An organization purchases a platform to solve a particular operational problem. The platform comes with its own capabilities, limitations, terminology, workflows, and assumptions about how people should behave. The organization configures what it can, works around what it cannot, and gradually adjusts its expectations to fit the technology it bought.
Eventually, the most important product question is no longer “What should the experience of receiving care be?” It becomes “What can the platform support?”
This is understandable. Healthcare organizations cannot build everything themselves. They should not build everything themselves. Buying technology is often faster, more economical, and more responsible than developing and maintaining it internally. The problem is not the use of vendors. The problem is allowing the capabilities of a vendor to determine the boundaries of institutional ambition.
A vendor can provide infrastructure, software, expertise, and speed. It cannot decide what the healthcare institution believes it owes the people it serves.
It cannot determine how much complexity the institution should absorb on behalf of the patient. It cannot decide when a person needs direction instead of more options. It cannot define the point at which automation should give way to human judgment. It cannot decide what kind of relationship the institution wants to create with someone trying to receive care.
Those are not software decisions. They are institutional decisions that become visible through software.
When the institution does not make them, someone or something else will. The electronic health record will make them. The vendor roadmap will make them. The organizational chart will make them. The most urgent operational demand will make them. The accumulated weight of everything that already exists will make them.
This is how procurement quietly becomes product strategy. Not through one dramatic decision, but through hundreds of smaller decisions that gradually narrow what the organization believes is possible.
The patient carries what the institution does not resolve
This may be one of the most important questions we can ask about a healthcare product: who carries the complexity?
Healthcare is complex because health itself is complex. Medicine, insurance, regulation, clinical operations, and the coordination of thousands of people across thousands of circumstances are all complex. We cannot simply design all of this complexity away. But we can decide where the complexity goes.
It can be absorbed by the product, handled by an operator, interpreted by a clinician, distributed across a coordinated system, or pushed onto the patient.
Too often, healthcare institutions push it onto the patient because resolving it internally would require difficult organizational decisions. We use clinical or insurance terminology because creating a shared language would require multiple departments to agree. We show people an exhaustive directory of providers because recommending one would require us to take responsibility for the recommendation. We make patients coordinate between different parts of the system because the systems themselves cannot communicate.
We call this self service. We call it choice. We call it giving patients more control.
Sometimes that is exactly what it is. A well designed product can give people greater independence, clearer information, and more agency over their health. But choice without meaningful direction is not necessarily empowering. Giving someone every possible option can be another way of declining to help them decide.
On paper, the product may still work. A patient can search the directory, complete the form, call the number, and eventually find an appointment if they understand what to search for, have enough time, and are willing to keep trying. Every individual transaction may be functioning exactly as designed.
But being able to complete a task does not necessarily mean someone was cared for.
A product can be technically usable while still asking too much of someone who is sick, frightened, exhausted, or unfamiliar with how healthcare works. A system can be operationally successful while transferring its most difficult work to the person least equipped to perform it.
Every unresolved institutional decision eventually becomes someone else’s burden. In healthcare, that person is often the patient.
Consumer technology is not a care model
The technology industry has trained us to think about product quality through the language of consumer experiences. Healthcare products should be easy, fast, personalized, intuitive, and convenient. They should reduce friction, increase engagement, and allow people to accomplish more through self service.
These are valuable ambitions. Healthcare has much to learn from the best consumer products. A person should not need specialized training to book an appointment, understand a benefit, or receive an answer to a basic question. There is no virtue in making an experience more difficult simply because healthcare itself is complicated.
But the consumerization of healthcare has limits.
A patient is not always a consumer making a preference based purchase. They may not know what they need. They may not be able to judge the quality of the available options. They may be making a decision under stress, pain, fear, or uncertainty. They may not have the freedom to walk away from the transaction. The consequences of a poor recommendation can be far more significant than an abandoned shopping cart.
Healthcare cannot simply imitate shopping, travel, media, or financial technology and declare itself patient centered. The relationship is fundamentally different because the responsibility is fundamentally different.
Removing friction is not enough. Some moments require deliberation. Some decisions require expertise. Some situations require a person to slow down, ask another question, or involve someone with more judgment. A product designed only around speed and completion may become very effective at moving people through the wrong experience.
The goal cannot be to make healthcare feel indistinguishable from every other consumer transaction. The goal should be to use technology in a way that is appropriate to the gravity and particularity of care.
That requires more than usability. It requires a philosophy.
