Growth has to read as a system — PE platforms

The growth story has to survive diligence, not just the pitch

PRESENTATION — WHAT THE OPERATOR SEES

Twelve to eighteen months from a process. The bankers want a growth narrative and diligence will want the mechanism underneath it. Right now the honest version of the story is that the platform bought well — and a buyer pays for a buying record differently than for a compounding one.

DIFFERENTIAL — LIKELY CAUSES, MOST LIKELY FIRST

01
The multiple arbitrage is narrower than the market still believes, so the premium has to be earned on the platform side

Buying at the practice tier and selling at the platform tier is the engine, and the spread has compressed. If the platform premium is not evidenced by something other than scale, the exit prices closer to the sum of the parts.

Brian Colao, Director of Dykema's DSO Industry Group — the only named source on record quantifying both ends. Smaller practices, which briefly reached multiples around 7× at the peak, have returned to traditional ranges of 5–6×. Larger DSO transactions once reached 13–16× EBITDA but have fallen closer to 9–10×. That is a spread of roughly 3–5 turns, and it is materially narrower than the peak-era arbitrage the market still discusses.

Brian Colao, Dykema DSO Industry Group, quoted verbatim in Group Dentistry Now

NOT THIS IF — Your add-ons clear $250m in enterprise value. Then you are in the large-cap market, a different set of comparables applies, and this differential is not describing you.
02
Same-store cannot be produced on demand, and it is the first number diligence builds

Private platforms have no external same-store benchmark, which means the number is judged on how it was constructed. A definition assembled during a process, or one that moved between quarters, arrives in the quality of earnings as a finding rather than as a metric.

No advisory source or transaction database publishes same-store growth for private healthcare platforms; only public filers disclose it, under legal liability — which is precisely why a buyer will judge yours on construction rather than on level. The two public disclosures available are instructive in opposite directions. The Joint Corp reported comp sales of (2.0)% in Q3 2025 while revenue rose 6% and adjusted EBITDA rose 36%. Acadia Healthcare's FY2025 same-facility table shows revenue +4.9% on patient days +2.1% and admissions +2.3% — and, in the same table, same-facility adjusted EBITDA down 3.6%, from $855.2m to $824.3m. A same-store number that reports only revenue is not the number, and a buyer will ask for the margin line beside it.

The Joint Corp. Q3 2025 earnings release; Acadia Healthcare Q4 and FY2025 earnings release, 25 February 2026 (SEC-filed), read at source; ADMEN PE/rollup source harvest, 48 sources

NOT THIS IF — Same-store already runs monthly on a fixed cohort with a written, unchanged definition and 24 months of history. Then this is a strength to lead with, not a gap.
03
There is no attributable demand mechanism — spend cannot be traced to a cohort with a payback

A growth story that cannot be decomposed into a repeatable acquisition mechanism is a story about the past. Buyers pay for growth they can re-run, and they discount unevidenced growth claims toward zero because they have no benchmark to check them against.

Bain's five value-creation levers for PE-owned physician groups are on record verbatim, and exactly one touches demand generation: "Enhance core business operations to better engage patients and referrers, while leveraging technology to drive productivity gains." Bain classifies it as an operations lever, not a media lever. Compounding the problem: no source, free or paid, publishes CAC, LTV or payback benchmarks for PE-backed healthcare platforms — so an unevidenced claim has nothing to be checked against and nothing to be credited against either.

Bain & Company, Global Healthcare Private Equity Report 2026, Physician Groups chapter (read at source); ADMEN PE/rollup source harvest, 48 sources

NOT THIS IF — You can produce a cohort-level payback for at least one channel at one site, traced from spend to first treatment to twelve-month revenue. That is the artifact; one real instance beats a platform-wide assertion.
04
The multiples inside your own materials may be contaminated

Private multiples are valuable and unpublished, so an SEO industry manufactures them and attributes them to real firms that never said them. A fabricated comp in a CIM is a credibility event, and it is the diligence finding you least want.

