DX / Differentials / SPECIALTY DENTAL

D-04 — MULTI-LOCATION MARKETING · SPECIALTY DENTAL

One playbook, different results by location — specialty dental

Same brand, same protocol, different starts per site

DIFFERENTIAL — LIKELY CAUSES, MOST LIKELY FIRST

1. 01

   Geography is doing the work, not the playbook

   Site performance tracks the market the site sits in. Comparing sites to each other rather than each site to its own market makes a normal site look broken.

   Median case starts by region for solo orthodontic practices: South 536, Midwest 374, West Central 329, Northeast 319, Pacific 306 — a 1.75× spread. Median gross income runs from $2,510,093 in the South to $1,227,727 in the Northeast, and JCO footnotes these differences as statistically significant at or below the .01 probability level. The Northeast has reported lower income than other regions in every study since 1981. Community size matters too: in cities of 50,000–500,000, 52% of practices describe themselves as "not busy enough" against 30% overall.

   2023 JCO Orthodontic Practice Study, Part 1, Tables 3 and 10 — medians, solo full-time US practices, n=116, calendar 2022. Texas and Utah both sit in JCO's West Central region.

   NOT THIS IF — All your sites are inside one metropolitan market with overlapping catchments. Then geography is held constant and the variance is real.

2. 02

   It is the lead specialist's career stage, not the site

   Practice output rises steeply with years in practice, peaks, then declines. A group that staffs new sites with early-career specialists and mature sites with mid-career ones will see a spread that has nothing to do with the sites.

   Median activity by years in practice: 2–7 years produces $902,500 gross income on 319 case starts at 52% overhead; 14–20 years produces $3,055,027 on 647 starts and 1,393 active cases at 60% overhead; 31+ years falls back to $1,729,709 on 294 starts at 58% overhead. That is a 3.4× spread in gross income and 2.2× in starts across career stage — larger than the entire regional spread.

   2023 JCO Orthodontic Practice Study, Part 1, Table 2. Solo full-time practices only; JCO's own narrative: "median income, expenses, and numbers of cases generally increased up to 14-20 years in practice and then declined."

   NOT THIS IF — Your site leads are all inside the same tenure band. Then career stage is held constant and the ranking must come from somewhere else.

3. 03

   Each site has its own referrer portfolio, and they are not comparable

   In endo, OMS and perio the site does not inherit the brand's demand — it inherits whatever referring doctors its own specialist has assembled. Two sites with identical marketing can have referrer portfolios that differ by an order of magnitude.

   Annual referred-patient volume per endodontic practice: 10% receive 100 or fewer, 13% receive 101–500, 31% receive 501–1,000, 25% receive 1,001–2,000 and 21% receive more than 2,000. The standard deviation is 3,234 patients against a mean of 1,930 — a coefficient of variation more than 1.5× the mean. The top ten responses ranged from 13,000 to 35,000 patients while fourteen respondents reported 50 or fewer. Solo practices run a mean 1,137 / median 1,000; group practices a mean 2,892 / median 1,500.

   AAE Referral Patterns Survey Report, April 2026, p.16 — n=581 endodontists. The report states the distribution explicitly and warns that any single average endodontic practice figure is a fiction.

   NOT THIS IF — Your sites are consumer-demand driven — pediatric dentistry, or adult-heavy orthodontics — where the demand engine is recare and consumer reach rather than a referring-doctor portfolio.

4. 04

   Payer mix, not performance — and in pediatric dentistry it is severe

   Two sites produce the same dentistry and collect very different amounts because they sit on different payer bases. The variance is in collections and never appears in production.

   Pediatric payer mix across a large client base: insurance 56.05%, fee-for-service 24.24%, Medicaid 19.72% — but rural pediatric practices run 52% Medicaid against 13.98% for suburban and urban. Collections as a percentage of production run 75.48% average (68.89% median), roughly twenty points below the 96–98% collection ratio the specialty's own consultant guidance calls healthy. And the underlying demand differs by state: child Medicaid/CHIP dental utilisation was 49% in 2024, ranging from 30% in Ohio to 62% in Texas, with 27 states below half.

