The Unified Interface Thesis

The Engagement Myth

The conventional wisdom says patients won't engage. They're treated as passive. They ignore apps. They end up in the ER for something urgent care could handle. That story is wrong.

Patients don't fail to engage because they're lazy or irrational. They fail because we built a system that requires them to be expert navigators of a fragmented mess, and then blamed them when they weren't.

The Fragmentation Problem

The average employer-sponsored health plan now includes five to ten point solutions: MSK, diabetes management, fertility, surgery navigation, mental health. Each has its own app, login, outreach cadence, and logic. We've given patients ten doors and expected them to know which one to open, when, and why.

Most don't. Most can't. That isn't a failure of will. It's a failure of design.

The Timing Problem

Point solutions depend on claims data to know when to intervene. But claims arrive after the fact. The surgery is already scheduled, the PT is booked, the prescription is filled. By the time outreach happens, the moment for steering has passed.

That isn't steering. Steering means being present at the moment of decision. Outreach after the fact is just commentary.

Health Insurance Isn't a Product

A product can be evaluated, compared, understood. Health insurance can't, not because it's complex, but because it was never designed as a whole. It's a stack of contracts, exclusions, network arrangements, and vendor partnerships that piled up over decades. No one architected it.

When we ask patients to "navigate their benefits," we're asking them to navigate something that isn't navigable. You don't get a map. You get a box of coupons to places you've never heard of, half of which have expired, none of which are indexed.

Why Navigation Platforms Fall Short

Navigation platforms were supposed to fix this: one front door, one phone number, one place to go. Instead they became another layer, not the unifying layer. Another app to download, another vendor relationship, another per-member-per-month line item.

They can point you at a door. They can't remove the door. The navigation layer doesn't control the underlying infrastructure. It can't bind point solutions into a unified experience with shared context, shared data, and shared logic.

What a Unified Interface Actually Requires

A unified interface isn't a portal with better design. It's a different architecture, and it needs three layers:

• Data Connectivity: Claims, eligibility, network status, PA decisions, accumulator data, care gaps. The system has to know where a patient is in their benefits right now.
• Decisioning Logic: Given a patient's situation, where should they go? What's clinically appropriate, financially optimal, and actually accessible?
• Consumer Interface: If the first two layers work, this becomes simple. The patient asks a question and gets an answer that accounts for everything the system knows.

The Dependency Chain

Those three layers aren't independent pieces. They're a sequence, and each one depends on the one beneath it: Data → Decisioning → Interface. You can't build the interface without the decisioning logic. You can't build the decisioning logic without the data. You can't skip steps.

Most solutions have one layer. Some have two. Almost nobody has all three connected. That's why we're still talking about engagement problems years into the point solution era.

Who Should Build This?

Legacy health plans have the data but can't move. Decades of technical debt, acquisitions that were never integrated, and a fundamental lack of member trust get in the way. Point solutions have engagement, but only for their slice. Navigation platforms have the ambition but not the infrastructure. Employers have the checkbook but not the capability.

Everyone has a reason to want the unified interface. Everyone has a reason to keep someone else from owning it. Whoever builds it owns the patient relationship, which is the strategic high ground for the next decade of healthcare.

The Stakes

Steering fails not because patients are unwilling. It fails because the architecture makes success nearly impossible. Too many apps. Too late to intervene. Too hard to understand.

The unified interface isn't a better app. It's the precondition for steering to exist at all. Without it, point solutions stay underutilized, navigation platforms stay approximations, and patients keep defaulting to the path of least resistance. Not because they're irrational, but because we made the rational path unreasonably hard.