One confirmed fabrication: ctacquisitions.com publishes DSO and dermatology platform ranges — including "12.0x to 18.0x adjusted EBITDA" — and attributes them to a specific Provident Healthcare Partners dental M&A update from Q4 2024. That 42-page PDF was downloaded and text-extracted: the phrase "adjusted EBITDA" appears zero times in all 42 pages, and no multiple range of any kind appears in its dental section. The same publisher cites 12.0x–18.0x on one page and 10.0x–15.0x on another for overlapping periods. Separately, FOCUS Investment Banking states plainly that its subsector EV/EBITDA ranges "represent synthesized estimates based on reported transaction data, sector commentary, and trend signals" — an honest disclosure, and a reason not to publish the numbers.

ADMEN verification of ctacquisitions.com against the Provident Healthcare Partners Q4 2024 PDF (downloaded and text-extracted); FOCUS Investment Banking healthcare EBITDA multiples dashboard (self-declared methodology)

NOT THIS IF — Every multiple in the materials links to a primary document you can open. If so, you are already ahead of most processes.
05
You are benchmarking against a source that structurally cannot see you

The reference document for healthcare private equity excludes the deal sizes a mid-market platform actually executes. Quoting its deal counts as though they describe the rollup middle market is a misreading a sophisticated buyer will catch.

Bain's own figure note discloses the counting rule: "deal count excludes add-on deals below $250 million; 2025E represents actual data through November 30, 2025, annualized." Essentially every tuck-in a mid-market platform executes sits below that floor and is therefore invisible in Bain's counts. Bain is entirely citable for direction — global healthcare PE deal value above $190 billion in 2025 on 445 announced buyouts, provider and related services up 57% to roughly $62 billion, exit value from $54 billion in 2024 to an expected $156 billion, exits above $1 billion from 16 to more than 40, and more than 150 sponsor-to-sponsor deals worth more than $120 billion. It is not a middle-market denominator.

Bain & Company, Global Healthcare Private Equity Report 2026 (figure note and headline figures, read at source; Bain's stated data sources are Dealogic, AVCJ, PitchBook and Bain analysis)

NOT THIS IF — Your comparable set genuinely sits above $250m in enterprise value. Then Bain is measuring your market and you should use it.

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 · The multiple arbitrage is narrower than the market still believes, so the premium has to be earned on the platform side

Model the platform twice: once at the platform tier and once at the add-on tier applied to the acquired EBITDA. Colao's published ranges give you the goalposts — 5–6× at the practice tier against 9–10× at the platform tier.

CONFIRMS IF

The gap between the two models is the entire equity story. Whatever justifies that 3–5 turns is what you must evidence, and if you cannot name it in a sentence, neither can the banker.

EXCLUDES IF

The platform premium is already anchored to something specific and testable — payer contracts, ancillary capture, a de novo record with disclosed economics.

02 · Same-store cannot be produced on demand, and it is the first number diligence builds

Ask your finance team to produce 24 months of same-store growth on a fixed cohort with a written definition — revenue AND margin, both lines, the way Acadia files them — and time how long it takes.

CONFIRMS IF

It cannot be produced inside a week, the cohort definition has to be invented to produce it, or only the revenue line exists. Acadia's public filing shows why the margin line matters: same-facility revenue +4.9% alongside same-facility adjusted EBITDA −3.6% in the same year.

EXCLUDES IF

It arrives in a day from an existing monthly readout with an unchanged definition.

03 · There is no attributable demand mechanism — spend cannot be traced to a cohort with a payback

Take one acquisition channel at one site, isolate a single month's patient cohort, and follow it to first treatment, second visit and twelve-month revenue. Compute the payback period on that one cohort.

CONFIRMS IF

No cohort at any site can be followed end to end. If it cannot be done once, it is not a mechanism.

EXCLUDES IF

The cohort resolves and the payback is computable. Then build the same view across channels and sites — that is the growth story.

04 · The multiples inside your own materials may be contaminated

Citation audit of the CIM and management presentation. For every multiple, valuation range and market statistic, open the primary document and search it for the number. Any figure attributed to a named firm gets checked against that firm's actual publication, not against the page that quoted it.

CONFIRMS IF

Any figure cannot be found in the document it is attributed to. That is a fabrication regardless of who repeated it — as with the 12.0x–18.0x range attributed to a Provident PDF containing no such range.

EXCLUDES IF

Every figure resolves to an openable primary.

05 · You are benchmarking against a source that structurally cannot see you

List your last ten add-ons by enterprise value and count how many exceed $250 million.

