The Dental Performance System

The practice plan

One page: what you are building, what has to be true for it to happen, and the three things you are doing about it this quarter.

The practice plan

This year's goals

Your future

Where this ends up

This year

Done by Collections Owner income One key number

Clinical model

Core values

What has to be true

Who owns what

What we are saying no to

Forces we are planning around

One thing to fix

—

One decision to make

One person to help

This week

Did we run the practice well this week? The targets change depending on what you are building.

	Target	This week	Prior

This month

Is the business healthy? Net production is calculated for you.

	Target	Month to date	Last month

This quarter

Three priorities for the quarter. Tap a dot to change its status.

How to read this page

Five minutes, once a week, before the first patient.

Where we are

Thirteen numbers, each against a target that came from your destination. Two practices with identical numbers and different destinations get different targets here.

What is off, and by how much

The bar is variance measured in bands — one band is one meaningful unit of concern for that particular number. It reads as a shape before it reads as numbers.

The one thing that deserves you

Not the worst-looking number. The highest weighted deviation among the things you can actually move, with the evidence and a first action attached.

What it will never tell you

It will not flag an outcome. Collections, production and owner income carry no weight, because "collections are down" is not something a person can do on Tuesday. It names the driver underneath.

The chain, not the number

Production → adjustments → net production → collections. A practice can grow production and collect nothing. The month block shows every link so the leak has somewhere to show up.

What it is not

It is not a dashboard to browse, and not a report that gets filed. If it takes longer than five minutes to read, somebody has added something that should not be there.

Owner: five minutes here, 45 in the weekly leadership meeting, 60 once a month on the financials, half a day once a quarter — about 85 minutes in a normal week. Office manager: 15 minutes on Friday updating this page, 10 minutes a day preparing the huddle sheet, plus the weekly and monthly meetings — about two and a half hours a week. Hygiene lead: the huddle, the weekly meeting, one priority — about an hour. The daily huddle is not on the owner's clock. It belongs to the practice and runs whether you are in the building or not.

Diagnostic

Twenty statements about how the practice runs. The owner scores each one, then you check it against what you can see. Write down where the two answers differ.

Score the twenty

1 means it does not happen. 3 means it happens sometimes, depending who is working. 5 means it happens every time, and it happened last week while the owner was out.

How the practice scored

The twenty, and the evidence behind them

Score each one 1 to 5. You have to see proof. What the owner thinks is a starting point, not the answer.

What each number means

1 — not at all. It does not exist.
2 — occasionally, with no standard. It happens when someone remembers.
3 — sometimes, inconsistently. It exists but depends who is working.
4 — usually, with known exceptions. It holds unless something unusual happens.
5 — consistently, and without the owner. It ran last week while the owner was out.

Rule: Owner proposes. Evidence sets the number. Difference gets written down. The gap is usually more useful than the score itself.

The twenty, and what you have to see

Patients · 1

Every new patient is asked where they came from, and we review the answers monthly.Evidence: Pull the source field for the last twenty new patients. Count the blanks. · Source: Front office

Patients · 2

We ask for a review at the chair, by name, when a patient says something good.Evidence: Count reviews left in the last 90 days against patients seen. · Source: Front office

Patients · 3

New patient calls follow a script that everyone who answers the phone has been trained on.Evidence: See the script. Listen to three recorded calls. · Source: Office manager

Patients · 4

Treatment is presented the same way regardless of who presents it.Evidence: Observe two presentations by different people. · Source: Implementer, in operatory

Patients · 5

Patients who leave without scheduling are contacted within a defined window.Evidence: See the written window and the last 30 days of contact log. · Source: Front office

Process · 6

Insurance verification happens before the appointment, on a documented schedule.Evidence: See the schedule. Sample five of tomorrow’s patients. · Source: Front office

Process · 7

Our recare reactivation sequence runs on its own, without me prompting it.Evidence: See the written sequence and evidence of the last run. · Source: Office manager

Process · 8

Claims are followed up on a defined aging cadence, not when someone notices.Evidence: See the schedule in writing and the aging worked date. · Source: Claims owner

Process · 9

We hold a morning huddle every day the practice is open.Evidence: Count huddles held in the last ten clinical days. · Source: Whole team

Process · 10

The five processes that matter most are written down, not carried in someone’s head.Evidence: Ask to be handed them. Reading time counts, describing does not. · Source: Office manager

Team · 11

Every function in this practice has one named person accountable for it.Evidence: The seat list, complete, with no blanks. · Source: Owner

Team · 12

I have a written list of things I have stopped doing personally in the last twelve months.Evidence: See the list. Verbal recall scores 2 at most. · Source: Owner

Team · 13

If our most senior hygienist resigned tomorrow, we have a plan that does not involve me working more.Evidence: Ask the owner to describe the plan in sixty seconds. · Source: Owner

Team · 14

New team members follow a written induction sequence rather than shadowing whoever is free.Evidence: See the sequence. Ask the most recent hire what happened. · Source: OM and newest hire

Team · 15

Someone other than me could run the practice for a week without calling me.Evidence: Count calls received on the owner’s last week away. · Source: Owner

Money · 16

I review a P&L against a budget every month.Evidence: See the last three months, with a budget column. · Source: Owner

Money · 17

I know my overhead by category as a percentage of collections.Evidence: Ask for the number without looking. Within 3 points scores 4+. · Source: Owner

Money · 18

I know what my compensation would be if I paid market rate to a doctor producing what I produce.Evidence: Ask for the number. Not knowing scores 1, regardless of anything else. · Source: Owner

Money · 19

AR over 90 days is reviewed and worked on a set schedule.Evidence: See the calendar entry and the last worked date. · Source: Claims owner

Money · 20

I know where the next dollar of surplus should go, and why.Evidence: Ask. A hesitant answer is a 2; a reasoned one with a number is a 5. · Source: Owner

How the score adds up

Area score = sum of its five items, out of 25.
Whole-practice score = all twenty added up, out of 100.
Bands: under 40 = no system · 40–65 = partial, owner still required · 65–85 = system exists · over 85 = runs without the owner.
Unanswered = zero. Never invent a score.

Having more than one person score it

Owner, office manager and one clinical team member score the same twenty independently, without seeing each other’s answers.
Divergence = owner score minus the mean of the other two, per item.
Divergence of two or more gets flagged. Items 1, 3, 4, 6, 7, 8 diverge most — owners overrate front-office process almost every time.
Show the divergence first. Show them the gap before you show them the score.

The first conversation, word for word

Ten questions, in order, and how to get the owner to pick one thing and put a date on it.

Ninety minutes, owner only. No numbers in the room. The implementer asks; the owner talks. Do not read the four destinations until question six — name them early and the owner picks the one that sounds best, then answers everything else to match.

"Before we look at a single number, I need one thing from you: what you are building, and when you expect to be done building it. Everything after today depends on that answer, and if it changes later the whole system changes with it. So take your time and do not give me the answer you think I want."

Opening frame. Say it close to verbatim. It sets up the forced choice later.

1 · "Fast-forward five years and you are happy with how this turned out. Describe the day. Where are you, how many days are you in a chair, who is running the place while you are not there?"

Listen for: clinical days, whether they appear at all, whether other locations appear, whether they lead with money or with time. Do not interrupt. Write down the first three nouns they use.

2 · "Do you intend to still own this practice five years from now?"

The first fork. Yes leads to owner, independent or group. No or unsure leads to sale. "I don't know" is an answer — record it and keep going.

3 · "If you do sell it, do you already know to whom — an outside buyer or a group, or someone who already works here?"

