Why the next phase of AI adoption will reward structural thinking over speed
When I first began working closely with AI companies a few years ago, it was clear to me that AI would eventually transform home-based care in a meaningful way. What I misjudged was the timeline.
I assumed it would take years for real movement to emerge. Instead, the conversation accelerated and quickly polarized. Leaders are often framed as either bold adopters or cautious laggards, and many operators I speak with still default to the “wait-and-see” approach that has defined much of healthcare innovation for decades. That instinct is understandable.
Technology has delivered incremental gains, but it has rarely transformed care delivery the way we once imagined. As a result, many assume the defining decision today is simply whether to embrace AI.
But in conversations across healthcare and other labor-intensive industries, a different reality is emerging. Most leaders are not debating whether AI matters. They are grappling with where it fits and how to introduce it without compounding the operational complexity they are already managing.
Balancing Pressure from All Sides
They are also navigating pressure from both directions. Boards and investors are pushing for urgency and a clear AI strategy. Frontline teams often carry hesitation rooted in fear, uncertainty, or skepticism about whether AI can improve outcomes without destabilizing the workforce.
Meanwhile, internal teams are experimenting independently, vendors are multiplying, and private equity platforms are consolidating agencies while simultaneously pushing for innovation to unlock value across portfolios. Research from McKinsey & Company shows that 62% of organizations are already experimenting with AI agents in some form. In that environment, urgency naturally drives action. Most organizations respond in predictable ways: launching pilots, testing tools, and layering automation onto existing workflows.
Yet despite the activity, the outcomes are often underwhelming.
In many cases, organizations are not even sure what success should look like. Instead of step-change performance, leaders encounter fragmented systems, operational fatigue, and modest returns. Not because the technology lacks promise, but because it is introduced without a structural context.
Why Point Solutions Fall Short
Healthcare leaders are accustomed to point solutions that aim to make individual operational workflows more efficient. As a result, technology ownership has historically been fragmented across HR, compliance, operations, and clinical teams, with IT often acting as gatekeeper rather than decision-maker. Now, CIOs are suddenly being asked to “own AI strategy,” often without clarity on what that truly means.
The organizations gaining traction are approaching this moment differently. They are reframing the central question. Rather than asking, “Which AI tools should we deploy?” they are asking, “Where is our operating model fundamentally constrained?”
Fragmented Workflows = Fragmented Foundation
Nowhere is this more evident than in distributed, labor-dependent sectors like home-based care, where performance is shaped by deeply interconnected workflows. Here, two lifecycle systems dominate outcomes:
- Attraction to Retention: The full lifecycle of the caregiver or clinician, spanning recruitment, onboarding, intake, scheduling, and retention.
- Referral to Bill: The end-to-end pathway from referral generation and intake through case conversion, staffing, and ultimately insurance reimbursement.
These lifecycles intersect at staffing, where workforce supply meets patient demand. Optimizing a single step within either lifecycle rarely produces lasting value. More often, doing so just shifts friction elsewhere in the system.
Take intake in home health, the process providers go through to bring on new patients. For years, tools have aimed to streamline referral conversion and reduce friction between sales and administrative teams. Yet the same frustrations keep surfacing. When discussions go deep enough, the true constraint often lies elsewhere, in documentation workflows that remain structurally misaligned.
Where Structural Misalignment Becomes Opportunity
The structural misalignment of AI is where the opportunity lies. Research from Deloitte has found that organizations generating meaningful value from AI tend to redesign workflows and operating structures rather than simply layering automation onto existing processes. AI becomes a fundamental structural layer, rather than a technological accelerator.
Organizations that treat AI as a systems-level redesign opportunity are beginning to unlock more meaningful advantages. Instead of chasing isolated use cases, they align adoption around lifecycle thinking.
Often, speeding one step in a workflow just shifts the bottleneck downstream. Administrative complexity remains one of healthcare’s most persistent operational challenges; research from Edge shows more than three quarters of healthcare leaders report high administrative workload, and over half experience operational backlogs on a weekly or daily basis. Referral intake may double in speed, yet clinical documentation still requires calls to a doctor’s office. More recruiting prospects may surface, but a single recruiter responsible for follow-ups slows conversion.
The conversation shifts towards asking how AI can relieve pressure across the entire workforce or client journey, rather than automating individual tasks.
Designing for the Next Phase of AI
This type of thinking prioritizes architecture before vendors, outcomes before pilots, and clarity before experimentation. It also requires leadership teams to resist the instinct to move quickly solely because the technology is advancing quickly.
In many ways, this moment mirrors earlier enterprise inflection points. Early adopters optimize for speed. Enduring leaders optimize for design.
Over the next several years, competitive advantage will likely accrue less to organizations that implemented AI first and more to those that integrated it most coherently. The differentiator will not be access to tools, but the quality of the operating model into which those tools are introduced, and therefore the value the tools can provide. Technological shifts reward momentum in the short term. But over time, thoughtful design compounds in ways reactive adoption rarely does.
That work requires discipline, structural clarity, and senior leaders willing to own the design of the system itself, not just the decision to invest in AI. It also requires choosing partners who are willing to understand the operating model and co-design around it, not simply deploy into it. The technology will keep advancing. The question is whether your organization is building something that can absorb it.
