Clinician compensation is the largest cost in a behavioral health group and the hardest thing to change once set, which is why most groups never change it; they accumulate. A split negotiated with the first hire, a salary offered to recruit a specialist, a per-session rate for a part-timer, a side arrangement for the clinician who threatened to leave — five years later the group has six arrangements, three of them undocumented, and a compensation structure that cannot be explained to a new hire, modeled for its effect on margin, or defended to a buyer. The redesign, in the order it has to happen.
Step one: the inventory
Every arrangement, written down as it actually operates. For each clinician: the model (percentage split, salary, salary plus productivity, per-session, hybrid), the parameters (percentage, base, threshold, rate), what the percentage is applied to (billed charges, contracted rate, or collected cash — three different numbers), benefits and their cost, paid time off, supervision given or received, any guarantees, bonuses, or side letters, and the date it was agreed and by whom. The inventory is reconciled to payroll for the trailing twelve months — because in most groups the payroll register and the owner's memory disagree on at least two clinicians.
The three findings the inventory usually produces. Splits applied to billed charges, which pays clinicians for revenue the group never collects; productivity thresholds set above break-even, which means the bonus is unreachable and the clinician knows it; and arrangements that differ for clinicians doing identical work, which is a retention problem waiting for a conversation in the break room.
Step two: the economics under each model
Compute what each clinician costs per completed session, today. Compensation as actually paid, benefits, employer taxes, divided by completed sessions; beside it the group's net revenue per session for that clinician's payer mix and the allocated overhead. The result is contribution per session by clinician under the current arrangement, and the spread — in most groups a few clinicians carry the margin and a few lose money on every session, with the model, not the clinician, as the usual cause.
Then model the same clinicians under each candidate model. A percentage split on collected revenue at 50, 55, and 60 percent; a salary at market with a productivity component above a threshold; a per-session rate by license type; and a hybrid with a base covering a minimum expectation and productivity above it — each at the clinician's actual utilization and payer mix, and at utilization five points higher and lower. The output is a grid: clinician by model by utilization, showing the clinician's pay and the group's contribution. The group is choosing a model for its economics and its ability to fill calendars, not for its familiarity.
Step three: set the threshold at break-even
The productivity threshold is the design decision that matters most. It is the number of completed sessions per week at which the clinician's fully loaded cost — base, benefits, taxes, allocated overhead — is covered by net revenue; in a group with a US$82,000 base, US$18,000 of benefits and taxes, US$25,000 of allocated overhead, and US$112 net per session, that is about 1,116 sessions a year, or 24 a week on 46 productive weeks. A threshold set at 24 is credible; the clinician who reaches 26 earns the productivity component on two sessions a week and feels it. A threshold set at 30 tells the clinician the bonus is theoretical, and the group gets the behavior of a flat salary with the cost of a bonus plan nobody earns.
The productivity rate above the threshold is the second decision. A share of net revenue per session above the threshold — 35 to 45 percent is common — that leaves the group meaningful contribution on marginal sessions while paying the clinician enough to want them; the rate is applied to collected revenue, never to billed charges.
Step four: the associate track
A pre-licensed clinician is on a different model with a defined step. A lower base reflecting narrower reimbursement and supervision cost, no productivity component or a lower threshold during the ramp, supervision hours funded by the group with a repayment term if the clinician leaves within a defined window after licensure, and a step at licensure that is large enough — 20 to 30 percent — that converting the caseload into a private practice is obviously worse than staying; the step is the retention mechanism, and it is written into the offer, not negotiated at licensure.
Step five: differentials and retention terms
Differentials for what the group needs and the market pays for. Specialty (child and adolescent, eating disorders, trauma), language, acuity-weighted caseload, evening and weekend blocks, and site lead or supervisory responsibility — each a defined amount or percentage, applied by rule rather than by negotiation.
Retention terms that make the model expensive to leave. A bonus vesting at 12 and 24 months, the funded-supervision repayment term for associates, licensing and continuing-education costs covered, and a documented path to clinical leadership with its own compensation — all of which cost less than a departure and all of which survive review better than a non-compete.
Step six: model every clinician's pay, before and after
No clinician should learn their new number in the meeting. The grid from step two is run on the final model: each clinician's current annual pay at actual utilization, their pay under the new model at the same utilization, and their pay at the utilization the group expects after the schedule and intake work; the group's contribution under both. The rule most groups adopt is that no clinician's pay falls at current utilization — the redesign is funded by the clinicians who were below break-even reaching it, by the splits moving from billed to collected, and by the thresholds that now get reached — and the handful of clinicians whose pay would fall get a transition guarantee for two to four quarters.