Internal succession is a sale with a financeability overlay, not a separate destination. Note it as sale, internal.

4 · "Three years from now, how many days a week do you want to be treating patients?"

Four or more points to owner practice. Two to three points to independent. Fewer than two with expansion talk points to group. This is the single most predictive question in the script.

5 · "When this ends, which matters more to you — the income you took out along the way, or the number on the cheque at the end?"

Along the way points to owner or independent. The cheque points to sale or group. If they say both, ask which one they would give up first.

6 · "Here are the four. Read them. Which one wins when two of them conflict?"

Now read the definitions. Force one. "Primary" is the word to use — you are not asking them to abandon the others, you are asking which one wins.

7 · "Give me a month and a year."

If they resist: "If you cannot name a month, name the year. If you cannot name a year, we do not have a purpose — we have a preference, and preferences do not change targets." Do not proceed without a date.

8 · "Tell me the second one you would have picked."

Record as secondary. Explain what secondary does: nothing to targets or weights, used only to break ties between priorities of equal fit. Say that plainly, or they will hear it as a consolation prize.

9 · The stress test. "If building the group meant your take-home dropped thirty per cent for two years, would you still do it?" Substitute the cost that actually bites for their choice.

Owner practice: "If a partner offered you double your income for four more clinical days, would you take it?" · Independent: "If stepping back cost you fifteen per cent of production for a year?" · Sale: "If getting a premium meant two years of earn-out working for someone else?" A destination that does not survive its own cost is not a destination.

9b · If the practice has associates: "Do your associates know what you are building? And does what you just chose make their deal better or worse?"

Most owners have not thought about this and the honest answer is usually "worse, or at least tighter." A sale destination means scrutiny on lab spend, chair-time density and scheduling; a group destination means standardised protocols. Both land on the associate as reduced autonomy. Get the answer written down now, because it becomes the hardest conversation in month two.

10 · "Say it back to me in one sentence."

If they cannot say it in one sentence without hedging, the interview is not finished. Go back to question four.

What a real answer looks like

One primary destination, named by the owner, not offered by the implementer
A month and a year
It survives the stress test in question nine
The owner can state it in one sentence without qualifying it
It does not contradict a commitment already made — a signed ten-year lease, a partnership promise to an associate, a note with a personal guarantee maturing inside the horizon

What does not count

"All of them." Return to question five.
No date. Stop. Reschedule. Targets built on a date nobody named are targets nobody believes.
A spouse or partner in the room who visibly disagrees. Stop and get them in the next conversation. The destination that holds is the household's, not the owner's.
A destination chosen to impress the implementer. The tell is speed — an owner who answers question six in under ten seconds usually has not answered it at all.

When two answers fight, in this order

1 · Transfer wins. If there is a transfer with a date, it wins over any growth or lifestyle objective.
2 · The earlier date wins. Two destinations, two dates, the nearer one sets the targets.
3 · Cheque wins income. Where both are stated and neither of the above resolves it, money at the end wins over money along the way — because it is the harder programme, and a practice built for it can still pay well on the way.

Which one applies, and how to tell

What each of the four actually means, how to work out which one fits, and what to do with the one they almost picked.

Owner practice

The owner intends to keep producing at three or more clinical days a week through the horizon, and no transfer of ownership is planned inside sixty months. Success is measured in owner income and schedule control. How to tell: the owner answers four or more to question four, and answers "along the way" to question five.

Independent practice

The owner intends to keep ownership but reduce to three or fewer clinical days, and no transfer is planned inside thirty-six months. Success is measured by whether production and operations continue in the owner's absence. How to tell: two to three days at question four, no transfer at question two, and the owner can name at least one thing they currently do that someone else should.

Dental group

The owner intends two or more locations under common management inside the horizon. Success is measured as same-store performance plus new-store contribution, reported separately and never blended. How to tell: expansion language appears unprompted at question one, and the owner can name a funding source. An owner who wants a group and has no capital plan has an ambition, not a destination.

Sale

The owner intends to transfer majority ownership inside thirty-six months. Success is measured by normalised EBITDA and by how little of production depends on the owner personally. How to tell: a transfer intent at question two and a date inside thirty-six months at question seven. Internal succession to an associate is recorded as sale, internal — same targets, plus a financeability overlay, because the associate has to service acquisition debt out of what they produce.

What this does to your associates

Every destination except owner practice narrows what an associate controls, and an associate producing a third of the practice has real bargaining power. This is what usually goes wrong in months two and three. It is not resistance — the associate and the owner now want different things.

Sale — the sharpest conflict. Normalised earnings means pressure on lab spend, chair-time density and scheduling, and the associate carries the cost while the owner collects the proceeds. Ask before you start: does the associate have equity, an earn-out participation, or a retention payment at close? If none of the three, expect resistance and treat it as rational.
Dental group — standardised clinical protocols. Acquired and employed doctors will tolerate centralised billing and scheduling. They will not tolerate dictated clinical decisions. Draw that line explicitly and in writing, or attrition draws it for you.
Independent practice — usually improves the associate’s position, because the owner is handing over production and presentation. Say so out loud; it is the one destination where the associate is a beneficiary.
Owner practice — the quiet conflict. An owner maximising their own income and schedule has structurally less to give an associate, and associates in this setting churn at eighteen to thirty-six months. If the practice intends to keep one, the comp model has to be worth staying for.

The rule: if an associate produces more than a quarter of the practice, the owner tells them the destination in their own words within thirty days of the first session. Not the implementer, not a memo. An associate who learns the practice is being prepared for sale by noticing the questions being asked is an associate who is already interviewing elsewhere.

Secondary objectives

Recorded on the practice plan, in the owner's words.
They change nothing — not targets, not weights, not the calculation.
Single use: when two best practices look equally worth doing, pick the one that also serves the secondary.
Tell the owner this explicitly at the interview. An implementer who lets a secondary quietly influence targets has reintroduced the problem the fork exists to solve.

If they change their mind later

Only at a quarterly reset, never between them.
Requires the full stress test again, not just a statement.
On change: targets and weights reset immediately, and any priority already under way that no longer serves the new destination is stopped rather than finished.
Record the change and the date. A practice that changes destination twice in a year has a decision problem rather than an operating problem, and that is what you report.

Twenty statements

These moved. They live on the Diagnostic tab now.

Scoring moved to the Diagnostic tab.

What every number means

One definition each, written down, and do not change them halfway through the year. Change a definition and the chart moves — it looks like the practice improved when nothing happened.

New patients

Unique patients completing a comprehensive or periodic exam, first visit ever at this practice. Counted at the completed appointment, not at scheduling. Excludes reactivated lapsed patients — track those separately, they behave differently and inflate the number if merged.

Gross production

Treatment completed, valued at full office fee, before any adjustment. By provider, doctor and hygiene separated. Scheduled production is a different number. Never let the two share a chart.

Adjustments

Three separate lines, never one: contractual insurance write-offs · courtesy and professional discounts · membership plan discounts. Collapsed into a single figure, you cannot tell a fee-schedule problem from a discounting habit.

Net production

Gross production minus adjustments. The number that can actually be collected, and the one that belongs on the Snap.

Hygiene share of production

Hygiene gross production ÷ total gross production, same period. Hygiene means the hygiene columns, not prophy codes performed by a doctor. A practice below the low twenties is usually short of hygiene hours rather than short of patients. Check the schedule before you blame reappointment.

Production per chair-hour

Gross production ÷ clinical hours available, tracked separately for doctor and hygiene columns. The only production number that compares across practices, days and provider counts. A four-op practice and an eight-op practice are not comparable on monthly production and are comparable on this.