   Cain Watters & Associates, "Pediatric Practice Comparison Report, 2022 Edition" — real client books, no sample size disclosed. Collection-ratio target: Weir J, "Business Metrics for the Pediatric Dental Practice," Pediatric Dentistry Today, March 2017 — published by the AAPD but explicitly disclaimed as "opinions and recommendations of the author," so it is a target, not a measurement. Utilisation: ADA HPI analysis of CMS-416 EPSDT reports, "Dental Care in Medicaid Programs," December 2025 update. NOTE: the 75.48% ratio is of production (pre-adjustment) and the 96–98% target is of what can be collected (post-adjustment) — the gap is contractual write-offs, not collection failure.

   NOT THIS IF — Your sites share one payer contract portfolio and one payer mix. Then collections variance is operational.

5. 05

   Scale bought the strong site a case type the weak site cannot do

   The high site is not executing better; it has access to sedation, general anaesthesia or hospital privileges that change its case mix. This is a capital and credentialing question wearing a marketing costume.

   Hospital affiliation runs 66.04% across pediatric practices overall — but 85.71% of large practices against 14.29% of small ones, the sharpest large/small separator in the dataset, with daily hospital general-anaesthesia production of $9,895. In oral surgery the analogue is priced by the capital markets directly: valuation bands make "multi-site, anesthesia capabilities" the named characteristic of the $3–5M EBITDA tier at 9–11× EV/EBITDA.

   Cain Watters & Associates, "Pediatric Practice Comparison Report, 2022 Edition." OMS bands: FOCUS Investment Banking, "Oral Surgery Practice Valuation 2026 Update," which names its inputs (PitchBook Q1 2026, Bain Global Healthcare PE Report 2026, VMG Health 2026) but does not attribute figure to source line by line, and disclaims that actual pricing varies by transaction.

   NOT THIS IF — Every site has equivalent sedation and general-anaesthesia capability and equivalent hospital access. Then case mix is available to all of them and the difference is in how it is booked.

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

Geography is doing the work, not the playbook	Rx 04 · marketing attribution →	The fix is a comparison set, not a campaign: each site against its own market rather than against the group's best site. Half of multi-site variance conversations end here, and end with nothing being bought.
It is the lead specialist's career stage, not the site	Rx 03 · patient conversion →	You cannot buy tenure. What is transferable is the observable practice of the peak-cohort site — consultation structure, follow-up on patients placed on observation, how the treatment recommendation is presented. Note that in orthodontics the observation pool is a real pre-treatment asset with no general-dentistry analogue: the median solo practice carries 243 patients on observation against 334 annual starts.
Each site has its own referrer portfolio, and they are not comparable	Rx 01 · patient acquisition →	Referrer development is site-local because referring doctors are site-local. A group-level program that does not resolve to named referring doctors per site is spending on an average that does not exist anywhere in the group.
Payer mix, not performance — and in pediatric dentistry it is severe	Rx 04 · marketing attribution →	Production, adjustments and collections reported separately, per site, every month. Collapsing them into one figure is the single most common way a payer-mix problem gets misdiagnosed as a marketing problem — and it is the distinction dentistry cares most about.
Scale bought the strong site a case type the weak site cannot do	Rx 05 · preparing for sale →	This is a capital and credentialing decision, not a demand problem, and we would say so plainly rather than sell into it. It belongs in the same conversation as valuation, because the capital markets already price it: anaesthesia capability is a named driver of the OMS multiple band.

HOW THIS DIFFERS BY SCALE

How this differs by scale

Single site	"Solo" in orthodontics does not mean one office — the median solo practice runs 0.7 satellite offices and 49% have at least one. A single-owner practice with a satellite already has multi-site variance and usually has never measured the satellite separately.
Group	Two to nine doctors is where the variance is most often misread, because the consolidated number is healthy. Every diagnostic above has to be run per site before it is run for the group, and referrer counts have to be per site, because referring doctors do not follow a brand across a metro.
Platform	Above ten locations there is no external benchmark at all: the ADA publishes no DSO or group affiliation figure by specialty, and no specialty-dental analogue of a cross-platform new-patient index exists. A platform's only honest comparison set is its own site distribution and its own trailing cohorts — which, run properly, is a more useful instrument than any industry average would have been.

OTHER PRESENTATIONS — SPECIALTY DENTAL

• Plateaued growth in a specialty whose aggregate market is still growing
• Rising cost per start — and cost per referral, which nobody publishes
• Referrals arrive; starts and treatment do not follow
• Making the referral book read as a system, not a relationship