REQUIRED STILLS
108
12 locations × 9 · fixed set
Real operators, real rooms. Stock reads as stock — patients can tell, and so can a diligence team. Every frame on this list has a destination before the camera comes out of the bag.
SCOPE — The library covers all 12 locations from the day each one opens. The same-store financial instruments (Form 247-B, Form 610-M) report the 11 sites open at Day 0 and hold Parkway — 11 (opened Mar '26) on its own line. The library was built to the Day-0 roster and did not follow that rule, which is why sheet 3 has a zero row. The locked brand elements this footage must satisfy live on Form 302-G — check #7 of the monthly brand sweep fails any published photograph whose asset ID does not resolve to this library.
01 · SLOT MAP — WHAT EACH ASSET IS FOR — an asset with no destination does not get shot. That is the whole discipline.
| SLOT | WHERE IT LANDS | FORMAT · DELIVERY PX | LENGTH | PER LOC | WHY IT EXISTS |
|---|---|---|---|---|---|
| L1 | Location page — arrival block | 3:2 · 1600×1067 | still | 1 | the door they will actually walk through; the only frame where signage belongs |
| L2 | Location page — the rooms | 3:2 · 1600×1067 | still | 3 | answers "is it clean, is it modern" before a patient has to ask |
| L3 | Location page — the desk | 4:5 · 1200×1500 | still | 1 | the people who answer the phone (Form 247-K); faces need a signed release |
| L4 | Location page — provider block | 4:5 · 1200×1500 | still | 1 / provider | named-provider blocks; also feeds the referring-GP kit |
| L5 | Local listing — profile + posts | 1:1 · 1200×1200 | still | 2 | crops square, so shoot centre-safe or lose the frame |
| L6 | Paid social — feed & story | 4:5 video · 1080×1350 | 8—12 s loop | 1 | autoplays muted. No shot may depend on speech to read. |
| L7 | Paid search — image assets | 1:1 1200×1200 + 1.91:1 1200×628 | still | derived | never shot separately — cropped from L1 / L2 |
| L8 | Booking confirm + reminder email | 3:2 · 1200×800 | still | 1 | shows the room they are coming to. HYPOTHESIS — untested, see the note below. |
| L9 | Careers page + job card TPL-07 | 4:5 · 1200×1500 | still | 1 | hiring is a live constraint — an open desk req and a hygiene req |
| L10 | Group hub — network montage | 16:9 · 1920×1080 | 15—20 s loop | group | the hub carries authority; one plate per location cuts into one montage |
| L11 | Referring-GP kit TPL-04 | 3:2 · 1600×1067 | still | derived | cropped from L2 / L4 — referrals are a channel in specialty dental |
| 11 SLOTS | REQUIRED SET PER LOCATION | 9 STILLS + 1 LOOP + 1 PORTRAIT / PROVIDER | 9 | 2 of 11 slots are crops, not shots |
Two of the eleven slots are never photographed — they are crops. Shooting them separately is how a library becomes four hundred files nobody can find. The required set is nine stills (1 arrival + 3 rooms + 1 desk + 2 square details + 1 careers + 1 reception wide), one vertical loop, and one portrait per credentialed provider on the page. Nine, not "as many as we get." A fixed denominator is what makes sheet 3 auditable.
DELIVERY CROPS — SHOOT LOOSE, CROP LATER, NEVER THE REVERSE relative proportions, to scale
WHAT WE CAN PROVE, AND WHAT WE ARE REFUSING TO CLAIM
Facts about the pages. At Day 0, 9 of the 11 location pages ran stock photography. Three stock images appeared on 5 of them. Two offices nine miles apart showed the identical operatory. Those are observations about files, and any diligence team can re-run them in an afternoon.
What we have not proved. We cannot say that real photography books more consults than the stock it replaced at this group. The location-page template (TPL-01, Form 302-G) changed in the same release as the Block 1 photography, so the two effects are confounded and the comparison is worthless. There is no published benchmark for how location photography affects booking rate in dentistry. We looked. We are not going to manufacture one, and we are not going to reuse an agency-blog number as though it were research.