Collections percentage

Collections ÷ net production, same period, on a rolling three-month basis. Single-month figures are dominated by insurance payment timing and will mislead you in both directions.

Case acceptance

Dollars of treatment accepted ÷ dollars presented, same period. Excludes hygiene and recare. Excludes treatment diagnosed but not presented. The harder version counts only treatment scheduled within 30 days. It scores 10–20 points lower. Pick one, write down which, and say so every time the number is quoted.

Hygiene reappointment

Completed hygiene visits where the patient left with their next recare visit on the schedule ÷ completed hygiene visits. Measured at the point of departure. Backfilling at month end produces a flattering number and hides the handoff problem entirely.

Unscheduled treatment

Diagnosed, presented, not scheduled, diagnosed within the last twelve months. Without the rolling window the number inflates forever and the team stops believing it.

Open chair time

Unfilled available clinical hours ÷ available clinical hours, doctor and hygiene tracked separately. A blended figure hides which department has the problem, and the two have completely different fixes.

Broken appointments

Cancellations inside 24 hours plus no-shows ÷ total scheduled appointments. Practices that exclude "rescheduled on the call" understate this badly. Include them unless the patient was seen that week.

AR over 90

Balance aged 90+ days ÷ total AR, reported as a percentage. The dollar figure alone tells you nothing in a growing practice — it rises with volume even as collection improves.

Overhead

All operating expense excluding owner-doctor compensation ÷ collections, grouped as staff and benefits, clinical supplies, lab, facility, marketing, admin. Excluding owner compensation is the only way the number is comparable to anything outside this practice.

Owner income

Salary plus distributions plus personal expense run through the practice. One number, stated honestly. Most owners cannot say it without looking it up, and the looking-up tells you something on its own.

Normalised EBITDA

Earnings after replacing owner compensation with a market rate for a doctor producing what the owner produces, and after removing personal expense. This is the number a buyer prices. It is usually far below what the owner thinks the practice earns, and showing it early prevents a painful surprise later.

Owner share of production

Owner gross production ÷ total practice gross production. The clearest sign of how much of this business would survive the owner leaving. Three of the four destinations depend on it falling.

Active patients

Patients seen within the last eighteen months. Twelve-month and twenty-four-month definitions are both in common use and produce wildly different counts. Buyers usually use eighteen.

Where every number comes from

Which report each number comes out of, how often someone pulls it, and whose job that is.

Metric	Source	Frequency	Owner	Target method
New patients	PMS, completed exam codes	Weekly	Front office	Destination
Case acceptance	PMS, presented vs accepted	Weekly	Treatment coordinator	Destination
Hygiene reappointment	PMS, at departure	Weekly	Hygiene lead	Fixed 90%
Broken appointments	PMS, cancel and no-show	Weekly	Scheduling	Fixed 5%
Open chair time	PMS, schedule utilisation	Weekly	Scheduling	Destination
Gross production	PMS, by provider	Monthly	Office manager	Destination
Adjustments	PMS, three types split	Monthly	Claims owner	Destination
Hygiene share of production	PMS, production by column	Monthly	Office manager	Destination
Production per chair-hour	PMS production ÷ scheduled clinical hours	Monthly	Office manager	Derived per practice
Net production	Calculated	Monthly	—	Derived
Collections % of net	Calculated, 3-mo rolling	Monthly	Office manager	Fixed 98%
AR over 90	PMS aging	Monthly	Claims owner	Fixed 15%
Overhead ex-owner	Bookkeeping, recast	Monthly	Owner with bookkeeper	Destination
Owner income	Payroll plus distributions	Monthly	Owner	Owner-set
Unscheduled treatment	PMS, 12-mo window	Monthly	Treatment coordinator	Fixed
Owner share of production	PMS, by provider	Monthly	Office manager	Destination
Team turnover	Payroll, rolling 12 mo	Quarterly	Office manager	Fixed 15%
Vacant seats	Seat list	Quarterly	Owner	Fixed 0

One producer per number. Two people producing the same metric is how two versions of the truth appear. If the named owner leaves, reassign before the next Snap, not after.

Every target and every weight

Every number the tool does its maths on, laid out so you can argue with it. Do that before you put this in a practice, not after.

Proposed · not agreed

The numbers the tool runs its maths on. They were worked out by reasoning rather than taken from data, so argue with them before you use them in a practice. A band is one unit of concern for that particular number, which is why a metric with a 5% target is not treated the same as one with a 90% target.

Targets by destination

Metric	Unit	Better	Band	Owner	Independent	Group	Sale
New patients	count	higher	4	22	26	34	30
Case acceptance	%	higher	5	68	65	65	65
Hygiene reappointment	%	higher	5	90	90	90	90
Broken appointments	%	lower	2	5	5	5	5
Open chair time	%	lower	2	4	5	6	5
Collected, weekly	$	higher	2500	30,000	32,000	35,000	34,000
Gross production, monthly	$	higher	9000	140,000	148,000	160,000	155,000
Hygiene share of production	%	higher	3	28	28	28	28
Production per chair-hour	$	higher	30	340	340	340	340
Adjustments	%	lower	2	20	21	24	22
Collections % of net	%	higher	2	98	98	98	98
AR over 90 days	%	lower	4	15	15	15	15
Overhead ex-owner	%	lower	2	58	60	62	60
Unscheduled treatment	$	lower	30000	120,000	120,000	120,000	120,000
Owner share of production	%	lower	6	70	55	45	50
Owner clinical days	days	lower	1	4	3	2	3
Team turnover, 12 mo	%	lower	8	15	15	15	15
Vacant seats	count	lower	1	0	0	0	0
Owner income, monthly	$	higher	3000	24,000	24,000	24,000	24,000

Weights by destination

Metric	Owner	Independent	Group	Sale
New patients	0.40	0.70	1.00	0.90
Case acceptance	1.00	1.00	0.90	1.00
Hygiene reappointment	1.00	1.00	1.00	1.00
Broken appointments	0.60	0.60	0.60	0.60
Open chair time	0.90	0.70	0.70	0.70
Adjustments	0.80	0.80	0.80	0.90
Collections % of net	0.95	0.95	0.95	1.00
AR over 90 days	0.70	0.70	0.70	0.70
Overhead ex-owner	0.90	0.90	0.80	1.00
Unscheduled treatment	0.50	0.50	0.50	0.50
Owner share of production	0.20	1.00	0.90	1.00
Owner clinical days	0.40	0.80	0.70	0.40
Team turnover	0.60	0.60	0.60	0.60
Vacant seats	0.55	0.55	0.55	0.55
Gross production	0 · outcome	0 · outcome	0 · outcome	0 · outcome
Collected	0 · outcome	0 · outcome	0 · outcome	0 · outcome
Owner income	0 · outcome	0 · outcome	0 · outcome	0 · outcome

What the weights do

Weight decides how loudly a metric may shout, not whether it is tracked. Everything is tracked; only some things can become the one thing to fix.
Zero means never flagged. The three outcomes carry zero under every destination. That is not a setting anyone tunes. It is the rule, and it does not move.
The signature differences. New patients runs 0.40 for an owner practice and 1.00 for a group. Owner share of production runs 0.20 for an owner practice and 1.00 for independent or sale. Overhead runs 1.00 for a sale and 0.80 for a group, because a group carrying unabsorbed capacity should not be scolded for it.
Best practices are not weighted. The six under each area are things a well-run practice does. Nothing ranks them, and nothing scores them against the destination — the owner picks which one becomes a priority.

How the one thing gets picked

How the maths works, when a number counts as off, how to trace a bad one back to its cause, and when to escalate.

