From Referral Network to Operating Network
How a designed joint venture became a weekly operating layer — and patient engagement moved from 22.6% to 37.1% and held.
37.1%
Patient engagement, up from 22.6%
At or above 36% for seven consecutive months (Oct 2025 → Aug 2026). The network's reported result, not a projection.
74
Consecutive weekly status reports
Decisions on Monday, a written status every Friday — a 74-week record that serves as institutional memory.
537%
Direct Access capacity growth
Monthly appointment slots grew from 391 to 2,489 — managed as inventory, not counted as completed visits.
81.8%
Timely Access, up from 67.1%
Average time to first appointment stayed under the 7-day standard in all eleven months.
The network’s reported results, not a projection. The health system and participating providers are not named. An engaged patient received a referral and attended the referred provider at least once; the definition was the same across all months reported.
The handoff was the problem
An established behavioral health network was already referring patients, but too few of those referrals became visits. Behavioral healthcare breaks down most often at the handoff: a patient is referred, an appointment may be offered, and then the connection is lost.
In October 2025, 22.6% of referred patients were engaged in care. The network worked. But running it took more than an understaffed team could do every week — provider-by-provider performance management, follow-through on individual referrals, capacity planning, onboarding, and governance reporting. Much of that work wasn’t getting done, so referrals went out and too many never became a visit. The constraint was not the network. It was management capacity.
The same team that designed it, ran it
Aveniq was brought in to design BHNewCo as a corporate joint-venture spin-off meant to break a 30-year behavioral-health stalemate. Over six months the team built the venture’s brand, financial logic, and operating logic. All of it lived in Jumpscript — the cohesion layer that held the interviews, operating logic, financial logic, and stakeholder landscape as one working context.
In October 2025, the work moved from designing the venture to running the core network. The context was already organized and the cadence was already running: decisions on Monday, a written status report on Friday. The 74 weeks are one record, not two projects.
A hands-on model with six parts
What changed was how the network was run. Each part of the weekly model managed a specific asset.
Providers
Every provider held to the same standards, plus workflow diagnosis, review of individual referrals, and 1:1 coaching. Patterns that didn’t change were escalated to executives.
Slots
Direct Access capacity matched to real service, payer, location, and population demand. Slots were managed as inventory, not counted as appointments.
The funnel
Acceptance, scheduling, outreach, attendance confirmation, and loop closure run as one job instead of five handoffs. When a referral leaked, it was traced to the organization responsible.
Users
New users, permission changes, platform refreshers, and provider-profile upkeep handled every week as routine operations.
Governance
Provider performance reviewed in Clinical Operations. Sensitive issues went to the Network Committee and the Board.
Signal
A formal monthly close — with matured values, rolling averages, and provider drill-downs — separated real movement from temporary noise.
The stalemate didn’t break on brand. It broke on conversion.
Volume
Four providers handled 73.3% of referrals — about 325 a month. A small improvement at those four moved the network average more than a large change at a low-volume provider could. Management attention went there first.
A slot is not a visit
Direct Access capacity grew more than sixfold — largely from two major providers opening capacity and a new provider joining. But a slot only has value when it is visible, matched to the right patient and payer, used, and accepted. Management is what turned that inventory into visits.
Engagement did not cost access
Extra outreach that led to attended care was worth a small cost in access time. Average time to first appointment stayed under the 7-day standard in all eleven months.
Eleven months, five phases
Oct–Nov 2025
Diagnose
Identified the high-volume providers that drove engagement, how they performed, and where referrals failed to convert.
Baseline 22.6% engagement · 391 Direct Access slots
Dec 2025–Jan 2026
Build
A major provider expanded Direct Access capacity; a new provider joined; engagement became the network’s top measure.
24.1% engagement in January
Feb–Mar 2026
Inflect
A second major provider opened significant capacity; governance incentives were aligned; warm handoffs took hold.
36.2% engagement in February
Apr–May 2026
Stabilize
Referral deep dives, 1:1 coaching, platform training, provider-profile cleanup, and tighter loop closure.
Timely Response reached its 90% goal
Jun–Aug 2026
Institutionalize
Performance governance was made durable so the gains would hold without heroics.
37.1% in August; at or above 36% since February
October 2025 → August 2026
| Measure | Oct 2025 | Aug 2026 | Change |
|---|---|---|---|
| Patient Engagement | 22.6% | 37.1% | +14.5 pts≈64% relative gain |
| Timely Access | 67.1% | 81.8% | +14.7 pts |
| Timely Response | 80.6% | 90.2% | +9.6 ptsat or above the 90% goal four straight months |
| 7-Day Follow-Up | 67.3% | 81.8% | +14.5 pts |
| 30-Day Follow-Up | 72.7% | 86.7% | +14.0 pts |
| Declination Rate | 4.5% | 3.5% | −1.0 pt |
| Avg. Time to First Appointment | 3.9 days | 3.5 days | under the 7-day standard all 11 months |
| Direct Access Monthly Capacity | 391 | 2,489 | +537%available slots, not completed visits (Nov 2025 → Sep 2026) |
| Hospital Avg. Length of Stay | 3.29 days | 2.50 days | −24%reported alongside the engagement lift; not isolated as a single cause |
The engine produced the scorecard. It did not have the conversation.
In May 2026, an engine the team built took over quarterly provider quality scoring, and scoring time fell by about 75–80%. That same week, people were still in corrective conversations about intake. Recurring production got faster; the relationships still needed people.
Reporting changed the same way. The performance view moved from PowerPoint to an interactive dashboard that could be refreshed in about five minutes and opened to the provider. The team stopped rebuilding the deck — and spent the time on the work only people can do.
A weekly operating layer, and a capability that stays
What moved was the connection between a referral and a visit. It wasn’t a new directory or a larger team. It was a weekly operating layer: capacity treated as something to manage, attention spent where the volume was, referral follow-through treated as one job, and a small senior team working from context the organization already had.
The capability stays with the network: the dashboard, the scoring engine, the monthly close, and a 74-week written record that serves as institutional memory. Seventy-four weeks in, the gains have held — and engagement is still short of the network’s 40% goal. The open question is how to make that management capability permanent and keep going.
Aveniq builds this with your team, around the knowledge you already have: a working context, a weekly cadence, and people accountable for the relationships and the follow-through. The rhythm can start quickly. This result was built over 74 weeks of it.