By: Shannon Werb, Chief Executive Officer at Array Behavioral Care
Over the past decade, we’ve watched mental health care undergo a rapid transformation. Telehealth expanded access at unprecedented speed, and contractor-based, gig-style clinician networks played a meaningful role in that early acceleration. Having spent decades building and scaling clinician networks across multiple specialties, both in telehealth and in‑person settings, I’ve seen firsthand how contractor-based models can accelerate access in the right contexts. While it might be tempting to apply that playbook broadly, including mental health, I’m convinced the underlying care structures won’t support it. As the market matures, and as expectations around quality, outcomes, and accountability rise, it’s becoming increasingly clear that not all care models are equally suited for the realities of behavioral health.
This isn’t an ideological argument, and it’s not anti-technology. It’s a structural one. Behavioral health is fundamentally different from other forms of care, and the way we design care delivery models matters deeply. The growing “gigification” of mental health introduces risks that are often underappreciated, especially when contrasted with the advantages of a practice-based, employed clinician (W2) model.
The Structural Challenges of Gig-Based Mental Health Models
1Continuity of care is structurally weak
Effective behavioral health treatment depends on sustained relationships between patients and clinicians. Trust begins at the time of initial access, clearly, consistently, and transparently. Trust builds over time. Clinical insight deepens through continuity. Yet contractor networks, by design, optimize for availability, often at the expense of durability over time.
High clinician churn, fluctuating schedules, and loose affiliation make it difficult to ensure that patients see the same clinician they chose, initially or consistently, particularly those with higher acuity, comorbidities, or complex social needs. When care becomes episodic rather than continuous, outcomes suffer, and patients are more likely to disengage when they need support most.
2Accountability for outcomes is diffuse
In contractor-heavy models, clinicians operate as largely independent units. When something goes wrong, when a patient deteriorates, drops out of care, or presents in crisis, ownership of that outcome is often unclear.
Who is responsible for monitoring progress? For identifying early warning signs? For managing transitions between levels of care?
When accountability is fragmented, gaps emerge downstream. Those gaps often show up as emergency department utilization, inconsistent quality, or avoidable lapses in follow-up. Without a clear owner of outcomes, it’s difficult to systematically improve them.
3Risk management and escalation are harder to operationalize
Mental health care inherently involves clinical risk. Suicide risk, substance use, and acute psychiatric deterioration require more than policies and checklists; they require real investment from teams, supervision, and shared responsibility.
Contractor-based models typically rely on handoffs and asynchronous protocols rather than embedded clinical teams. That makes consistent escalation pathways harder to execute in real time. Supervision can be uneven. Decision making becomes isolated. In moments where coordination matters most, the system is least equipped to respond.
4Integration with health systems is shallow by design
There is an important distinction between networks and practices. Networks connect clinicians to patients. Practices integrate clinicians into care systems enabled through true investment in clinical care programs, models and systems.
Many contractor-driven models sit adjacent to hospital, payer, and primary care workflows rather than being embedded within them. That limits care coordination across settings and acuity levels. Data doesn’t flow cleanly. Communication breaks down. The result is parallel care, not integrated care, which is an increasingly unacceptable tradeoff as health systems demand tighter alignment and shared accountability.
5Ghost networks are an inevitable byproduct
Large contractor networks can look robust on paper, listing thousands of clinicians across geographies and specialties. In practice, availability is often far more limited.
Even well-run clinician networks experience substantial churn, often losing roughly a third to half of participating clinicians in a given year as contractors cycle in and out. Practice-based models, by contrast, are intentionally designed to invest in retention, supervision, and long-term career durability, and have consistently demonstrated meaningfully lower turnover as a result. In contractor‑heavy models, misaligned incentives can lead to inconsistent participation over time, contributing to “ghost networks” in which theoretical capacity far exceeds real‑world access. Regulators and payers are increasingly focused on this gap, scrutinizing whether listed networks can reliably serve patients in a timely and sustained way.
Why the Practice-Based, W‑2 Model Has a Durable Advantage
1Employment enables real clinical governance
Employing clinicians allows for true clinical governance: standardized training, ongoing supervision, peer review, and clear escalation pathways. It creates an environment where quality is managed proactively, not retroactively.
This isn’t about control; it’s about support. Clinicians practice better when they’re part of a team that invests in their development and holds shared responsibility for patient care.
2Practices support true continuity of care
In a practice-based model, clinicians own patient panels. Care teams are designed intentionally. Coverage is coordinated. Patients aren’t starting over every few months with a new provider.
That continuity matters, not just for outcomes, but for trust. It’s foundational to effective mental health treatment.
3Capacity planning is based on real availability
4Integration is foundational, not optional
Practice-based models are built to integrate with EMRs, hospital workflows, and payer infrastructure. Collaboration across care settings isn’t an afterthought; it’s a requirement.
That level of integration enables better care coordination, smoother transitions, and clearer accountability across the continuum of care.
5Accountability is explicit and durable
Designing for the Reality of Behavioral Health
The rise of gig-style mental health platforms was a response to a real access crisis, and in many ways, it helped move the field forward. But inconsistent access alone is not the finish line.
Behavioral health requires structures that support continuity, accountability, and trust over time. As the market matures, models built around practices, not just networks, are better aligned with that reality.
This isn’t about rejecting innovation. It’s about designing care in a way that reflects the seriousness of the work and the needs of the patients we serve. In mental health, how we build matters just as much as how fast we grow.
