Demand / Capacity Market Map

RX 01 · Q3 '26 MAP · BUILT AUG 7 '26 ON DATA THROUGH THE JUN '26 CLOSE · 12 LOCATIONS (11 AT DAY 0) · 6 MARKETS · SPECIMEN

REVIEWED WITH THE CFO AND THE REGIONAL DIRECTORS — QUARTERLY, 45 MIN · MONTHLY DELTA CHECK RUNS WITHOUT A MEETING · NEXT REFRESH NOV 6 '26

CONFIDENTIAL · RX 01FORM 118-M

Demand curve against capacity curve, overlaid, one panel per market. Some markets deserve zero spend — this document names them, states why, and states the condition that would overturn each ruling. Geography at metro and region grain only.

MARKETS / LOCATIONS

6 / 12

11 open at Day 0 · Parkway 11 opened Mar '26 and is excluded from every same-store line

OPEN CHAIR-HOURS, JUN '26

1,590

≈ $349K/mo of latent net production at the group's own $219.4/utilized hr — a ceiling, not a plan

RECOMMENDED SPEND CHANGE

−$6,500

per month · −13.1% · reallocation first, reduction second — the money moves before it shrinks

MARKETS RECOMMENDED AT ZERO

2 of 6

plus 1 gated to a $2,500 floor until its desk answers

THE OVERLAY — MARKET DEMAND vs. OPEN CHAIR-HOURS · TWELVE MONTHS TO JUN '26 · EACH SERIES INDEXED TO ITS OWN 12-MONTH MEAN = 100

M-01Capital Metro — Core

CENTRAL 01 · WESTLAKE 03 · NORTH 04

DEM 12,400/MO · OPEN 195 HRS · IN PHASE 2/12 MO

M-02Capital Metro — Inner Ring

MIDTOWN 02 · HILLCREST 05 · UPTOWN 12

DEM 8,600/MO · OPEN 348 HRS · IN PHASE 2/12 MO

M-03North Corridor

SAME-STORE: ROUND ROCK 06 · LAKELINE 07

DEM 9,600/MO · OPEN 221 HRS SS · IN PHASE 2/12 MO

M-04South Ring

SOUTH 08

DEM 3,800/MO · OPEN 136 HRS · IN PHASE 3/12 MO

M-05Riverside Corridor

RIVERSIDE 09 · CEDAR PARK 10

DEM 4,900/MO · OPEN 408 HRS · IN PHASE 3/12 MO

M-06High Plains — Regional

NO LOCATION · NEAREST CHAIR 41 MIN

DEM 2,300/MO · OPEN 0 HRS · NO CAPACITY CURVE EXISTS

How to read a panelMarket search demand, monthly, from Google Keyword Planner. Your open chair-hours, monthly, from your own scheduling template. Each line is indexed to its own twelve-month average, so the dotted mid-rule is that line's normal.

The ticks are the pointA tick marks a month when demand and open capacity sat on opposite sides of their own averages: the market wanted you in a month your chairs were full, or your chairs were empty in a month nobody was looking. Count the ticks. That is the misphase.

Panels are not comparable by heightEach panel is scaled to its own two series; height tells you shape, the printed figures tell you scale. M-03's capacity line is same-store — Parkway 11 is excluded. M-06 has one line because it has no chairs.

The signals, per market

DEMAND · CAPACITY · PAYER MIX · SATURATION · DRIVE-TIME · SHEET 2 OF 4 · SPECIMEN

CONFIDENTIAL · RX 01FORM 118-M

One row per market. Four of these columns come from sources a diligence team can re-run without us; the rest come from your own systems. Sheet 4 says which is which, when each last refreshed, and how confident we are in it.