How it is worked out

Deviation, in bands. Higher-is-better: (target − actual) ÷ band. Lower-is-better: (actual − target) ÷ band. A negative result means better than target.
Score = deviation in bands × weight. Computed only where deviation is positive and weight is above zero.
The one thing to fix = the highest attention score across all metrics, weekly and monthly together.
Ties go to the metric with the shorter feedback loop — a weekly number beats a monthly one, because the owner can move it before the next Snap.

When a number counts as off

Green — deviation of zero or less. At or better than target.
Amber — deviation above zero up to one band.
Red — deviation above one band.
Severe — deviation above three bands. The variance bar clamps at three; the number stays uncapped in the calculation, so a severe metric will win the fix and should.
Not yet in the tool: the Snap currently renders a binary on-target / off-target state. The three-band status above is the specification.

Which way it is heading

A metric that has improved for two consecutive periods is demoted one status band for display, but not in the calculation — the practice should see that it is working without the maths losing sight of the gap.
A metric that has deteriorated for three consecutive periods is promoted one band, whatever its absolute level. Near the edges, which way it is moving matters more than where it sits.
Not yet implemented. The tool holds a prior value for every metric but does not yet hold a series.

Working back to the cause

Outcomes never get flagged. When one moves, work back through the chain in this order and report the first link that is off target.

Collections down → adjustments → collections % of net → AR over 90 → gross production. Four links; the answer is almost never the last one.
Gross production down → open chair time → broken appointments → case acceptance → new patients. Check capacity before demand. Most practices do it the other way round and buy marketing they did not need.
Owner income down → overhead by category → collections % → adjustments → gross production.
Hygiene production down → hygiene reappointment → open chair time in the hygiene column → active patient count.

When to escalate

Any metric red for three consecutive weeks has to be considered at the next quarterly reset. It does not automatically become a priority — it must be argued against the other candidates.
Any metric red for two consecutive quarters goes back to the destination itself: either the target is wrong for this practice or the destination is not reachable on this timeline. Both are conversations, not adjustments.
Vacant seats above zero for a full quarter goes on the list regardless of weight. An unfilled seat is a decision being made by default.

What we leave alone on purpose

Any metric whose weight is zero for the current destination is tracked, displayed, and never flagged. Tell the owner which ones these are at the targets session — an owner practice that is not being nagged about new patients should know that is a choice, not an oversight.
Metrics with fewer than four weeks of history are displayed without a status. Do not flag a number you cannot yet trust.

What is on the Practice Snap

What is on the page, what the page works out for itself, and who updates it on which day.

What is on the page, top to bottom

Attention block — one thing to fix (computed), one decision to make (owner input), one person to help (owner input).
This week — six rows: new patients, case acceptance, hygiene reappointment, broken appointments, open chair time, collected. Columns: metric, target, this week, prior, variance bar.
This month — seven rows in chain order: gross production, adjustments, net production, collections % of net, AR over 90, overhead, owner income.
This quarter — three priorities, each with owner and a red / amber / green marker.

What the page works out for itself

Net production = gross production × (1 − adjustments %). Both current and prior.
Variance bar = deviation in bands, clamped to ±3, drawn left of centre for better than target and right for worse.
The fix, its evidence sentence and its first action.
Every target, read from the destination.

Weekly workflow

Friday, office manager, fifteen minutes. Pull the six weekly numbers and update the month-to-date figures. Prior-period values shift automatically.
Monday, owner, five minutes. Read the attention block, write the decision and the person, check the three priorities.
Monday, weekly meeting. The Snap is the agenda’s first five minutes. Read them out. Do not discuss them yet.
If the numbers are not in by Friday, the meeting still happens and the Snap shows last week. Never delay the routine for data.

The five-minute test. If reading the Snap takes longer than five minutes, something has been added that should not have been. Thirteen numbers and three priorities is the ceiling, not a starting point.

The four meetings, minute by minute

The agendas do not change. These meetings usually die by drifting: the huddle turns into case review, the weekly turns into everyone reporting status, the monthly turns into reading the P&L out loud.

Daily huddle — 10 minutes, standing, before the first patient

0–2

Yesterday: produced, collected, and anyone who left without their next appointment.

2–5

Today's schedule by column, production goal, and the gap. Where the open time is.

5–8

Patients on today's schedule with unscheduled treatment — by name, procedure and dollars. Patients on today's schedule who are due or overdue for recare.

8–9

Accounts needing a financial conversation today.

9–10

One clinical note per complex case. Anything the team needs from the doctor.

Owner note: this is not on the 90-minute clock. It runs inside the clinical day and it has to run whether the owner is in the building or not.

Weekly leadership — 45 minutes, same day and time

0–5

The six week numbers against target. Reading them is not discussion.

5–15

Priority status, red / yellow / green only. If this takes more than ten minutes, the priorities are written wrong.

15–40

Issues: list them, prioritise them, solve the top three. Solve means a decision, an owner and a date — not a discussion.

40–45

Who needs to know what, and who tells them.

How it drifts: into reporting. If nobody left the room with an action, it was a status meeting and it will be cancelled within a quarter.

Monthly financial — 60 minutes, second week of the month

0–15

P&L against budget. Variances over 5% only — everything else gets skipped.

15–25

Overhead by category as a percentage of collections, against last quarter.

25–40

The chain: production, adjustments by type, net, collections percentage. Then AR aging.

40–50

Cash position, what is due, what is committed.

50–60

One action, with an owner and a date.

Most single-location practices have never held this meeting. It is usually the most valuable hour in the whole system, and the one that makes the owner take the rest of it seriously.

Quarterly reset — half a day, first week of the quarter

0–45

Score last quarter. Priorities done or not done, no partial credit. Numbers against target, quarter over quarter.

45–90

Revisit the destination. Still true, still the same date.

90–120

Re-score the twenty statements. Report score movement and number movement separately.

120–150

Re-rank the priorities against the destination and choose three.

150–180

Make one decision that has been deferred all quarter.

What to actually say in each meeting

Who says what, in which order, and the one or two things that get written down afterwards.

Weekly leadership meeting

Owner opens: "Six numbers. I am reading, not discussing."

Reading the numbers is not the meeting. If discussion starts here, the issues block loses its time and nothing gets solved.

Owner: "Priorities. Red, amber or green. One word each."

If anyone explains, interrupt. Explanation belongs in the issues block. Ten minutes maximum — longer means the priorities are written wrong.

Owner: "Issues. What is in the way this week?" Then: "Which of these actually matters most?" Then: "What are we deciding, who owns it, by when?"

Three questions, in that order. List, prioritise, solve. Solving it means a decision with a name and a date on it, not a discussion that stops when the clock runs out.

Owner closes: "Who needs to know what, and who is telling them?"

Five minutes. This is the step most often dropped, and dropping it is why the team hears about decisions secondhand.

Written down: the three decisions with a name and a date each, and where the priorities stand. One page, same place every week.

Monthly financial meeting

Implementer: "Variances over five per cent only. Everything else we skip."

The failure mode is reading the whole P&L aloud. Fifteen minutes, exceptions only.

Implementer: "Overhead by category, as a percentage, against last quarter. Not in dollars."

In a growing practice the dollars always rise. The percentage is the only honest number, and owners consistently reach for the dollars.

Implementer: "Walk me down the chain. Production, adjustments, net, collections. Where does it leak?"

Let the owner find it. An owner who traces their own leak remembers it; one who is shown it forgets by Thursday.

Implementer closes: "One action. Who and when."

One. A financial meeting that produces five actions produces none.