How we get a clean answer. Block 3's locations roll staggered — photography first, page template two weeks later — so the two changes separate. First honest read: Feb '27, on n=5, which we will report as small. Until then the argument for real photography is the one on the site: a judgement, not a measurement. Slot L8 carries the same warning — showing a patient the room they are coming to may reduce broken appointments, and at 26% at the weak desk that would matter, but it is a hypothesis and it stays marked one until it is tested.
The same run, in the same order, at every location — because a library is only auditable if the twelve halves of it were shot the same way.
02 · THE PER-LOCATION RUN — 11 SETUPS, 3.5 HOURS — the same run at every location, in this order, so 12 libraries are comparable to each other.
| # | SETUP | FEEDS | FRAME | CAPTURE | MOVE / LIGHT | THE CONSTRAINT |
|---|---|---|---|---|---|---|
| S01 | Arrival — facade + door from the parking approach | L1 | 3:2, loose | 12+ frames | tripod, eye level · avoid midday top-light | signage in frame here and nowhere else |
| S02 | Reception, wide, empty | L8 | 3:2 | 8 frames | tripod · kill overheads, bounce off ceiling | no patients in frame. No exceptions. |
| S03 | Front desk — staff working, hands and mid-body | L3 | 4:5 | 20+ frames | tripod · window key | releases signed before the camera comes out |
| S04 | Operatory A — chair-side wide | L2 | 3:2 | 10 frames | tripod · practicals + bounce | screens off, or the demo record only |
| S05 | Operatory B — tray, cabinetry, instrument detail | L2 L5 | 3:2 + 1:1 safe | 12 frames | handheld ok · practicals | nothing patient-identifying, ever |
| S06 | Operatory C — the chair from the patient's eyeline | L2 | 3:2 | 8 frames | tripod low · practicals | the frame patients want and nobody shoots |
| S07 | Provider portrait — standing, room soft behind | L4 | 4:5 | 25+ / provider | tripod · single soft key at 45° | one per credentialed provider on the page |
| S08 | Sterilisation / turnover workflow detail | L5 | 1:1 centre-safe | 10 frames | handheld · overhead + bounce | no faces, so it never blocks on a release |
| S09 | Careers frame — team in the corridor, mid-action | L9 | 4:5 | 15 frames | tripod · ambient | needs 3+ staff and 3+ signed releases |
| S10 | Vertical loop — one continuous action, rest to rest | L6 | 4:5 video, 24 fps | 4 takes × 20 s | locked off · as S03 / S04 | muted — start and end on the same pose |
| S11 | Hub montage plate — one wide of the space | L10 | 16:9 video | 3 takes × 15 s | locked or very slow push · ambient | one plate per location, one group montage |
| 11 | SETUPS · 3.5 HRS PER LOCATION | ~140 frames + 7 clips | ~40 min of it is waiting for a room |
The binding constraint is not the camera — it is an empty operatory. Roughly 40 minutes of every 3.5-hour run is spent waiting for a room to clear, which is why the run is scheduled against the location's own book and not against a crew calendar. Two locations in one day works only where both schedules carry a midday block. That is true at every site in this group, and it is the single fact that makes the economics on sheet 4 work.
GOVERNANCE — HOW THE LIBRARY STAYS TRUE, AND WHO OWNS IT AT DISCHARGE
{loc}-{slot}-{seq}-{yymm}, e.g. 11-L2-03-2609. Check #7 of the monthly brand sweep (Form 302-G) fails any published photograph whose ID does not resolve here. An asset with no ID is not in the library.
24 months, or immediately on: remodel, signage change, provider hire or departure, rebrand, a new location at lease signature, or any room reconfiguration a returning patient would notice.
One per identifiable person, three-year term, stored against the asset ID. Six expire in Q1 '27 — a re-shoot trigger, already on the calendar, and the reason Block 5 is sized at two days.
Inventory reviewed monthly, 20 minutes, first Friday, with the group marketing coordinator, alongside the pack (Form 610-M). Shoot blocks quarterly. This document is the agenda; there is no separate deck.
The coordinator owns the inventory and the trigger calendar; ADMEN owns the shot list, crew and post. At discharge the coordinator owns all of it — this document is written so that hand-off needs nobody in the room.
MARGIN NOTE — WRITTEN IN THE FIRST-FRIDAY REVIEW
The instrument said 100% for three months because we wrote the denominator by hand. Count what the client has, not what we planned for.