MARKET ROSTER — Q3 '26 · CAPACITY AS OF THE JUN '26 CLOSE · DEMAND IS A TWELVE-MONTH AVERAGE TO JUN '26
MARKET	LOC	DEMAND SEARCHES/MO	CHAIR-HRS AVAIL/MO	UTIL %	OPEN HRS/MO	DENSITY SRCH/HR	IN PHASE	FFS + CASH %	AUCTION OVERLAP †	POP ≤20 MIN †
M-01 Capital Metro — Core	3	12,400	2,016	90%	195	6.2	2/12	44%	17	78%
M-02 Capital Metro — Inner Ring	3	8,600	1,512	77%	348	5.7	2/12	36%	12	64%
M-03 North Corridor	3	9,600	2,016	75%	503	4.8	2/12 ss	33%	9	59%
M-04 South Ring	1	3,800	504	73%	136	7.5	3/12	31%	7	52%
M-05 Riverside Corridor	2	4,900	1,176	65%	408	4.2	3/12	27%	6	46%
M-06 High Plains — Regional	0	2,300	0	—	0	n/a	none	—	4	0%
GROUP · ALL 12	12	41,600	7,224	78%	1,590	5.4	12/60	36%	†	†
same-store — the 11 open at Day 0	11	—	6,552	80%	1,308	—	—	—	—	—

DEMAND = GOOGLE KEYWORD PLANNER, 22-TERM HIGH-INTENT SPECIALTY-DENTAL SET, GEO-TARGETED, 12-MO AVG TO JUN '26. CAPACITY = YOUR OWN SCHEDULING TEMPLATE, 21-CLINIC-DAY MONTH, TIED POD-FOR-POD TO FORM 610-M P.4. DENSITY = DEMAND ÷ AVAILABLE CHAIR-HOURS; THE GROUP FIGURE EXCLUDES M-06 FROM BOTH SIDES (39,300 ÷ 7,224). ss M-03'S PHASE SCORE IS SAME-STORE — PARKWAY 11 OPENED MAR '26 AND IS EXCLUDED. † NON-ADDITIVE. NO PUBLISHED DENTAL BENCHMARK EXISTS FOR ANY COLUMN ON THIS TABLE — EVERY COMPARISON IS THIS GROUP AGAINST ITSELF.

THE FINDING THE OVERLAY EXISTS TO PRODUCE

In 48 of 60 market-months, the demand you could buy and the chairs you had open were on opposite sides of their own averages. You are not short of demand and you are not short of chairs. You are short of them at the same time.

January is the worst month in the network: benefits reset, search demand peaks in four of the five markets that have chairs, and open chair-hours hit their twelve-month low in all five. December and February behave the same way. No amount of spend fixes a January — the chairs are already full. The fix is a scheduling-template change made in October and it belongs to operations, not to media. What media can do is stop paying peak-season prices for a peak we cannot serve: invert the seasonal spend curve, weighting it into the months your panels show capacity open. That change costs nothing and is worth more than the reallocation on sheet 3.

WHAT WE DO NOT KNOW — STATED BEFORE YOU FIND IT

Parkway 11 opened Mar '26, so M-03 has no twelve-month same-store capacity series. Its curve and its phase score use Round Rock 06 and Lakeline 07 only; Parkway's 282 open chair-hours are reported as a number and never as a trend, and every same-store line in this document excludes them. Eleven of twelve locations were open at Day 0. A same-store comparison that quietly includes the twelfth is not a comparison, it is an advertisement.
The 68% displacement rate in M-01 carries the largest ruling in this document, and it is one month of one log. It is the share of M-01's new-patient bookings created by moving or declining another appointment — scheduling reschedule log, Jun '26, n = 88 bookings. The trade on sheet 3 survives down to a 26% displacement rate; below that our recommendation is wrong. One month is thin and we are saying so before you ask. The Nov refresh runs it on six.
We broke our own kept-rate rule once, in the direction of our own argument. We apply one group booked-to-kept rate of 74% everywhere (Form 247-K). For M-06 we used its own measured 41%, because applying 74% would have hidden the finding. At 74% M-06's cost per completed first visit is $501 against $534 of twelve-month contribution — it would clear, barely, and the ruling would flip. So the 41% is the ruling. It is measured on 220 booked consults over twelve months. Dispute that number and you dispute the recommendation; we would rather have that argument out loud than buried.
Two load-bearing numbers come from one cohort. The 2.25× ratio of twelve-month to 90-day collected value, and the 6.2 chair-hours a new patient consumes in its first year, both come from this group's Jul '25 cohort followed to Jul '26. One cohort, two dependencies. The Nov refresh adds the Oct '25 cohort and we will report if it disagrees.
M-06's demand figure is the only one we cannot cross-check. Our impression share there is 9%, too thin to validate Keyword Planner against our own search-terms data. The ruling does not rest on it — a market with no chair in it is not made servable by more demand — but the number is unverified and labelled as such. Drive-time coverage is also modelled, not observed: a road-network isochrone over census population, not a count of where your patients live.