Written down: the variances you went through, the one action with a name and a date, and the overhead table for the month.

Daily huddle

Office manager opens: "Yesterday we produced X and collected Y. Z patients left without their next appointment."

That third number is the one that changes behaviour. Say it every day even when it is zero — especially when it is zero.

Office manager: "Today’s goal is X, we are scheduled at Y, the gap is Z, and the open time is here."

Office manager: "On today’s schedule with unscheduled treatment: name, procedure, dollars." Then recare due. Then accounts needing a conversation.

By name. A list of counts changes nothing; a patient’s name in the room changes what happens at the chair.

Doctor closes: "Anything you need from me today?"

The whole huddle is ten minutes, standing. If people sit down it becomes a meeting and it will be cancelled.

Written down: nothing. The huddle sheet gets printed, used and thrown away. Write it up and it turns into paperwork, and then it stops happening.

What a good priority looks like

If you cannot say done or not done at the end of the quarter, it was never a priority.

One named person. Not a department, not two people, not "the team."
A date. The last day of the quarter unless there is a reason for an earlier one.
Observable done. Written so that a person who was not in the room could walk in and say whether it happened.
First step, three mid-steps, final step — each with a by-date.
Resources named — money, hours, software, or a decision from someone else.
Fit to the destination. If it would rank low for what this practice is building, it does not belong in the three, however appealing it is.

Not a priority

Improve collections.

Get the team more engaged.

Work on case acceptance.

Document our processes.

A priority

Reduce AR over 90 from 19% to 12% by 31 March, by working a standing Tuesday hour with a named owner.

Hire and start an associate by 15 November so the owner drops Friday clinical from 1 January.

Move hygiene reappointment from 79% to 90% by 31 March by booking the next visit in the operatory, measured at departure.

Writing a priority that can be scored

What counts as one, what does not, and how you settle it at the end of the quarter.

Rules that do not bend

Three maximum. Not four. A practice carrying four finishes two, and the two it finishes are the easy ones.
One named person each, and no person holds two. At Foundation scale: one to the owner, one to the office manager, one to the hygiene lead.
It has to close a gap you can point at — a number that is off, or a statement that scored low. A priority that fixes nothing you have measured is a preference.
Each maps to exactly one area. If it spans two, it is two priorities or it is badly written.
A date — the last day of the quarter unless there is a reason for earlier.

Required fields

Owner, by name · due date · title · what done looks like, in observable terms · why it matters this quarter · first step, three mid-steps, final step, each dated · resources needed, including any decision required from someone else.

What does not count

Owned by a department, a pair, or "the team".
Written as a direction rather than a destination — improve, increase, work on, focus on.
No number and no observable end state.
Dependent on someone outside the practice with no commitment from them.
Something that is really a task. If it takes under a week, it belongs on the weekly issues list.
Something that is really a year-long project. If it cannot finish inside the quarter, cut the first quarter of it out and make that the priority.

Deciding whether it happened

Binary. Done or not done. No partial credit, ever. The first quarter a seventy per cent priority is marked green is the quarter the system stops meaning anything.
Judged against the written "done", not against effort. A priority that was well worked and did not land is not done; say so, and say the work was good.
Scored at the reset by the implementer, not by the owner. The owner argues; the implementer records.
A priority carried into a second quarter counts as not done in the first, and is rewritten before it is carried.

Disqualified

Improve collections.

Get the team more engaged.

Work on case acceptance.

Document our processes.

Grow new patients.

Qualifies

Reduce AR over 90 from 19% to 12% by 31 March, working a standing Tuesday hour. Owner: Marisol.

Hire and start an associate by 15 November so the owner drops Friday clinical from 1 January. Owner: Dr. Reyes.

Move hygiene reappointment from 79% to 90% by 31 March by booking the next visit in the operatory, measured at departure. Owner: Kara.

Writing a process down properly

What has to be on the page before a process counts as written down. Anything less is just a conversation, and conversations walk out the door with the person who had them.

Name and trigger — what starts it, and what time or event.
Owner — one person, by name, plus the backup.
Steps in order, with the system or screen each one happens in.
Timing — how often, how long, and the deadline for each step.
What done looks like — the observable end state.
What to do when it breaks — the escalation, and to whom.
Where it lives — one location, findable by someone hired next month.
Review date — when it gets checked against what people actually do.

The test: hand it to someone who has never done the task and watch them run it once. Every place they hesitate is a place the card is wrong.

Running the quarterly reset

The order to run it in, the questions to ask, and what actually changes at the end.

How to run the half day

0–45 · Score. "Priority one. Done or not done." Three times. No explanation until all three are scored.

Implementer scores, owner argues, implementer records. Scoring before discussing prevents the quarter being relitigated into a pass.

45–75 · The numbers. Each metric this quarter against last quarter and against target. Direction first, position second.

75–120 · The destination. "Is it still what you are building? Is it still the same date?"

If either answer changed, run the stress test again before accepting it. A changed destination is legitimate; a drifted one is not.

120–150 · Re-score the twenty. Against day one, with the same evidence standard.

Report score movement and number movement separately, and say which is soft and which is hard. Scores can move because people feel better; numbers cannot.

150–170 · Re-rank and choose three. Written to the standard, owners named, dates set, before anyone leaves.

170–180 · One deferred decision. "What have you been carrying all quarter?" Make it in the room.

What resets

Targets — only if the destination changed. Otherwise they hold for the year.
Weights — only if the destination changed.
Priorities — always. Three new, or two carried and rewritten plus one new.
Seat list — always reviewed, risk marks re-set independently by the office manager.
The metric dictionary — never mid-year. Changes take effect at the fiscal year, and are annotated on the trend line.

The ninety days, at a glance

Ninety days from the first call to a practice running itself, then it repeats every quarter. Nine sessions, about fourteen hours of the owner’s time in total.

Before

Fit conversation — is this practice at the stage this system fits

45 min · owner

Day 1

Declare — what are you building, and by when

90 min · owner

Days 2–10

Baseline — the data pull, the twenty statements, the recast P&L

2 hrs · owner, OM

Day 12

Targets — the destination sets the numbers

60 min · owner

Day 14

Priorities — three, with owners and dates

90 min · leadership

Days 15–30

Start the routine — huddle and weekly, nothing else

daily 10 min, weekly 45

Days 31–60

Fix the leak — first financial meeting, two route cards

60 min + build

Days 61–90

Make it stick — stop-doing list, quarterly reset, re-score

half day · leadership

Day 90

The test — the huddle happens when the owner is out

pass or restart

What happens, step by step

What happens at each step, who is in the room, what you leave with, and the particular way each step goes wrong.

Before

Fit conversation

One call, no materials. Three questions: what do you think you are building, what is actually bothering you week to week, and what would have to be true in a year for this to have been worth it. Listen for whether the answers to one and two point the same direction — when they do not, the gap between them is what you have been hired to fix.

Qualify out if the practice collects under roughly $700K (the overhead of the system exceeds what it can move), if it already sits inside a group with imposed reporting, or if the owner will not name a date. A destination without a date is a preference, and preferences do not change targets.

In the room: owner only — not the office manager, not yet · Time: 45 minutes · Out: go or no, and the destination they are leaning toward

How it fails: selling into a practice whose real problem is clinical, personal, or partnership conflict. None of those are operating-system problems and the rollout will stall around them.

Day 1

Declare

The first and most important session. Not values, not why the practice exists, not a mission statement. One question: what are you building, and by when.

0–20

The four destinations, read aloud. Owner practice · independent practice · dental group · sale. The owner argues with all four. Let them — the arguing is how they find out which one they actually mean.

