The first location of a behavioral health group succeeds on the founder's reputation, referral relationships, and willingness to work evenings. The second location tests whether the model works without those things, and the test is failed more often by the finance than by the clinical work: a lease signed on demand, clinicians hired before credentialing, a ramp assumed in months when it takes quarters, and no date on which anyone decides whether it is working. The model that prevents that, built before the lease is signed.
What to know before choosing the site
Demand is necessary and not sufficient. The group's waitlist, inquiry volume by geography, and referral sources that specify a location are the demand evidence; a waitlist of forty in a neighborhood thirty minutes from the current site is a real signal, and an owner's sense that "there is nothing out there" is not. The model starts from the inquiries the group already turns away by location and the share that would convert at a closer site.
Payer geography decides the margin before anyone is hired. The payers that dominate the target area's population may differ from the first site's mix — a suburb with employer-sponsored commercial coverage and a corridor with heavy managed-Medicaid enrollment produce different net revenue per session at the same rates; the group's existing contracts may or may not extend to the new address, and some payers require a separate location credentialing that takes 60 to 120 days; the model carries the target area's expected payer mix, not the first site's.
Clinician supply sets the ramp. A site can be staffed by relocating existing clinicians — which cannibalizes the first site's capacity and the caseload that does not follow — or by net-new hiring, which runs 60 to 120 days to fill plus 60 to 120 days to credential plus three to six months to ramp; the local labor market for the license types and specialties the site needs is checked by posting a role before the lease is signed, and the model's staffing dates come from that evidence.
Competition and referral sources are mapped, not assumed. The existing groups within a reasonable radius, their payer participation, their specialties, and their waitlists; the primary care practices, schools, employee assistance programs, and regional centers that would refer; and the one or two referral relationships the group can realistically establish before opening.
The lease math
The lease is a monthly revenue requirement, not a monthly cost. A 2,400-square-foot suite with six offices at US$32 per square foot gross is US$6,400 a month, plus utilities, furniture, build-out amortized over the term, signage, technology, and insurance — call it US$8,900 a month of occupancy; at a blended net revenue of US$114 per session and an occupancy target of 8 to 10 percent of revenue, the site must deliver about 800 to 950 sessions a month to carry the space, which is five to six full-time clinicians at target utilization. The model asks whether the staffing plan reaches that volume, by which month, and what the occupancy ratio is in every month before it does.
Term, options, and the tenant-improvement allowance are finance terms. A five-year term with a two-year early-termination right, a tenant-improvement allowance that funds the build-out rather than the group's cash, free rent for the first three to six months while the site ramps, and an expansion option on adjacent space are each worth modeling in dollars; a landlord who will not give the termination right is pricing the group's risk into the group's pocket.
Shared space and sublease alternatives. Before a lease, the model compares a sublease of two or three offices from an existing practice, a shared-suite arrangement, or a telehealth-first launch with a smaller footprint — each with lower fixed cost and a slower ceiling — against the full suite; the right answer depends on the demand evidence and the group's cash.
The staffing and credentialing sequence
Credentialing starts before the lease is signed. Location credentialing with each payer is submitted the day the address is known; clinician credentialing is submitted the day an offer is accepted; the site opens on the date the largest payers are live for at least two clinicians, not on the date the furniture arrives.
The site lead is the first hire, and the model budgets their management time. A clinician at 70 percent clinical and 30 percent site management, compensated for both, who owns the schedule template, intake routing, referral relationships, and the ramp — a site without a lead is a site the founder is running from thirty minutes away.
The staffing plan is phased to demand and credentialing. Two clinicians at opening (one relocated with a caseload that follows, one new hire credentialed before opening), a third at month three, a fourth and fifth at months six and nine as utilization at the site passes 75 percent and the intake funnel confirms demand — each hire gated on a utilization and funnel trigger rather than on a calendar.
Front desk, billing, and intake are costed from month one. A part-time front desk and intake coordinator at opening, scaling with inquiries; billing absorbed by the group's existing function with a per-claim allocation; the model carries these as site costs, not as free capacity from site one.
The ramp and the cash curve
Utilization ramps by clinician, not by site. A relocated clinician with a caseload that follows opens near 60 percent and reaches 80 in two to three months; a new hire opens at 30 to 40 percent and reaches 75 in four to six months; the site's blended utilization is the weighted average, and the model carries it month by month.
Revenue follows utilization with the payer lag. Sessions in month one are collected in months two through four by payer; self-pay collects immediately; the site's cash receipts lag its revenue by the group's weighted days to cash, which is why the cumulative cash requirement is larger than the cumulative operating loss.
The cumulative cash requirement is the number the owners approve. Build-out and furniture net of the allowance, deposits, pre-opening payroll for the site lead and first hires during credentialing, the operating loss during the ramp, and the receivables lag — summed month by month to the trough, which usually arrives in month five to eight; a site that reaches monthly break-even in month nine may need US$250,000 to US$400,000 of cumulative cash before it does, and the owners should see that number before the lease, not discover it in month six.
Break-even is a month, and the model names it. Monthly contribution at the site — net revenue less clinician compensation, site staff, occupancy, allocated billing and management — crosses zero in a named month under the base case; cumulative break-even, where the site has repaid its investment, comes twelve to eighteen months later.