The group's number. Total compensation cost before and after at current utilization, and at the expected utilization; in most redesigns the group's cost is flat to slightly higher in year one and the contribution improves because utilization rises when the threshold is reachable.
Step seven: the transition conversation
One model, one document, one conversation per clinician, in a window of thirty days. The clinical director and the practice owner, with the CFO's one-page summary for the clinician: the model, the parameters, their modeled pay at their utilization, the differentials that apply, the retention terms, the effective date, and the transition guarantee if any. The conversation is a presentation of a group-wide change, not a negotiation; side arrangements are what the redesign is removing. Clinicians who are better off say so quickly; clinicians who are protected by the guarantee need to hear that it is a guarantee; the one or two who are worse off at current utilization and better off at target get the schedule and intake plan that gets them there.
Effective date at a payroll boundary, with a quarter of overlap reporting. The new model starts on the first day of a pay period; for the first quarter, each clinician's statement shows pay under both models so the transition is visible and disputes are about facts.
Step eight: the documentation that survives diligence
A compensation policy, a signed agreement per clinician, and a payroll that reconciles. The policy describes every model in force, the parameters, the differentials, the retention terms, and the review cadence; each clinician signs an agreement that references it; payroll runs from the practice management system's completed sessions for productivity calculations, with a monthly reconciliation that someone other than the person running payroll reviews; and the whole package is what a buyer's diligence team asks for first. Undocumented arrangements are the single most common reason behavioral health groups are re-underwritten downward in a sale.
Step nine: the twelve-month review
The model is reviewed against the market and the economics once a year, on a date. Posted market rates by license type checked; the group's net revenue per session and overhead refreshed, which moves the break-even threshold; utilization by clinician against the band; turnover and the exit themes; and adjustments applied across the model rather than negotiated clinician by clinician. The annual review is what prevents the next five years of accumulation.
Why this is CFO work
The inventory reconciles to payroll, which only finance can do; the economics by model require net revenue per session by payer and allocated overhead, which only finance computes; the threshold is a break-even calculation; the before-and-after grid is a model; the transition guarantee is a cost decision; and the documentation is a diligence deliverable. The clinical director owns the conversations and the clinical judgment on differentials; the CFO owns the numbers and the design.
Worked example
An 18-clinician outpatient group, US$3.6 million net revenue, six arrangements in force (three undocumented), net revenue per session US$112, allocated overhead US$25,000 per full-time clinician.
| Finding at inventory | Figure |
|---|---|
| Clinicians on a 60% split of billed charges | 5 — paid on US$0.14 per US$1 the group never collected; effective split of collected 68% |
| Salaried clinicians with productivity threshold at 30 sessions | 7 — none reached it in 12 months; average utilization 69% |
| Per-session part-timers at a rate above the group's net per session for their payer mix | 2 |
| Associates with no defined step at licensure | 4 — two licensed in the last year and left within 90 days |
| Contribution per session, best vs. worst clinician | +US$31 vs. −US$22 |
The redesign: one model for licensed full-time clinicians — US$82,000 base, productivity at 40 percent of collected net above 24 sessions a week; splits converted to 55 percent of collected for the two clinicians who preferred a split, with a floor; per-session part-timers at a rate by license type; an associate track at US$54,000 with funded supervision, a two-year stay term, and a 25 percent step at licensure; differentials for child and adolescent (US$4,000), Spanish-language (US$3,000), and site lead (US$12,000); retention bonus of US$2,500 at 12 months and US$4,000 at 24.
| Before-and-after (at current utilization) | Before | After |
|---|---|---|
| Group compensation cost | US$1,958,000 | US$1,989,000 (+1.6%) |
| Clinicians whose pay rose | — | 11 |
| Clinicians whose pay was flat | — | 4 |
| Clinicians with a transition guarantee (pay would have fallen) | — | 3 (two quarters) |
| Modeled contribution at expected utilization (+6 points, threshold reachable) | US$1,044,000 | US$1,206,000 |
| Departures in the transition quarter | — | 0 |
| Departures in the following 12 months | 5 prior year | 2 |
Illustrative figures for a hypothetical organization; not a client's data.
The split-on-billed clinicians had been paid about US$38,000 a year in aggregate on revenue the group never collected; the seven salaried clinicians had a bonus plan none of them could reach, and four of them reached the new threshold within two quarters once the schedule templates were rebuilt. The three transition guarantees cost US$14,000 over two quarters and retained three clinicians whose departures would have cost over US$300,000. The signed agreements and the reconciled payroll were the first items in the data room eighteen months later.