What exists, what is missing, what is stale — per location, against a fixed denominator. A library you cannot audit is a folder, not a library.
REQUIRED STILLS
108
12 locations × 9 · fixed set
ON HAND, IN DATE
91
84% · 17 frames short
LOOPS
9 / 12
one shot and never cut — our backlog
PROVIDER PORTRAITS
25 / 31
6 short · one is a declined release
LOCATIONS WITH A GAP
7 / 12
5 partial · 1 stale · 1 never shot
STILL COVERAGE BY SHOOT BLOCK this group against itself — there is no industry figure here, because none exists
| LOCATION | CHAIRS | LAST SHOOT | STILLS / 9 | LOOP | PORTRAITS HAVE / NEED | EXPIRES | STATUS | THE GAP, AND WHOSE IT IS |
|---|---|---|---|---|---|---|---|---|
| Central — 01 | 4 | Oct '25 | 9 | ✓ | 3 / 3 | Oct '27 | COMPLETE | — |
| Midtown — 02 | 3 | Oct '25 | 9 | ✓ | 2 / 2 | Oct '27 | COMPLETE | — |
| Westlake — 03 | 4 | Oct '25 | 9 | ✓ | 2 / 3 | Oct '27 | PARTIAL | associate hired Jun '26, portrait owed. Ours — the hire trigger did not fire. |
| North — 04 | 4 | Oct '25 | 9 | ✓ | 3 / 3 | Oct '27 | COMPLETE | the reference set — every run is graded against it |
| Hillcrest — 05 | 3 | Oct '25 | 9 | — | 2 / 2 | Oct '27 | PARTIAL | loop shot Oct '25, never cut. Ours — nine months of post backlog. |
| Round Rock — 06 | 4 | Oct '25 | 6 | ✓ | 3 / 3 | VOID JUL '26 | STALE | remodelled Jul — 3 room frames show the old cabinetry |
| Lakeline — 07 | 4 | Feb '26 | 9 | ✓ | 3 / 3 | Feb '28 | COMPLETE | — |
| South — 08 | 3 | Feb '26 | 9 | ✓ | 2 / 2 | Feb '28 | COMPLETE | — |
| Riverside — 09 | 4 | Feb '26 | 9 | ✓ | 2 / 3 | Feb '28 | PARTIAL | an associate declined a release. Stays open. We do not shoot people who have not signed. |
| Cedar Park — 10 | 3 | Feb '26 | 8 | ✓ | 2 / 2 | Feb '28 | PARTIAL | no exterior — scaffolding on the facade until Oct. Nobody's fault. |
| Parkway — 11 | 4 | never | 0 | — | 0 / 3 | — | NONE | opened Mar '26, between blocks. Ours. Page has run a stock operatory for five months. |
| Uptown — 12 | 3 | Feb '26 | 5 | — | 1 / 2 | Feb '28 | PARTIAL | we cut the day short to make a flight. Ours entirely — pickup is on our nickel. |
| 12 LOCATIONS | 43 | 91 / 108 | 9 / 12 | 25 / 31 | 5 ✓ · 5 ◐ · 1 ✕ · 1 ∅ | 125 of 151 assets in date — 83% |
Four of the seven gap rows say "ours". That is not humility, it is the only way this table stays usable: a coverage report that only ever blames the client is one nobody trusts the next time it flags something real. The row that matters is Parkway. It opened in March, sixteen weeks into the second half of the engagement, and no shoot was ever scheduled — because the shoot calendar, like the brand sweep on Form 302-G, was built from the Day-0 roster of eleven. Until June the inventory reported 11 of 11 locations covered: 100%, green, and wrong. Parkway was not in the denominator. The denominator now comes from the PMS location table, and a new site enters the shoot calendar at lease signature rather than at opening.
A crew day costs the same whether it covers one location or two. Everything affordable about a per-location library follows from that one sentence.
