Growth has flattened — PE platforms

Same-store flat while the unit count rises

PRESENTATION — WHAT THE OPERATOR SEES

Total revenue is up and the board deck looks fine, because six add-ons closed in eighteen months. Same-store is flat or negative and nobody can say which sites are carrying it. The sponsor sees a platform that is buying its growth rather than compounding it — and inorganic growth is priced at the add-on multiple, not the platform multiple. Nothing here is readable until the platform reports same-store separately from add-on growth, on a cohort definition it wrote down first.

DIFFERENTIAL — LIKELY CAUSES, MOST LIKELY FIRST

01
Growth is entirely inorganic and the reporting cannot separate it

Unit count, total revenue and adjusted EBITDA can all rise while the comparable base declines. They are independent series, and a platform that reports only the consolidated line has no way to see the divergence — or to prove to a buyer that it isn't there.

The Joint Corp, Q3 2025: comp sales (2.0)% and system-wide sales $127.3m down 1.5% — while revenue rose 6% to $13.4m and adjusted EBITDA rose 36% to $3.3m from $2.4m. Net unit change was negative in the same quarter: 9 franchised clinics opened against 11 closed, plus 3 company-owned or managed closed and 1 refranchised, on a base of 962.

The Joint Corp. Q3 2025 earnings release (SEC-filed), read at source

NOT THIS IF — Same-store is already reported separately from de novo and acquisition, on a written cohort definition, and is positive. Then the plateau is somewhere else in this list.
02
De novo openings are being counted inside "organic"

Organic growth includes de novo. Same-store deliberately excludes both de novo and acquisitions. A platform that opens units and reports "organic growth" has measured the openings, not the base — and a buyer's quality of earnings work will separate the two in week one.

Acadia Healthcare puts a number on what the ramp costs, and it is the only quantified de novo figure in this vertical that is filed under legal liability. FY2025 startup losses were $56 million, with FY2026 guided to $47–53 million and Q1 2026 to approximately $14 million. Acadia defines the line precisely: "Startup losses represent the anticipated net operating loss for new facilities opened over the previous twelve months and, to a lesser extent, preopening costs associated with facilities expected to open in future periods." In the same year Acadia added 1,089 licensed beds — 311 to existing facilities and 778 from newly opened ones. LifeStance shows the same structural point at the unit level: 572 centers and 8,040 clinicians, with a new center specified at 4,000–5,000 sq ft and 12–15 clinicians. A unit entering the base at that scale is a materially different thing from a mature one, and blending them hides both the drag and the ramp.

Acadia Healthcare Q4 and FY2025 earnings release, 25 February 2026 (SEC-filed), read at source; LifeStance Health FY2025 Form 10-K

NOT THIS IF — De novos are excluded from the comparable base until their thirteenth full month, the exclusion is documented, and the startup drag is reported as its own line. Note the reverse error also exists: excluding de novo from organic understates a platform that is genuinely building.
03
The acquired base was never brought to platform-level operating performance

Add-ons are underwritten on synergy and integrated on back office. The clinical and demand operations are left where they were, so the acquired sites keep producing at the practice-level median while the platform is priced as though they produce at the top.

Planet DDS, 2026 Outlook: revenue per chair at DSO-affiliated practices averages $205,690 with a median of $156,741, across 59,139 chairs — against $236,286 average and $184,502 median at solo practices. On the per-chair measure, the consolidated base underperforms the independent one.

Planet DDS, 2026 Outlook / 2026 Deep Dive

NOT THIS IF — Post-close production per chair, room or provider at acquired sites matches or exceeds their own pre-acquisition trailing twelve months. Then integration is not the constraint.
04
A conversion ceiling, not a demand ceiling

Demand arrives and does not become booked, shown, treated work. Spending more against the same conversion buys the same plateau at a higher cost, which is why this cause is so often misdiagnosed as an acquisition problem.

