DX / Differentials / PE PLATFORMS & ROLLUPS

D-04 — MULTI-LOCATION MARKETING · PE-BACKED PLATFORMS

One playbook, different results by location — PE platforms

Same brand, same playbook, a distribution instead of a result

DIFFERENTIAL — LIKELY CAUSES, MOST LIKELY FIRST

1. 01

   The spread is real, large, and normal — its existence is not evidence the playbook failed

   Provider-level and site-level output varies enormously at identical inputs. A platform that treats any variance as a failure will chase noise; the useful question is which term of the spread you can actually move.

   MGMA found a 7× revenue spread per FTE physician at identical staffing across 141 groups. MGMA DataDive's own medians show how much sits in specialty alone: total medical revenue per FTE physician of $760,383 in primary care, $697,712 nonsurgical, $687,652 surgical.

   MGMA DataDive Financials & Operations (2024 data); MGMA multi-group analysis, n=141

   NOT THIS IF — The spread persists inside a single specialty, a single payer mix and a single site format. Then it is not structural and something below explains it.

2. 02

   Service mix, not volume

   Two sites see comparable patient counts and bill radically different amounts, because they perform different work. No demand-side intervention changes this, and mistaking it for a marketing gap is the most expensive error on this page.

   CMS Medicare PUF, CY2024, 12,616 dermatology-typed providers: the top decile bills 4.1× the median on 2.7× the patients — the gap is mix, not panel size. Mohs first stage pays $581.52; actinic keratosis destruction beyond the first lesion pays $6.22. Allowed amounts run $26,773 at the 10th percentile, $188,072 at the median, $777,069 at the 90th.

   CMS Medicare Physician & Other Practitioners Public Use File, CY2024

   NOT THIS IF — Revenue per encounter is tight across sites and only encounter counts differ. Then it is a volume problem and demand work is legitimate.

3. 03

   Utilization — the same playbook run against differently filled calendars

   Identical marketing, identical brand, wildly different schedule density. Utilization is the single widest published operating spread in these businesses and it is rarely on the platform readout.

   Zenoti, CY2025: staff utilization 38% median, 56% at the 75th percentile, 80% at the 90th. Revenue per location tracks it — $1,860,000 median, $2,340,000 at the 75th, $4,250,000 at the 90th. Planet DDS shows the same site varying with the calendar: Friday production $7,388 against Tuesday $10,152, 27% lower.

   Zenoti CY2025 medspa platform data; Planet DDS 2026 Outlook

   NOT THIS IF — All sites sit inside a narrow utilization band and the revenue spread survives anyway.

4. 04

   Asset base — chairs, rooms and hours, not demand

   A site with fewer operatories or rooms has a hard ceiling. Comparing its total revenue to a larger site's produces a variance that is pure capacity and reads as performance.

   Planet DDS, across 59,139 chairs: revenue per chair averages $205,690 with a $156,741 median at DSO-affiliated practices, against $236,286 average and $184,502 median at solo practices. In veterinary the equivalent normalizer is published too — AVMA reports median revenue per exam room of $444,668, or $371,500 at companion-exclusive practices, and $538 revenue per square foot.

   Planet DDS 2026 Outlook (59,139 chairs); AVMA 2025 Economic State of the Profession

   NOT THIS IF — Revenue per chair or per room is tight across the platform and only site totals differ. The variance is then capacity, correctly, and it is a capex conversation.

5. 05

   Cost structure, not top line

   The revenue spread is modest and the margin spread is not, because cost per provider diverged. Nothing on the demand side touches this and it is where a great deal of platform variance actually lives.

   MGMA DataDive: median total operating cost per FTE physician rose from $620,098 to $801,938, up 29.3% over five years. Over 2011–2024, CPI rose 39.5% while operating cost rose 71.6% at physician-owned practices and 83.3% at hospital-owned. AMGA, across 5,700+ clinics, puts the median loss per physician at system-affiliated groups above $249,000 — revenue $719,901 against expenses $1,036,238.

   MGMA DataDive Cost and Revenue; AMGA 2023 (5,700+ clinics)

   NOT THIS IF — Site-level contribution margin is tight once corporate allocation is stripped out. Then the spread is genuinely top-line.

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.

WHAT RESOLVES EACH

What resolves this, and how you will know it resolved

The spread is real, large, and normal — its existence is not evidence the playbook failed	Rx 04 · marketing attribution →	A per-site readout on normalized denominators, not a platform average. The deliverable is the ability to say which part of the spread is structural and which is not — that sentence is worth more in diligence than a better average.
Service mix, not volume	not a marketing engagement →	We can make mix visible; we cannot change it. What a provider is credentialed, trained and contracted to perform is a clinical, recruiting and payer-contracting decision. Spending media against a mix gap buys more of the low-value work.
Utilization — the same playbook run against differently filled calendars	Rx 03 · patient conversion →	Route demand to the empty hours across the platform rather than to the site that happens to rank. This is the one variance cause where platform scale is genuinely an advantage.
Asset base — chairs, rooms and hours, not demand	not a marketing engagement →	A capacity ceiling is a capex decision. The useful contribution is refusing to let a capacity gap be reported as a performance gap — and refusing to sell demand into a site that cannot deliver it.
Cost structure, not top line	Rx 04 · marketing attribution →	Contribution margin by site, tied to the income statement, allocation stripped. Outside that, this is a CFO's problem and not ours.

HOW THIS DIFFERS BY SCALE

How this differs by scale

Single site	A single site has no variance problem, it has a performance level. The comparison that matters is against its own specialty's published distribution — see /marketing-problems/dental-groups-dsos/multi-location-marketing/ and /marketing-problems/dermatology-groups/multi-location-marketing/ for how those distributions are actually shaped.
Group	At a few sites the temptation is to explain each difference individually, site by site, in narrative. Normalize first. Most of what looks like five different stories is one denominator problem.
Platform	At platform scale variance is the asset, not the embarrassment — a documented distribution with a stated mechanism reads as control, while a single flattering average reads as an average. A buyer will build the site-level distribution regardless; the question is whether you brought it or they found it.

OTHER PRESENTATIONS — PE PLATFORMS & ROLLUPS

• Same-store flat while the unit count rises
• Blended CAC is rising and nobody can say why
• Demand arrives; booked, shown, treated work does not
• The growth story has to survive diligence, not just the pitch