Operations is where promised denial reductions become real, where credentialing cycles get measured in days not months, and where AR aging compresses one Friday review at a time. Built on Musk's deletion discipline and Jensen's flat-structure signal — designed to execute, not coordinate.
CHS Operations symptoms look different across credentialing and RCM, but the root cause is the same in both: workflows lack disposition discipline. Every claim, every payer application, every denied appeal needs an explicit named outcome with an owner and a date — or it drifts. Drifted work doesn't get billed; drifted work doesn't get paid.
Five symptoms keep recurring. They look like five problems. They're not.
The fix is structural — turn every workflow into a stage-gated pipeline with explicit dispositions and named owners.
Each part is modular. Implement in sequence or start with the pillar where the team's bleeding most cycle time today (for most CHS Ops teams, that's Pillar 04 — Handoffs, or Pillar 07 — the Denial Disposition Workflow).
Six stages from Sold to Renewal — explicit entry/exit, SLAs, named owners, handoff artifacts at every transition.
Pipeline · Stage GatesQuestion, delete, simplify, accelerate, automate — applied monthly to one process. Required output: one deletion.
5 Steps · MonthlyT5T from credentialing leads, RCM leads, customer success — friction visible to the CEO in week 1, not Q4.
T5T · Flat · Group FeedbackFive critical handoffs in CHS ops — each with named owner on each side, an artifact (not a meeting), an SLA, a failure mode.
★ Star PillarDaily, weekly, bi-weekly, monthly forums. Each with a single decision mandate. No status updates — those live in the system.
Daily · Weekly · MonthlyTen operating rules for the COO and ops leadership. Time allocation. Stop-doing list. The standard for what good looks like.
10 RulesSix stages from sold to renewal. Each has explicit entry/exit, SLA, owner, and a handoff artifact at the boundary. Stage 4 (Stabilization) is the highest-risk window — that's where credentialing applications either start moving or get stuck.
Five steps. The order is non-negotiable. Optimizing before deleting is the most common mistake in healthcare ops — sophisticated, fast versions of work that shouldn't exist.
Attach a person's name to every requirement. "Compliance" is not a name. If no one owns it, it doesn't exist.
Most credentialing checklists can lose 30–50% of items and gain quality. Most denial workflows have steps that exist because they used to.
Only after deletion. Never simplify what should be eliminated.
Speed up what survives. Halve cycle time. Halve it again. Cycle time is the metric.
Last. Automating a broken process makes you faster at the wrong thing.
CAQH attestation runs annually. Most providers have at least one stale field. Stale CAQH gates new payer applications — the first we hear about it is when a payer rejects.
Quarterly attestation cycle with automated diff. CAQH stale rate becomes a tracked metric, not an annual scramble. Payer applications gate-check CAQH hygiene before submission, not after rejection.
Every denied claim gets a free-text note. Categorization is inconsistent across team members. Appeal-or-write-off decisions made ad hoc, often weeks after the denial.
9 standardized denial categories. Each has a named disposition rule (Appeal / Adjust / Write-off) and an owner. Every denial dispositioned within 5 business days. Weekly Denial War Room.
Re-credentialing cycle dates live in payer portals, scattered spreadsheets, and individual heads. Providers fall out of network silently when nobody catches the date.
Re-cred calendar lives in HubSpot per provider × payer. Alerts fire at 180/120/60 days before expiry. Service Manager owns the queue. Surprise lapses become near-zero.
Five handoffs in CHS operations quietly determine whether the company scales smoothly or grinds. Each one needs a named owner on each side, an artifact — a single document the next owner can read and run with — an SLA, a failure mode, and a detection signal. Click any handoff to expand.
One-page scope document in HubSpot. Auto-generated from the deal at signature, completed by AE within 24h. Includes signed scope, provider list with NPIs, payer list, current denial baseline (if RCM), current credentialing state (if cred), success metrics, customer stakeholders, day-0 expectations.
Failure: AE moves to next deal; verbal commitments surface mid-onboarding as "we thought you were doing X." Provider list is incomplete; payer list is wrong; baseline numbers don't match.
Detection: Any onboarding ticket in week 2+ that references a commitment not in the Charter. Should trend to zero.
Multi-section sign-off package. What is live (every contracted provider × payer combo with status), provider data confirmed in CAQH, system access granted (clearinghouse, payer portals, EHR if applicable), first 30-day SLA targets, customer escalation tree.
Failure: Onboarding never actually leaves. Onboarding Lead keeps fielding tickets that should be Service Manager's. Customer treats onboarding as the relationship; service team is invisible to them.
Detection: Onboarding Lead with billable hours on a customer 30+ days post-go-live. Service Manager not present in customer-facing communications. Both should trend to zero.
One-page (max two) report sent to customer the 5th business day after month-end. Numbers in the customer's language: denial rate vs. baseline, AR aging buckets, credentialing apps in queue by stage, re-cred upcoming. Plain-English narrative on what changed and why.