The stress cases
Run the model at the ramp that actually happens. Credentialing 30 days slower than planned; the first new hire resigning at month four; utilization reaching 70 rather than 80 percent; the target area's payer mix ten points heavier in managed Medicaid; no-shows at 15 rather than 11 percent — each as a toggle, and the question is whether the site survives a ramp that takes 50 percent longer than planned; if it survives only the base case, it is not a plan, it is a hope.
Cannibalization is a stress case, too. The share of the new site's early clients who would have been served at the first site anyway, and the first site's utilization after the relocated clinician leaves — the group's net gain is the new site's revenue less the first site's loss, and in a group where the first site has slack, the second site's first six months can be close to net-neutral.
The decision gate
Day 120 is a decision, not a review. Specific targets set before opening — site utilization, completed sessions, intake conversion, referral-source activity, cumulative cash against the model — and a written set of outcomes: on track, continue; behind on volume with a credible cause, extend 90 days with a revised plan; behind on volume with no credible cause, stop hiring and consolidate; and the owners decide on that date with the numbers in front of them. The most common failure in a two-site group is a second site that never got its own numbers and was carried for three years by the first.
The site's profit and loss statement exists from month one. Its own revenue by payer, its own clinician and staff cost, its own occupancy, and its allocated share of billing and management — reported in the management pack beside site one, with the ramp model as the comparison column.
Why this is CFO work
The demand evidence comes from the intake data; the payer geography from the payer file and the target area's enrollment; the lease math from the group's net revenue per session and occupancy ratio; the staffing sequence from the credentialing clock and the recruiting function; the ramp and cash curve from utilization and collection timing the group already measures; and the decision gate from someone who will put a slow site on the agenda on the day it was agreed. None of it happens in a group where the owner is seeing clients and the lease is on the desk.
Worked example
A 16-clinician group at one site, US$3.4 million net revenue, opens a second site twenty-five minutes away: 2,200 square feet, five offices, US$8,400 monthly occupancy all-in, three months' free rent, a US$40,000 tenant-improvement allowance against a US$65,000 build-out. Staffing: a site lead relocated with 60 percent of her caseload following, one new hire credentialed before opening, a third clinician at month three, fourth at month six, fifth at month nine. Blended net revenue US$112 per session (target area is six points heavier in managed Medicaid than site one). Payer lag: weighted 34 days.
| Month | Clinicians | Blended utilization | Sessions | Net revenue | Cash receipts | Site cost | Contribution | Cumulative cash |
|---|---|---|---|---|---|---|---|---|
| −2 to 0 (pre-open) | lead + 1 | — | 0 | 0 | 0 | 61,000 (build-out net, deposits, pre-open payroll, credentialing) | — | −61,000 |
| 1 | 2 | 44% | 148 | 16,600 | 5,200 | 31,400 | −14,800 | −87,200 |
| 2 | 2 | 56% | 189 | 21,200 | 14,900 | 31,400 | −10,200 | −103,700 |
| 3 | 3 | 55% | 279 | 31,200 | 20,800 | 40,800 | −9,600 | −123,700 |
| 4 | 3 | 63% | 319 | 35,700 | 29,600 | 49,200 (free rent ends) | −13,500 | −143,300 |
| 5 | 3 | 70% | 355 | 39,800 | 34,900 | 49,200 | −9,400 | −157,600 |
| 6 | 4 | 66% | 446 | 50,000 | 38,600 | 58,600 | −8,600 | −177,600 |
| 7 | 4 | 72% | 487 | 54,500 | 47,100 | 58,600 | −4,100 | −189,100 |
| 8 | 4 | 77% | 521 | 58,400 | 53,000 | 58,600 | −200 | −194,700 |
| 9 | 5 | 73% | 617 | 69,100 | 57,300 | 68,000 | +1,100 | −205,400 (trough) |
| 10 | 5 | 77% | 651 | 72,900 | 66,200 | 68,000 | +4,900 | −207,200 |
| 11 | 5 | 80% | 676 | 75,700 | 71,900 | 68,000 | +7,700 | −203,300 |
| 12 | 5 | 82% | 693 | 77,600 | 75,400 | 68,000 | +9,600 | −195,900 |
| 13–18 | 5 | 83% avg | ~700 / mo | ~78,400 / mo | ~77,500 / mo | 68,000 | ~+10,400 / mo | −133,000 at month 18 |
Illustrative figures for a hypothetical organization; not a client's data.
Monthly contribution turns positive in month nine; the cash trough is about US$207,000 in month ten; cumulative break-even arrives around month 31 under the base case. The stress case — the month-three hire resigning at month five and credentialing on the largest payer running 40 days late — pushes the trough to US$268,000 and monthly break-even to month thirteen; the site survives it. Day-120 targets were set at 60 percent blended utilization, 310 sessions, and cumulative cash no worse than US$150,000; the site reported 63 percent, 319 sessions, and US$143,300, and continued. The owners had approved a US$275,000 cash requirement and a line draw plan before the lease was signed, which is why month ten was a line item rather than a crisis.