Where the money goes — and stops

RANKED SPEND REALLOCATION · Q3 '26 · SHEET 3 OF 4 · SPECIMEN

CONFIDENTIAL · RX 01FORM 118-M

Ranked by where the next dollar does the most work, and it ends where it should end: in the markets that get nothing. Some markets deserve zero spend. These are the two, and here is why — including the one with the best numbers on the page.

RANKED REALLOCATION — ENDS IN ZERO-SPEND MARKETS WITH STATED REASONS · MEDIA SPEND ONLY, EXCLUSIVE OF FEES
#	MARKET	NOW $/MO	COST / COMPLETED FIRST VISIT ‡	CONTRIB @ 90 D	CONTRIB @ 12 MO	CLEARS	12-MO HRS CLAIMED / OPEN	REC. $/MO	Δ
1	M-03 North Corridor	$12,200	$288	$323	$727	90-day	+211 / 503	$22,000	+$9,800
2	M-02 Capital Metro — Inner Ring	$9,800	$321	$340	$764	90-day	+48 / 348	$12,300	+$2,500
3	M-04 South Ring	$4,300	$392	$310	$698	12-mo only	+30 / 136	$6,200	+$1,900
4	M-05 Riverside Corridor	$7,500	$373	$313	$704	gated — desk	releases 83	$2,500	−$5,000
5	M-01 Capital Metro — Core	$8,900	$241	$375	$844	90-day	229 / 195 ✗	$0	−$8,900
6	M-06 High Plains — Regional	$6,800	$905	$237	$534	NEITHER	no chairs	$0	−$6,800
	GROUP	$49,500	—	—	—	—	—	$43,000	−$6,500

‡ COST PER COMPLETED FIRST VISIT = YOUR COST PER CLICK ÷ (YOUR CLICK-TO-BOOKED RATE × YOUR BOOKED-TO-KEPT RATE), ON THE 22-TERM INTENT SET. "CLEARS" COMPARES IT TO CONTRIBUTION, NOT COLLECTIONS. "12-MO HRS CLAIMED / OPEN" IS THE CHANGE IN PAID NEW PATIENTS × 6.2 CHAIR-HOURS EACH — WHAT THE PROGRAMME CONSUMES A YEAR FROM NOW — AGAINST OPEN HOURS TODAY. M-01 HAS THE CHEAPEST NEW PATIENT AND THE HIGHEST COLLECTED VALUE IN THE GROUP, CLEARS AT NINETY DAYS, AND IS STILL GOING TO ZERO. A COST-PER-PATIENT REPORT WOULD HAVE SAID SPEND MORE THERE. THAT IS WHY WE DO NOT SEND YOU ONE.

M-01 CAPITAL CORE — ZERO · NO ROOM IN THE CHAIRS

229 hrs vs 195

Best economics in the group — $241 a patient, $605 of collected value, clears at ninety days. And we are turning it off. At 90% utilization there are 195 free chair-hours a month here, 0.77 per chair per day. The programme buys 37 new patients a month; at your own 6.2 chair-hours per patient per year that is a standing claim on 229 hours. The market does not have them.

It already shows: 68% of new-patient bookings here are made by moving someone else, third-available is 31 days against a group median of 12, and new patients break at 31% against the group's 26%. The miss is ours — third-available was 19 days at the Jan refresh. It got worse on our watch, because the acquisition programme worked. We are recommending you switch off the thing we built.