20–35

The tie-break. If there is a transfer with a date, the transfer wins. Someone growing in order to sell declares sale, because the buyer's arithmetic sets the targets, not the growth story.

35–45

The date. A month and a year, written down. Not a horizon, not "two to three years."

45–60

Clinical model in one sentence — comprehensive care or insurance volume. These need opposite scheduling templates, case presentation, hiring and fee strategy. A practice that refuses to choose runs both badly.

60–90

The seats. Every function in the practice listed, one name against each. Vacancies stay visibly vacant. People holding two seats stay visibly doubled. Both are findings.

In the room: owner · Time: 90 minutes · Out: destination with a date, clinical model, seat list with risk marks

How it fails: the session drifts into values and culture. The moment Purpose is treated as "why we exist," the operating consequences disappear and you get a nice conversation that changes nothing. Hold the question to what is being built.

Days 2–10

Baseline

Three things run at once: the data pull, the twenty statements, and the recast P&L. The office manager runs the first, the owner does the second, you do the third.

Pull

The report list below, put on the definitions below. Expect the first pull to be wrong — reports default to different date ranges and different provider filters, and the numbers will not tie. Reconcile before anything is entered.

Score

Twenty statements, five per area, five-point scale. Owner scores alone first. If you are running multi-rater, the office manager and one clinical team member score the same twenty independently and you report the divergence.

Recast

The P&L restated: owner compensation normalised to a market rate for a doctor producing what they produce, personal expenses separated out, overhead grouped into staff, clinical supplies, lab, facility, marketing, admin. Twenty-four months, monthly.

In the room: owner and office manager for the kickoff, then asynchronous · Time: 2 hours of meeting, 4–6 hours of pull · Out: scored baseline, twenty-four months of history on one set of definitions, recast P&L

How it fails: accepting the practice's own numbers without putting them on one definition. If case acceptance was computed one way last year and another way this year, every trend line you build is measuring definition drift. This is the least glamorous phase and the one that determines whether anything after it is true.

Day 12

Targets

The destination sets the targets and the weights. The Practice Snap gets populated with the practice's own numbers, and for most owners this is the session where it becomes real — because it is the first time they have seen their own P&L with their compensation normalised out of it.

0–20

The recast P&L. Show it before the targets. Overhead by category, EBITDA after a market-rate doctor. Sit with the silence.

20–40

The chain. Production → adjustments → net production → collections, with their own twelve months in it. Find the leak together rather than presenting it.

40–55

Targets and weights, set by the destination. Explain which numbers the system will now leave alone and why — an owner-practice destination means new-patient growth stops being nagged about.

55–60

Who updates the Snap each week, and when. Usually the office manager, Friday afternoon.

In the room: owner · Time: 60 minutes · Out: a live Practice Snap with real numbers and a named person who updates it

How it fails: negotiating the targets down to what feels comfortable. The targets come from the destination, not from the current performance. If the owner wants different targets, the honest move is to change the destination.

Day 14

Priorities

Each area lists six best practices. Three get chosen. Never four — a practice carrying four finishes two, and the two it finishes are the easy ones.

0–30

Read the six best practices in each area alongside the diagnostic scores. The ones the practice is worst at are the candidates.

30–60

Choose three. One to the owner, one to the office manager, one to the hygiene lead. One person owns it. Not a department, not a pair.

60–90

Write each one to the priority standard below: observable done, first step, three mid-steps, final step, each with a date, plus resources needed.

In the room: owner, office manager, hygiene lead · Time: 90 minutes · Out: three written priorities with named owners and dates

How it fails: priorities written as aspirations. "Improve collections" is not a priority. "Reduce AR over 90 from 19% to 12% by 31 March by working a standing Tuesday hour" is.

Days 15–30

Start the routine

Huddle starts, every clinical day. Weekly leadership meeting starts. Nothing else changes for two weeks.

The only goal for this fortnight is that both meetings happen every single time they are supposed to. Adding a third thing now is how this usually dies. The owner feels momentum, takes on a project, misses a huddle, and the habit never sets.

Week 1

You attend every huddle and the weekly meeting. You are not running them — the office manager runs the huddle, the owner runs the weekly. You are watching for whether the sheet is printed and whether anyone speaks who is not a doctor.

Week 2

You attend two huddles and the weekly. Correct the agenda drift, not the content. Huddles slide into clinical case review; weeklies slide into status reporting.

In the room: whole team for huddle, leadership for weekly · Time: 10 minutes daily, 45 weekly · Out: a routine that holds without you

How it fails: the huddle becomes a schedule read-out. If nobody names a patient with unscheduled treatment or an account that needs a conversation, it has become a ritual and it will be cancelled within a month.

Days 31–60

Fix the leak

First monthly financial meeting, then the two process fixes that pay back fastest in almost every practice.

Fix 1

Recare reactivation. A written sequence with a trigger, an owner, a schedule, and who to tell when it stalls. Runs without the owner prompting it. Most practices have a list and no sequence.

Fix 2

Claims on an aging cadence. A defined day and hour, a defined order of work, a point at which a claim is chased harder or written off. Most practices work claims when someone notices.

Both

Written as route cards, not taught verbally. A process that exists only in a conversation is not a process, and it leaves when the person does.

In the room: owner and office manager for the financial meeting, process owners for the builds · Time: 60 minutes plus build time · Out: two route cards in use, first month closed properly

How it fails: building the route card and never watching it run. Sit with the person the first two times they use it. The gap between the card and what they actually do is where the card is wrong.

Days 61–90

Make it stick

Second monthly financial meeting. The stop-doing list. The quarterly reset.

Stop

The owner writes five things they personally still do that someone else could. One gets handed off this month, with a named recipient and a date. Not five. One that actually happens beats five that are listed.

Reset

Score the quarter. Done or not done, no partial credit. A priority at 80% is not done, and calling it done is how the next quarter's three become five.

Re-score

The twenty statements again, against day one. Movement in the scores is soft evidence; movement in the numbers is hard evidence. Report both, and say which is which.

Set

Three new priorities, re-ranked against the destination. The ranking will have changed, because the numbers changed.

In the room: leadership · Time: half a day · Out: a scored quarter, one real handoff, three new priorities

How it fails: partial credit. The first quarter where a 70% priority gets marked green is the quarter the system stops meaning anything.

Day 90

The test

The huddle happens on a day the owner is not in the building, with nobody reminding anyone. If it does not, the system is still the owner and the scores did not mean anything.

This is the only pass-fail in the programme, and it is deliberately not a number. Every destination except the owner practice depends on the business functioning without the owner, so this test predicts more about whether they get there than any number on the page.

Out: evidence, or an honest restart

Ongoing

Every quarter, start again

Every quarter the destination comes back on the table before anything else. Are we still building what we said we were building? If the answer changed, the targets change, the weights change, and the priority ranking changes with them. If it did not, score the quarter and set three more.

0–45

Score last quarter. Priorities done or not done. The numbers against target, quarter over quarter.

45–90

Revisit the destination. Still true? Still the same date? A changed destination is not a failure — an owner who decides in month seven that they are selling rather than holding has just saved themselves two years of building the wrong thing.

90–150

Re-rank and choose three. Written to the standard, owners named, dates set.

150–180

One deferred decision that has been carried all quarter gets made in the room.

Time: half a day, four times a year

What to ask them for

The full list of reports to request. Dentrix, Eaglesoft, Open Dental, Curve and Denticon all call them something different, but they all hold the same things.