Failure: Reports drift to bi-monthly, then quarterly. Customer's confidence in "did we get what we paid for" decays. Renewal conversations start uphill because customer can't recall what we did for them.
Detection: Any month where the report misses the BD-5 SLA. Track the streak. A miss is a leadership conversation, not a process flag.
Reconcile cash collected (RCM service) and billable scope (per HubSpot) every Friday. Differences flagged within the same week, not at month-end. Odoo reads from HubSpot for billing; never the reverse.
Failure: Cash and scope diverge silently. Customer disputes the invoice at month-end based on cash they actually received vs. what we believe we earned. Renewal conversations get hostile.
Detection: Friday-to-Friday delta. Any >2% delta unresolved within 7 days is a Service Manager 1:1 topic the following Monday.
Published to the customer in the Production-Ready Gate Doc. Account Manager owns the relationship. Production-side issues route to Service Manager. Internally, every inbound is owned within 24h or escalated. Customer mental model: "one person to call."
Failure: Customer sends the question to the AE who closed the deal 8 months ago. AE forwards to Onboarding. Onboarding forwards to Production. Customer escalates to CEO. The whole org has now been pulled into one ticket.
Detection: Track inbound emails forwarded more than once. Spike usually means the convention isn't being communicated to customers — gate doc is stale.
RCM's single largest revenue leakage point isn't denials themselves — it's denials sitting in indeterminate state for 15–30 days while appeal windows close. Every denied claim gets one of three explicit dispositions within 5 business days. No exceptions. No "we'll look at it later." Disposition discipline is the most important single discipline in this entire system.
Pulled into queue from clearinghouse / payer 835. Auto-categorized to one of 9 standard categories.
AR clerk validates category, confirms claim/payer/provider data, identifies appeal window if applicable.
One of three: Appeal, Adjust, or Write-off. Named disposition rule applied per category. Exceptions named explicitly.
Appeal filed / adjustment posted / write-off booked. Each disposition has a runbook owned by Service Manager. Outcome tracked in HubSpot.
Outcome categorized; recovery rate by category tracked weekly in Denial War Room. Patterns fed back into Algorithm Audit candidates.
The denial is wrong, recoverable within window, and recovery value justifies the work.
The denial is correct or partially correct; we accept it and move on without writing off the underlying receivable.
Recovery is uneconomical or impossible. Write off promptly and free the AR aging line — don't let it linger.
Five recurring forums, each with a single decision mandate. Hard agendas. Pre-reads required. The forums are deliberately separate from sales cadence — different signal, different decisions.
15 minutes. AR clerks + credentialing leads. Each lead names blockers and aging items. No status updates — those live in the system.
Friday. Review the week's denial volume by category. Recovery rate trend. Disposition SLA compliance. Patterns flagged for Algorithm Audit candidates.
Mondays. Every payer application in flight, by stage and age. Aging apps get named action items. Re-cred queue reviewed for next 60 days.
The only meeting where Production, Onboarding, Account Management, and Accounting are in the same room. Agenda: handoff health, recurring friction, Algorithm Audit candidates.
One process per month. Apply the 5 steps. Required output: one thing deleted. Not "flagged." Deleted. If nothing is deleted, name the failure publicly.
Top 10 accounts. Each Service Manager owns the read. Outputs: at-risk accounts, expansion-ready accounts, monthly report sentiment, billing reconciliation gaps.
Ten rules for the COO and CHS Operations leadership. Non-negotiable behavioral standards designed to make the right action obvious in the moment when speed and judgment are both at stake.
Appeal, Adjust, or Write-off. No "we'll look at it later." Disposition discipline is the difference between a profitable RCM book and an unprofitable one.
Every transition has a named owner on each side, a document, an SLA, and a known failure mode. If a handoff requires a meeting to function, the handoff is broken.
"Compliance" is not a person. If no one owns it, it doesn't exist. Precondition for Step 1 of Musk's algorithm.
Every provider × payer with active expiry. Alerts fire at 180/120/60 days. Surprise lapses are a leadership failure, not an oversight.
Odoo reads from HubSpot. ServiceNow (if used for payer apps) writes back to HubSpot. Three systems, one truth. Reconciled weekly, not monthly.
Every customer, every month, no exceptions. Customer can't recall what we did for them = renewal conversation starts uphill.
Not throughput. Not utilization. Cycle time per stage, per disposition, per handoff. Halve it. Then halve it again. Speed compounds.
Once is bad luck. Twice is coincidence. Three times is a system. Add it to the next Algorithm Audit and resolve it.
Algorithm Audit produces one removed step, every month, without exception. If the session ends without a deletion, name the failure publicly and rerun within 7 days.
Single point of contact. Internal handoffs are invisible to them. If a customer ever has to ask "who do I talk to about X" — that's a system failure.
The CHS Operations OS is the sister system to the CHS Revenue OS — sales gets the yes; operations makes the yes worth something. Both operate on the same discipline: T5T as signal, narrative as method, deletion as default, cycle time as the metric.