Unblock: third-available ≤14 days for four straight weeks, or the hygiene hire 610-M called for lands at North 04.

M-06 HIGH PLAINS — ZERO · NO CHAIRS AT ALL

0% within 20 min

You have no chairs in this market. Nearest one is 41 minutes away. Zero percent of this population is within twenty minutes of a chair you own.

And it shows: 59% of what we book here never arrives, against 26% everywhere else. A completed first visit costs $905 and returns $237 at ninety days, $534 at twelve months — short on both. The ones who do come travel once and largely do not return; their 90-day collected value is $383 against the group's $551.

Zeroing it releases $81,600 a year and gives up 90 new patients worth $48,060 of twelve-month contribution. Net +$33.5K a year, and 47 chair-hours a month handed back to Round Rock 06.

M-05 RIVERSIDE — GATED, NOT ZERO

$2,500 floor

This market has 408 open chair-hours — the second-largest pool you own — and a desk that answers 81% of its calls. Demand routed here leaks at the phone: about eight booked patients a month, roughly $3,900 of collected value, into voicemail.

Held at a $2,500 brand-defense floor. Unblock condition, stated in advance: answer rate ≥90% for four consecutive weeks on Form 247-K. On that week we restore M-05 to $7,500 and take it from M-03, without a meeting.

This is the one I'd overrule if the desk fix lands early. Don't wait for the quarterly.

WHAT THIS REALLOCATION IS ACTUALLY WORTH — INCLUDING THE PART THAT ARGUES AGAINST IT

At average cost per completed first visit this reallocation gains 47 new patients a month and loses 33 — net +14, on $6,500 a month less spend. Do not believe the +14. Average cost is not marginal cost: the thirty-fourth additional patient in M-03 will cost more than the first, because the cheapest inventory is the inventory we already bought.

So we model the gain at a 15% marginal penalty — not an assumption but a measurement: the M-04 spend-up test in Feb–Apr '26 raised spend 38% and cost per completed first visit 14%. At that penalty: +40 gained, −33 lost, net +7 new patients a month, worth about $64K of twelve-month contribution on top of $78K of released spend — call it +$142K a year, over half of which arrives as money that never leaves the account rather than collections someone has to earn.

Where this stops being true. The M-01 loss is counted at 32% of the patients it buys, because 68% of them displace a booking that would have happened anyway. That single number carries the recommendation. The trade survives down to a 26% displacement rate; below that we are wrong and you should overrule us. The number to watch is the reschedule log, not the ad account, and it is on the monitor on sheet 4.

Source and refresh log

SO A DILIGENCE TEAM CAN RE-RUN THIS WITHOUT US · SHEET 4 OF 4 · SPECIMEN

CONFIDENTIAL · RX 01FORM 118-M

Every figure in this document: where it comes from, who owns it, how often it moves, how confident we are. Four of the ten inputs can be re-pulled by someone who has never met us; the six that cannot are named as such, because a dependency you have not written down is a dependency you will discover late.