From the practice management system

Production by provider by month, 24 months — gross, at full office fee, before adjustment. Doctor and hygiene columns separated.
Adjustments by type and by payer, 12 months — contractual write-offs, courtesy and professional discounts, membership plan discounts, as three separate lines.
Collections by month, 24 months — split insurance payments and patient payments.
AR aging summary — current, and twelve monthly snapshots if the system retains them. Buckets at 0–30, 31–60, 61–90, 90+.
New patient count by month, 24 months — counted on completed comprehensive or periodic exam, first visit ever. Reactivated lapsed patients pulled separately.
Treatment presented and accepted, 12 months — dollars, excluding hygiene and recare.
Unscheduled treatment plan report — filtered to treatment diagnosed within the last twelve months only.
Hygiene visits completed by month, and the pre-appointment or reappointment rate if the system reports one.
Broken appointment and cancellation report, 6 months — cancellations inside 24 hours and no-shows, against total scheduled.
Schedule utilisation by column, 3 months — available clinical hours against filled, doctor and hygiene separately.
Active patient count — on an eighteen-month definition — and the attrition trend.
Payer mix by production, and the PPO participation list with fee schedules and effective dates.

From the books and the office

P&L, 24 months, monthly — not annual, not quarterly. Seasonality in dental is real and an annual view hides it.
Balance sheet — current and prior year end.
Payroll register by role, with hours, and benefits loading.
Debt schedule — balances, rates, maturities, covenants, personal guarantees.
Provider agreements — associate compensation terms, restrictive covenants, notice periods, and whether any equity, earn-out participation or retention payment exists.
Associate production share — what proportion of practice production each non-owner doctor generates. Above a quarter and the destination conversation includes them.
Team roster — role, hire date, compensation, hours, and who reports to whom in practice rather than on paper.
Lease — term remaining, renewal options, and whether the owner owns the building.
Any reporting software they already pay for, and which definitions it uses. If the practice already runs one, the dictionary has to reconcile to it or the team will see two different truths.

Expect the first pull to be wrong. Reports default to different date ranges, different provider filters and different inclusion rules for adjustments. Production will not tie to collections, and new patient counts will differ between two reports in the same system. Reconcile before entering anything.

Week by week, and what to do when it goes wrong

What to do each week, and what to do when the data is missing, the owner stops turning up, or the team digs in.

Week 1

Day 1 — purpose interview, ninety minutes, owner only. Seat list drawn at the end of the same session while the owner is still thinking about people.
Day 2 — issue the data pull list to the office manager with a named due date of day 8. Walk them through the first three reports personally; do not email the list and hope.
Day 3 — owner scores the twenty statements alone. Implementer does not sit in.
Day 4–5 — implementer scores the same twenty from evidence, in the practice, watching. This is the day that makes the assessment real.

Week 2

Day 8–10 — reconcile the pull. Expect it not to tie. Put everything on the same definitions before entering a single figure.
Day 10–11 — recast the P&L: owner compensation to market, personal expense removed, overhead grouped into six categories, twenty-four months monthly.
Day 12 — targets session. Show the recast P&L first, the chain second, the targets third. In that order, without exception.
Day 14 — priorities session with the leadership trio. Three chosen and written to the standard before anyone leaves the room.

Weeks 3–4

Huddle starts day 15. Implementer attends every huddle in week 3 and two in week 4.
Weekly leadership meeting starts day 15. Implementer attends both, running neither.
Correct agenda drift, not content. The huddle slides toward clinical case review; the weekly slides toward status reporting. Both within a fortnight, reliably.
Add nothing else. The only goal for these two weeks is that both meetings happen every single time.

Month 2

First monthly financial meeting, implementer running it.
Build the recare reactivation route card and the claims aging cadence card. Sit with the person the first two times each one runs.
Implementer attends one huddle a week and every weekly meeting.
Mid-month: check the Snap has been updated every Friday. If it has been missed twice, that is the conversation below, not a reminder.

Month 3

Second monthly financial meeting, owner running it, implementer silent.
Stop-doing list written; one item handed off with a named recipient and a date.
Quarterly reset, half a day, implementer running it and scoring.
Day 90 test: attend a huddle on a day the owner is out, unannounced.
Implementer attends the weekly meeting only. Stepping back is meant to happen.

When data is missing

Never impute. A blank is reported as a blank and the metric shows no status.
If a weekly number is unavailable for two consecutive weeks, find out whether the report exists, the person is unavailable, or the process does not happen. Three different problems, three different fixes.
If the practice management system genuinely cannot produce a number, substitute a manual count with a written definition and a named counter, and record the substitution in the definitions for that practice.
If more than four of the seventeen are unavailable at baseline, pause the rollout and fix reporting first. A system built on four-fifths of the numbers will produce confident wrong answers.

When the owner will not execute

First miss — name it in the room, privately, same week. "The Snap was not updated. What got in the way?" Listen; it is usually capacity, not resistance.
Second miss — reduce scope rather than repeat the ask. Drop to two priorities. A smaller system that runs beats a complete one that does not.
Third miss — stop the rollout and have the conversation about whether the destination is real. An owner who will not spend eighty-five minutes a week has told you the destination is not theirs.
Never take the work over. An implementer who updates the Snap for the practice has built a reporting service, and it ends when they leave.

When the team resists

Resistance is almost always one of four things: the change threatens someone’s standing, someone was not consulted about their own job, nobody explained the destination, or the destination genuinely costs them something. Diagnose which before responding — the fourth is not resistance, it is a correct reading of their own interests.
Standing — usually the long-tenured front-office person whose informal authority a written process removes. Give them ownership of the route card rather than subjecting them to it.
Not consulted — rebuild that process with them in the room. It will come out better anyway.
Not explained — the owner tells the team the destination, in their own words, at an all-hands. Not the implementer. This one is not delegable.
Genuinely costs them — almost always an associate, occasionally a long-tenured hygienist on a legacy comp arrangement. This one is not a coaching problem. It is an economics conversation for the owner, and the honest options are to change their deal, give them a share of the outcome, or accept that they may leave. Pretending it is an attitude problem is how a practice loses a producer six months before a sale.
If one person blocks after all four are addressed, that is a seat conversation for the owner, and it belongs on the Team tab, not in the meeting.

When to raise it

Two consecutive missed weekly meetings → conversation with the owner inside 48 hours.
Huddle held fewer than seven times in a fortnight → reset the huddle, do not add anything new.
Any metric red three consecutive weeks → must-consider item at the reset.
Any metric red two consecutive quarters → back to the destination itself.
Destination changed twice inside a year → stop rolling out and deal with how decisions get made. That is the problem you are actually looking at.

Checklists before each gate

Four points where you stop and check. Do not carry on until every box is ticked.

Before launch — end of week 2

Destination names one of the four with a month and a year.

Destination survived the stress test, and the owner stated it in one sentence.

Secondary objective recorded, and the owner has been told it changes nothing.

Seat list complete, every function listed, vacancies visible, risk marks set by the office manager not the owner.

All twenty statements scored by the owner and independently evidenced by the implementer.

At least thirteen of the seventeen metrics available and put on the same definitions.

Production reconciles to collections across the chain for the last three months.

P&L recast with owner compensation normalised to market, twenty-four months monthly.

Targets set from the destination, not negotiated down.

A named person owns updating the Snap, and knows the day and the time.

Three priorities written to the standard, each closing a gap you can point at.

Meeting times in everyone’s calendar as recurring appointments, not intentions.

Office manager has been told the real time cost of their part — about two and a half hours a week — and has agreed to it.

If an associate produces more than a quarter of the practice, the owner has a date to tell them the destination.

Before each weekly meeting

Snap updated by Friday, prior-period values shifted.