BLOCK 3 — 2 DAYS, 5 LOCATIONS, ONE INVOICE · PASS-THROUGH AT COST · GL 6200
LOADED CREW DAY
$3,850
lead $2,200 + second $800 + kit $250 + ground & per diem $600
BLOCK TOTAL
$9,650
2 crew days + post + releases & tagging · no lodging, no air
PER LOCATION TOUCHED
$1,930
1 full set, 3 pickups, 1 loop cut from footage we already own
THE WHOLE ARGUMENT
$4,560 → $710
the first location on a crew day vs the second one. The day is already bought; the second site costs only its own post.
| DAY | LOCATIONS | WHAT GETS SHOT | DRIVE | COST | WHAT IT CLOSES |
|---|---|---|---|---|---|
| 1 · Sep 9 | Parkway 11 + Round Rock 06 | full set (9 stills, 1 loop, 3 portraits) + 3 room frames post-remodel | 22 min | $4,890 | the zero row · brand exceptions E-08, E-11 |
| 2 · Sep 10 | Uptown 12 + Westlake 03 | 4 stills, 1 portrait, 1 loop · 1 provider portrait | 19 min | $4,510 | the Feb short day · the missed hire trigger |
| — · post only | Hillcrest 05 | loop cut from Oct '25 footage — no crew, no travel | — | $250 | nine months of post backlog |
| 2 DAYS | 5 LOCATIONS | 1 FULL SET · 3 PICKUPS · 1 LOOP CUT | $9,650 | 24 of 26 open asset gaps |
Two locations are deliberately not on this block. Cedar Park 10's exterior stays open until the scaffolding drops in October — and when it does, the office manager shoots it on a phone against a written frame spec, reviewed before it ships. We do not buy a crew day for one exterior. Riverside 09 stays open indefinitely: an associate declined a release, so that page runs eight photographs instead of nine for as long as that is true.
| MEASURE | JUL 31 '26 | AFTER BLOCK 3 | REMAINING GAP, NAMED | BASIS |
|---|---|---|---|---|
| Stills in date | 91 / 108 | 107 / 108 | Cedar Park 10 exterior — scaffolding until Oct | +16 frames |
| Vertical loops | 9 / 12 | 12 / 12 | none | 2 shot + 1 cut from existing footage |
| Provider portraits | 25 / 31 | 30 / 31 | Riverside 09 — release declined, stays open | +5 portraits |
| Total assets in date | 125 / 151 | 149 / 151 | two gaps, both named above | 83% → 99% |
| Full 12-location refresh, from zero | — | $31,620 | 6 crew days at two locations a day + post | $2,635 per location |
| Refresh amortised over the 24-month asset life | — | $1,318 / mo | 1.9% of $69.4K fully-loaded growth spend (Form 610-M) | GL 6200, pass-through |
Production is a pass-through, not retainer work, and we do not mark it up. Block 3 is 13.9% of one month's fully-loaded growth spend, spent once, and holds for two years. The margin in this deliverable is the shot list and the library discipline — the slot map, the fixed nine-frame set, the asset IDs, the trigger calendar. Not the camera. And a caution the board should hear now: these numbers are this friendly because all twelve sites sit inside one drive-time cluster. The first de novo outside it costs $6,040 — its own crew day, its own flight, its own hotel night, one location's post — and that trigger should fire at lease signature, not at opening. We have now made that mistake once.
WHAT WENT WRONG IN BLOCKS 1 AND 2four of the five are ours
A location opened in March and no camera has been in it. Our own site says stock reads as stock; for five months we were the ones publishing it, at one of this group's twelve locations. Worse, the inventory read 100% throughout, because the denominator was the Day-0 eleven. The roster now reads from the PMS and a new site enters the calendar at lease signature.
We cut the February day short to make a flight — five of nine stills, one of two portraits, no loop. Nobody's fault but ours, so the Sep 10 pickup sits inside this block at our cost rather than being re-billed. Flights are now booked for the evening after the last shoot day.
Hillcrest 05's loop was shot in October and has never been cut. Nine months of an asset the group already paid for, sitting unusable, because loop cuts queued behind page builds. New rule: a loop is cut inside 10 business days of the shoot, or the shoot is not closed.
Round Rock 06 remodelled in July and Westlake 03 hired an associate in June. Both should have fired a pickup inside a week; both sat until the July sweep found them. The trigger list existed — the calendar entry did not.
An associate declined a release. This is the one miss that is not a failure. We do not photograph people who have not signed, the page runs eight frames instead of nine, and the row stays flagged on sheet 3 rather than being quietly reclassified.