CONFIRMS IF

Few or none clear the threshold. Bain's deal counts do not describe your market and should not appear as your denominator.

EXCLUDES IF

Most clear it. Then Bain is measuring you and is the right reference.

WHAT RESOLVES EACH

What resolves this, and how you will know it resolved

The multiple arbitrage is narrower than the market still believes, so the premium has to be earned on the platform side	Rx 05 · preparing for sale →	Multiple-modeling assist and a growth narrative built to survive being checked. The frank version: if the only thing separating your platform tier from your add-on tier is scale, the honest advice is to spend the next four quarters building the mechanism rather than the deck.
Same-store cannot be produced on demand, and it is the first number diligence builds	Rx 04 · marketing attribution →	Install the fixed-cohort readout now, not during the process. Twenty-four months of history under one unchanged definition cannot be manufactured later, which is exactly why it carries weight. See /marketing-problems/private-equity/growth-plateau/.
There is no attributable demand mechanism — spend cannot be traced to a cohort with a payback	Rx 01 · patient acquisition →	Paired with Rx 04 · reporting, because the mechanism and the evidence of it are the same build. Bain frames the demand lever as operations — so the artifact a buyer credits is an operating system with attribution, not a media plan with a spend curve.
The multiples inside your own materials may be contaminated	Rx 05 · preparing for sale →	The cheapest credibility win available in a CIM is deleting every number whose primary you cannot open. It costs nothing, takes an afternoon, and removes the finding a sharp buyer uses to reprice everything else on the page.
You are benchmarking against a source that structurally cannot see you	Rx 05 · preparing for sale →	Cite Bain for direction and never for your denominator. Provident Healthcare Partners and Mertz Taggart count sub-$250m transactions and publish quarterly for free — Provident tracked 216 healthcare services transactions in Q1 2025, 174 strategic and 42 financial, with Multisite Provider Services the largest sector at 64.

WHAT "RESOLVED" LOOKS LIKE — Exit multiple at the platform tier, and the evidence supporting the spread over the add-on tier

MEDIAN

9–10× EBITDA at the large-DSO platform tier today, against 5–6× at the practice or add-on tier — roughly 3–5 turns of spread (Colao / Dykema). For ambulatory surgery, VMG Health reports many centers trading around the historical 7–8× benchmark.

TOP DECILE

Not published for this vertical, and no source produces one. The closest sourced statement of a top tier is VMG Health's, that "select best-in-class ASCs, especially those with strong strategic demand, have achieved double-digit valuation multiples" against that 7–8× benchmark — roughly 2–3 turns above. Treat any claimed top-decile platform multiple as unsourced until you open its primary.

TARGET

Hold the platform tier and be able to evidence the spread over the add-on tier with same-store history, a cohort payback and a documented site-level distribution. Do not model 13–16× — Colao states plainly that range belonged to the peak and has fallen. This is a distribution position in thin, named, checkable data, not a promise, and the market context is real: Bain reports exit value rising from $54 billion in 2024 to an expected $156 billion in 2025, while Dykema records more than 50 significant DSO sales processes abandoned since mid-2022.

Brian Colao / Dykema DSO Industry Group via Group Dentistry Now; VMG Health (read at source); Bain Global Healthcare PE Report 2026; Dykema M&A Sector Spotlight: Dental Service Organizations

HOW THIS DIFFERS BY SCALE

How this differs by scale

Single site	A single practice selling into a platform is on the other side of this trade and is being priced at the add-on tier — 5–6× for dental practices selling into a platform on Colao's published range. The relevant preparation is practice-level, not platform-level: see /marketing-problems/dental-groups-dsos/preparing-for-sale/.
Group	A sub-scale group is the hardest position, because it carries platform overhead while pricing near the add-on tier. The decision is usually whether to buy enough to reach the platform tier or to sell into someone who already has — and that decision should be made against a same-store number, not against a pipeline.
Platform	At platform scale the exit multiple is set by whether growth reads as a mechanism. Bain's five levers are the published framework a sophisticated buyer already uses, and only one touches demand. Being able to show the demand lever operating — cohort payback, same-store, site-level distribution — is what turns a pitch into a growth story that survives diligence, and it is precisely the part most platforms cannot produce.

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 →