Henry Schein One, 2026 Catalyst Index: new patients per location per month average 39 against 82 at the top 10%. The Index's own finding is that top-performing groups do not necessarily spend more on marketing — they make it easier for demand to convert: 7-day versus 23-day time to appointment, and 45% versus 75% case acceptance. Bain's five value-creation levers for PE-owned physician groups point the same way — exactly one touches demand generation, and Bain frames it as enhancing "core business operations to better engage patients and referrers," an operations lever, not a media lever.

Henry Schein One, 2026 Catalyst Index; Bain & Company, Global Healthcare Private Equity Report 2026, Physician Groups chapter (read at source)

NOT THIS IF — Third-available appointment is inside a week and case acceptance is already in the top band. Then the constraint is genuinely capacity or demand, not conversion.
05
Labor, not demand — the delivery hours do not exist to sell

A vacant provider seat is a hard ceiling on same-store growth. No amount of demand generation clears it, and generating demand against it actively damages the brand by lengthening waits.

APTA reports a national vacancy rate of 9.5% for outpatient PT practices, "nearly double the national average across all industries." Provident Healthcare Partners ties it directly to deal activity: "These labor challenges are driving strategic M&A activity, as many platforms struggle to meet growth targets through organic growth alone." A named advisor stating that organic growth failure is what forces acquisition.

APTA survey, quoted verbatim in Provident Healthcare Partners, Healthcare Services M&A Review Q4 2024 (read at source)

NOT THIS IF — Provider schedules carry open, sellable capacity. Then vacancy is not the binding constraint and demand work is legitimate.

HOW TO TELL THEM APART

How to tell these apart in your own numbers

Each of these is a measurement you can run yourself, without us.

01 · Growth is entirely inorganic and the reporting cannot separate it

Rebuild same-store on a strict cohort: sites owned and open for the full trailing 24 months, de novos excluded until their thirteenth month, acquisitions excluded until twelve months post-close. Report it beside total revenue for the same periods.

CONFIRMS IF

Same-store is flat or negative while total revenue grows — the pattern The Joint Corp disclosed in Q3 2025, comp sales (2.0)% against revenue up 6%.

EXCLUDES IF

Same-store growth is positive and roughly tracks total growth. The plateau is then a specific-site problem, not a structural one.

02 · De novo openings are being counted inside "organic"

Report organic growth and same-store growth as two separate lines for the identical period, each with its written definition attached — then add a third line Acadia files publicly and most private platforms cannot produce: the net operating loss attributable to units opened in the trailing twelve months, plus preopening cost.

CONFIRMS IF

The two growth numbers differ materially and only the higher one has ever reached the board, or the startup-drag line cannot be produced at all. Acadia's is $56 million against $3.31 billion of total facility revenue — the drag is real enough to disclose and guide.

EXCLUDES IF

The definitions are already written down, the de novo cohort is broken out, and the startup drag is a standing line item. Do not, however, divide startup losses by units opened to manufacture a per-unit cost — Acadia's $56 million covers units opened at different points in the year plus preopening cost for units not yet open, so the quotient is not a unit economic.

03 · The acquired base was never brought to platform-level operating performance

Rank every site by revenue per chair, treatment room or provider — and compare each acquired site against its own pre-acquisition trailing twelve months, not against the platform average.

CONFIRMS IF

Acquired sites sit at or below their pre-close run rate, and cluster near the Planet DDS DSO median of $156,741 per chair rather than the $184,502 solo median.

EXCLUDES IF

Acquired sites improved post-close on the per-unit measure. The dilution is then mix — you bought smaller sites — not degradation.

04 · A conversion ceiling, not a demand ceiling

Two numbers per site, measured the same week: third-available appointment in days, and case acceptance (or consult-to-treatment) by provider rather than by site.

CONFIRMS IF

Third-available runs toward Henry Schein One's 23-day typical figure rather than its 7-day top-performer figure, or acceptance sits near 45% rather than 75%.

EXCLUDES IF

Access is inside a week and acceptance is in the top band at every site. Conversion is not your constraint.

05 · Labor, not demand — the delivery hours do not exist to sell

Percentage of budgeted provider hours unfilled, by site, for the trailing quarter — vacancy plus unfilled shifts, not headcount against plan.