INPUT REGISTER — Q3 '26 · RE-RUN INSTRUCTIONS IN THE FORM 118-M SOP APPENDIX
#	SIGNAL	SOURCE	OWNER	HOW WE GET IT	REFRESH	LAST	CONF.
1	Search demand, 22-term intent set, per geo	Google Keyword Planner — historical metrics	ADMEN	Ads API · KeywordPlanIdeaService	Quarterly	JUL 28 '26	MOD.
2	CPC, impression share, auction overlap	Client Google Ads account — auction insights, geographic view	Client · ADMEN has access	Ads API · scheduled	Monthly	JUL 31 '26	HIGH
3	Click → booked consult	Client site + call tracking, offline conversion import	ADMEN — tracking	GA4 + offline upload	Monthly	JUL 31 '26	HIGH
4	Booked → kept · completed first visits	Practice management system	Client operations	Client must send · scheduled CSV to SFTP	Monthly	JUL 31 '26	HIGH
5	Available & utilized chair-hours, per location	PMS scheduling template	Client operations	Client must send · manual export, no API	Monthly	JUN 30 '26	HIGH
6	Displacement rate & third-available appointment	PMS scheduling reschedule log	Client operations	Client must send · manual export, one month held	Quarterly	JUN 30 '26	LOW ^D
7	Payer mix — FFS + cash share of production	PMS ledger reconciled to GL	Client CFO	Client must send · manual export	Quarterly	JUN 30 '26	HIGH
8	Collected value per NP & chair-hours per NP, 90-day / 12-month	PMS ledger, cohort-followed	Client CFO + ADMEN	Client must send · cohort join runs client-side	Quarterly	JUN 30 '26	HIGH / MOD. ^C
9	Population within 20-minute drive	US Census ACS 5-yr block-group population + OSRM isochrones on OpenStreetMap	ADMEN	Script — fully re-runnable	Annually / new location	MAY 14 '26	HIGH
10	Contribution margin — 62%	Client general ledger, 5000-series direct clinical cost	Client CFO	Client must send · manual	Annually / cost change	JAN 15 '26	MOD. ^G

^D LOW — ONE MONTH, n=88 BOOKINGS, AND IT CARRIES THE M-01 RULING. ^C MODERATE ON THE TWELVE-MONTH FIGURES — ONE COHORT. ^G MODERATE — 62% IS GROUP-LEVEL; PER-MARKET DIRECT COST EXISTS IN THE GL AND WE HAVE NOT SPLIT IT. ROWS 1, 2, 9 AND 10 CAN BE RE-RUN BY A THIRD PARTY WITH ACCOUNT ACCESS AND NO HELP FROM US. ROWS 3–8 REQUIRE YOUR SYSTEMS; EXPORT SPECS ARE IN THE SOP APPENDIX SO A SUCCESSOR CAN REQUEST THEM CORRECTLY ON DAY ONE.

PHI POSTURE — A COMPLIANCE STATEMENT, NOT A LABOR ONE

Nothing in this register carries protected health information. Rows 4–8 originate in a practice management system that does; they are exported as aggregates only — counts, hours and dollars by location and month. The cohort join in row 8 executes inside your environment and returns counts and dollars only; no patient-level record leaves it and ADMEN holds no PMS credential. If a future refresh needs a patient-level join it stops and goes to your compliance officer first — whether or not anyone is watching.

CADENCE — WHY THIS IS NOT A MONTHLY DOCUMENT

This map refreshes quarterly. Its load-bearing inputs — payer mix, collected value per new patient, capacity template, displacement rate, contribution margin — move quarterly at best, and five need a human at your end to export a file. A monthly version would restate the same numbers with new noise on top and teach you to stop reading it. Between quarters a delta check runs monthly with no document and no meeting — an automated monitor producing one paragraph, same-day, only when something breaks. Thresholds are fixed in advance so nobody argues about them afterwards:

Cost per completed first visit in any funded market moves ±15% against this map — or M-03 crosses $360, where sheet 3's largest increase stops clearing.
Impression share in any funded market moves ±10 points, or auction overlap gains 3+ new advertisers.
Open chair-hours in any market move ±20% — usually a provider left and nobody told media.
A zero-spend ruling's condition is met — M-01's third-available comes inside 14 days for four consecutive weeks, or M-05's desk clears 90% answer rate for four consecutive weeks.
M-01's displacement rate falls below 35% — eleven points above the 26% break-even, so the alarm fires before the recommendation does.
The monitor fails to run, or an input goes stale past its refresh interval. Silence is reported as a failure, not as good news.

Next full refresh: NOV 6 '26 — first Friday, with the Oct board pack, six months of reschedule log behind the displacement rate and a second cohort behind the twelve-month figures. Both zero-spend rulings and the M-05 gate re-open then, flagged in the file so nobody has to remember.

— specimen —
PREPARED — ADMEN ANALYST
— specimen —
REVIEW GATE — SECOND PARTNER
— specimen —
ACCEPTED — CLIENT CFO · DATE