Madisonhealth

Madison Memorial Hospital serves Eastern Idaho, and a hospital website carries a burden most sites don't. People arrive stressed, often one-handed on a phone, looking for exactly one thing — a provider, a department, a location, a phone number. Nobody browses a hospital site for pleasure.

The site already existed and already worked in the narrow sense. The problem was everything around that: finding a specific provider took more steps than it should, the structure didn't match how patients actually search, and the experience degraded on exactly the devices and connections most people were using.

The scale is easy to underestimate. This isn't a practice site with a services page. It carries provider profiles, specialties, multiple locations, a patient portal, online bill payment, careers, community CPR and Basic Life Support courses, and a hospital café — each serving a different person arriving for a different reason. The home page alone links to seventy internal destinations.

Which is why designing for “the user” doesn't survive contact with a hospital. A patient hunting a phone number, a family checking visiting hours, a physician making a referral, a nurse looking at jobs and a resident booking a CPR class are five different errands. A structure that serves one of them well tends to bury the other four.

I rebuilt the front end around wayfinding over presentation. Provider search, departments and locations come out from behind the menu and onto the first screen. The information architecture follows the questions patients arrive with, not the hospital's internal org chart — those are rarely the same shape.

Accessibility was treated as a requirement, not a checklist item. Healthcare serves people with visual, motor and cognitive impairments at a higher rate than the general population, and public-facing healthcare carries real obligations here. Semantic structure, keyboard paths, contrast and focus order were part of the build from the start.

Performance mattered more than usual because of where this site gets used: hospital wifi, rural mobile connections, older devices. Core Web Vitals work targeted the conditions the actual audience is on, not a fast desktop connection in a dev environment.

The routine paths surface early. Provider search, locations and specialties sit near the top, because that's what most visits are for. Billing and the patient portal are one step away for the same reason — those are errands, and every extra click on an errand becomes a phone call to a desk that's already busy.

And it had to stay maintainable by the hospital. A site with this many moving parts goes stale quickly if every change needs a developer. Provider listings change, departments reorganize, courses get scheduled. The build had to leave the team able to do all of that without touching layout.

A site where patients and staff find what they need without hunting for it — faster, more accessible, and structured so the hospital's own team can keep it current without a developer in the loop for routine changes.

It's one of two healthcare builds on this site, and the one behind the phrase “hospital-scale” used elsewhere. The complexity isn't visual; it's the number of audiences, departments and regulatory considerations that have to coexist in one navigable structure.

It's also local. Madison Memorial serves Rexburg, where I live and work, and it's one of five businesses inside the town I've built for. That's a different kind of accountability from a client three time zones away — the people using this site are the people ahead of me in the queue at the grocery store.

The hard part was approvals. Too many people had to agree before anything moved, and nobody could say yes on their own, so anything touching more than one department sat until enough of them did. The architecture was the easy half. Worth knowing before you scope work for an organization this size.

What's deliberately not on this page is a number. No traffic percentage, no ranking claim, no conversion lift. Those would need measurement I don't have for this build, and a hospital is the last place to start estimating. What's here is what was decided and why, which is the part that transfers to your project regardless.