CONFIRMS IF

Unfilled provider hours are the same order as the growth shortfall, or vacancy approaches the 9.5% APTA reports for outpatient PT.

EXCLUDES IF

Provider hours are substantially filled and sellable capacity is going unused.

WHAT RESOLVES EACH

What resolves this, and how you will know it resolved

Growth is entirely inorganic and the reporting cannot separate it	Rx 04 · marketing attribution →	A readout that separates same-store, de novo and acquisition on a fixed cohort, monthly. This is a measurement fix before it is a growth fix — and it is the number a buyer opens first.
De novo openings are being counted inside "organic"	Rx 04 · marketing attribution →	Write the cohort definition down and hold it constant. A definition that moves between quarters is worse than no definition, because diligence will find the change.
The acquired base was never brought to platform-level operating performance	Rx 03 · patient conversion →	The demand-side operations that were never integrated — scheduling, routing, capacity matching, front-desk instrumentation. Adding media across an unintegrated base scales the variance, not the mean.
A conversion ceiling, not a demand ceiling	Rx 03 · patient conversion →	Bain classifies this as operations and so do we. What does not work: raising spend. Against a 45% acceptance rate, incremental spend buys incremental unconverted demand at full price.
Labor, not demand — the delivery hours do not exist to sell	not a marketing engagement →	We would not take this on as a demand problem. A 9.5%-order vacancy rate is a recruiting and compensation problem, and demand generation against it makes the patient experience worse. Fix the seat, then call.

WHAT "RESOLVED" LOOKS LIKE — Same-store growth, reported separately from de novo and acquisition on a fixed cohort

MEDIAN

No published benchmark exists for private healthcare platforms — this is a gap in the market, not in our research. Across a 48-source harvest, no advisory firm or transaction database publishes same-store growth for private platforms. The only honest anchors are public filers, and they bracket wide: The Joint Corp comp sales (2.0)% in Q3 2025; Acadia Healthcare FY2025 same-facility revenue +4.9% on patient days +2.1%, admissions +2.3% and revenue per patient day +2.8%. Read Acadia's full line before treating +4.9% as the goal — in the same table, same-facility adjusted EBITDA fell 3.6%, from $855.2m to $824.3m, while same-facility revenue rose. Same-store revenue growth and same-store margin can move in opposite directions, and this one is flattered downward by a $52.7m professional and general liability reserve adjustment Acadia took in Q4.

TOP DECILE

Not published for this vertical by any source. Treat any figure claiming a private-platform same-store decile as manufactured until you can open its primary.

TARGET

Positive same-store growth on a written, unchanged cohort definition, reported monthly, with de novo and acquisition contribution shown beside it rather than inside it. The target is a distribution position and a measurement standard, not a promise — and the ceiling that matters is set by the operating metric underneath the platform, not by the platform line.

The Joint Corp. Q3 2025 earnings release; Acadia Healthcare FY2025 Form 10-K; ADMEN PE/rollup source harvest, 48 sources

HOW THIS DIFFERS BY SCALE

How this differs by scale

Single site	A single site or an add-on has no same-store problem — it has a production problem, and it is legible without any of this machinery. Read the differential for the specialty itself: dental at /marketing-problems/dental-groups-dsos/growth-plateau/, dermatology at /marketing-problems/dermatology-groups/growth-plateau/.
Group	Three to ten sites is where the error enters, because total growth still looks like performance and one strong site can carry the average. This is the cheapest moment to install the cohort definition — before there is an acquisition history to restate.
Platform	At platform scale the consolidated line is nearly uninformative and the site-level distribution is the whole story. Expect a buyer to rebuild same-store themselves from your site-level data; the only question is whether your number and theirs agree. See also /marketing-problems/private-equity/multi-location-marketing/.

OTHER PRESENTATIONS — PE PLATFORMS & ROLLUPS

A differential narrows the field. It does not replace the examination — that is what the six weeks are for. Every figure above is an industry reference range, not a client's numbers; those stay sealed. Sources are set out at /sources.

APPLY — 6 / QUARTER →