The front door is not a feature
The front door of healthcare is where a person first attempts to turn a need into action. It may begin with a search box, a phone call, a scheduling flow, a chatbot, or a message to a care team. But beneath the interface is something far more consequential: a set of decisions about how the institution will understand the person and what it will do in response.
What questions will it ask? What information will it use? How will it interpret what the person needs? Where will it send them? When will a human become involved? What happens when the expected path fails?
These decisions determine who receives care, how quickly they receive it, and whether they reach the right place. They influence the distribution of clinical resources and the work expected of patients, clinicians, and operators. They shape whether the institution behaves as one coherent system or as a collection of disconnected services.
The front door is not merely a channel. It is institutional infrastructure.
This becomes even more important as artificial intelligence moves closer to the beginning of the care journey. A healthcare chatbot may begin as another way to retrieve information. But once it starts interpreting needs, assessing urgency, recommending a type of care, selecting a provider, or determining when a human should become involved, it is participating in the care model.
The most important question is not whether the chatbot can generate an accurate answer. It is what responsibility the institution is allowing it to assume.
The model cannot answer that question for us. Neither can the vendor that provides it. Artificial intelligence can help an institution deliver its care model, but it cannot tell the institution what kind of care model it should believe in.
If a healthcare organization gives away the logic that determines how people are understood and where they are sent, it may be giving away more than a digital experience. It may be surrendering control of the relationship itself.
Efficiency for whom?
Healthcare institutions have real operational and financial constraints. People, time, capacity, and money are finite. A care experience that is humane but impossible to sustain is not a successful care experience.
Efficiency matters. But whenever we say a product is more efficient, we should ask a second question: more efficient for whom?
A digital intake form might save an employee ten minutes while asking the patient to find and enter information they do not understand. An automated denial might resolve a case quickly while forcing the person to call three different departments to learn what happened. A self service experience might reduce support volume because people can complete common tasks independently, or it might reduce support volume because finding a human has become nearly impossible.
An organization can become more efficient by making the patient do more work.
That is not always innovation. Sometimes it is simply a transfer of labor.
This does not mean automation is inherently uncaring or that every healthcare interaction needs a human. Automation can remove tremendous amounts of unnecessary work. A great product can give people faster answers, clearer choices, and more control over their health. But efficiency cannot only mean that the organization spends less or processes more transactions.
A healthcare product is truly efficient when the entire system works better, including for the person the system exists to serve.
The difference matters because every efficiency creates a theory about whose time is valuable, whose work is visible, and whose burden is acceptable. Those are not only operational judgments. They are moral judgments expressed through the design of the system.
Healthcare institutions must become authors again
When healthcare teams begin new product initiatives, they often begin with the thing they believe they need to build. We need a new app. We need a chatbot. We need to redesign the portal. We need a better scheduling experience.
These statements sound like direction, but often they are solutions standing in for questions that have not been answered.
What problem is the institution taking responsibility for? What should the person be able to understand or accomplish? Which parts of the experience must the institution own? What promise is it making? What should happen when the ideal path breaks? What kind of relationship does it want to create with the people it serves?
Ultimately, what kind of care model is the institution trying to make real?
These questions have to come before the interface because the interface will eventually answer them, whether we answer them intentionally or not. This is why design is most useful before the answer feels obvious. Design can help an institution explore different futures and expose the assumptions living inside presentations, strategy documents, operating models, and people’s heads.
A prototype is not valuable merely because it makes an idea look real. It allows us to experience the consequences of an idea before we build the entire system around it. It shows what the future will ask of the patient, the clinician, the operator, and the business. It reveals whether an idea that sounds coherent in a conference room remains coherent when a real person has to move through it.
This is design as a form of institutional authorship.
The failure of healthcare technology is often described as a failure of usability, interoperability, incentives, or innovation. All of these failures are real. But beneath them is a more uncomfortable failure: too many healthcare institutions no longer behave as authors of the experiences they create.
They procure technology, implement workflows, measure transactions, and optimize individual parts of the system. But they have not always decided what the whole should amount to or what obligations they are willing to assume on behalf of the person seeking care.
Those obligations cannot be delegated to an electronic health record, a chatbot, a design system, or a vendor roadmap. Technology can embody a care model, but it cannot tell an institution what kind of care it should believe in.
That decision belongs to the institution. It always has.
The future of healthcare will not be determined by which organizations acquire the most advanced technology. It will be determined by which ones recover the conviction to decide what their technology is for, the imagination to see what could exist beyond their current constraints, and the courage to take responsibility for what they make real.