Any metric with no data is blank, not estimated.

Priority statuses set by their owners before the meeting, not in it.

Last week’s three decisions checked: done, or explicitly carried with a reason.

Before the quarterly reset

Twelve weeks of Snap history available.

Priorities scored by the implementer before the room opens.

Baseline re-scored with the same evidence standard as day one.

Metrics red for three or more consecutive weeks listed as must-consider items.

Seat list refreshed and risk marks re-set independently.

The deferred decision identified in advance, so it cannot be avoided on the day.

Before declaring the ninety days successful

The huddle happened on a day the owner was not in the building, unprompted.

Both meetings held at least eighty per cent of the times they were scheduled.

At least two of three priorities done, judged binary.

Two route cards written and observed running.

One item genuinely handed off the owner’s stop-doing list, with a named recipient.

The Snap updated by the practice, not the implementer, for at least the last six weeks.

At least one number moved in the direction the destination requires.

The whole-practice score improved, and the improvement is explainable by something that actually changed.

Failing the ninety-day gate is a legitimate outcome. Record it honestly and restart the rollout at week three. A rollout declared successful on a technicality means the next quarter is built on a floor that is not there.

What the practice walks away with

What the practice gets, what form it comes in, and when they get it.

What they get	Form	When	Who signs it off
Destination statement	One paragraph, on the practice plan	Day 1	Owner
Seat list	One page, with risk marks	Day 1	Owner
Baseline report	Scores by area, whole-practice score, and the gaps between raters	Day 12	Implementer
Recast P&L	24 months monthly, owner comp normalised	Day 12	Implementer and bookkeeper
Practice plan	One page	Day 14	Owner
Practice Snap	One page, live, updated weekly	Day 12, then weekly	Office manager
Three priorities	One page each, to the standard	Day 14	Named owners
Huddle sheet	Printable, daily	Day 15	Office manager
Meeting agendas	Four cards	Day 15	—
Route cards	Two minimum by day 60	Days 31–60	Process owners
Quarterly report	Scored quarter, re-scored baseline, next three	Day 90, then quarterly	Implementer

Not on this list: a slide deck, a written strategy document, or a forty-page report. If the practice receives something they will not touch again, it was produced for the implementer’s comfort, not theirs.

What’s settled and what isn’t

The parts that never move, the parts still worth arguing about, and the things nobody gets to change once they are in a practice.

Implementer Manual v1.0 · draft

Locked — do not change in the field. The four destinations and the transfer-wins rule · the four areas and their order · the outcome rule (outcomes carry zero weight) · the production chain · the maximum of three priorities a quarter · the 85-minute owner budget · the metric definitions.
Proposed, not agreed. Every target value, every weight and every band width in the matrices below. These were set by reasoning, not by data. They are internally consistent and they are not validated. They become the method the moment someone uses them, so agree them on purpose rather than by default.
Unverified. All benchmark ranges quoted anywhere in this manual are working knowledge, not sourced figures. Verify against ADA HPI, Dental Intelligence or portfolio data before anything is shown to a client.
An implementer may never change what a number means partway through, lower a target because the practice finds it uncomfortable, award partial credit on a priority, or add a fifth priority.

Twelve rules

These are the twelve rules everything else sits on. If something in the system cannot be traced back to one of them, it should not be there.

1 · Pick one, and put a date on it

Four destinations, one chosen, each carrying a month and a year. Owner practice · independent practice · dental group · sale. Tie-break: if there is a transfer with a date, the transfer wins. Someone growing in order to sell declares sale, because the buyer's arithmetic sets the targets. Without this rule the four categories collapse into three overlapping programmes.

2 · How much system the practice gets

The destination sets how much system the practice gets, because it sets how much non-clinical time the owner has. Owner practice, four clinical days, 85 minutes a week — the floor, and nothing may be added to it. Independent practice, same 85 minutes, different content. Sale, roughly two hours, adding normalised EBITDA and diligence prep running alongside. Dental group, three hours, adding per-location Snaps, enforced common definitions, function reviews by discipline, an accountability layer and thirteen-week cash. Specified here, not yet wired into the tool — the dropdown currently changes targets, weights and ranking, but every destination still receives the same seventeen metrics and four meetings.

3 · Where everything files

Patient, Process, Team, Money — in that order, clockwise from the top. The order is causal: the patient model decides the process, the process decides who you need, and money is what is left. Every number, priority, route card and question files under exactly one area, which is how the practice knows who owns a problem. Money sits last deliberately. It is nearest the outcome, and the system does not treat outcomes as peers of drivers.

4 · Production to cash

Gross production → adjustments → net production → collections → owner income. Five links, each measured separately, because a practice can grow production fifteen percent and collect nothing. No general-purpose operating system tracks this, which is why a single-number scoreboard tells a dentist almost nothing.

5 · How the one thing gets picked

For every metric: deviation measured in bands, multiplied by a weight. Highest product becomes the one thing to fix, and arrives with its evidence and a first action. Bands, not percentages — one band is one meaningful unit of concern for that particular number. Otherwise a metric with a 5% target screams every time it moves half a point while one sitting eleven points low looks calm.

6 · Never chase an outcome

Outcomes carry a weight of zero. Collections, production and owner income can never be flagged. It names the cause underneath instead. "Collections are down" is not something a person can act on at eight in the morning. "Hygiene reappointment is at 79% and the handoff is happening at the front desk" is.

7 · Six best practices an area

Each area carries six things a well-run practice does. They are not ranked and not weighted. The owner picks which one becomes a priority for the quarter, and the diagnostic says which of the six this practice is worst at. A best practice that cannot be tied to a number or a statement does not belong on the list.

8 · The one-to-five scale

Twenty statements, five per area, anchored at 1 not at all · 3 sometimes and inconsistently · 5 consistently and without the owner. Bands: under 40 no system · 40–65 partial and owner-dependent · 65–85 a system exists · over 85 it runs without the owner. Every item is behavioural and single-barrelled. "We have great systems and the team follows them" cannot be answered honestly; "our recare sequence runs without me prompting it" can.

9 · Seven seats

Five of these exist in any business: finance, operations, sales, marketing and people. Dentistry adds two: clinical (standard of care, protocols, who gets referred out) and hygiene (the recare engine, which carries about a third of production and deserves its own seat). Sales in a practice means presenting and closing the case. Marketing means new patients, reviews and referrals in. In a small practice one person holds several seats, and that is fine as long as it is written down. Sales and marketing are the two that are usually nobody’s. Both land on the owner by default, and both are where growth comes from. A dentist who bristles at the word sales is exactly the practice where that seat is empty.

10 · One name per seat

Every function in the practice, one named person, plus a departure-risk mark. Vacancies and doubled seats stay visible. The two findings fall straight out of the list: work with no owner lands on whoever is nearest, and one person holding two seats is a single point of failure that looks fine until they take a week off. Have the office manager set the risk marks independently — owners systematically underrate front-office risk.

10 · What makes a priority real

Adapted from MIT Sloan research on goal setting. Frequently reviewed — status every week, not at quarter end. Ambitious — difficult, not impossible. Specific — a metric and a date, scorable done or not done. The fourth test, transparency, is amended for dentistry. Provider production is personal compensation, so a public production board is a compensation disclosure. Transparent at department and practice level, private at individual provider level, individual numbers reviewed one to one.

11 · Eighty-five minutes

Snap five minutes, weekly leadership 45, monthly financial 60 amortised to fifteen, quarterly reset four hours amortised to twenty. About 85 minutes of owner management time. The daily huddle sits off that clock — it runs inside the clinical day and happens whether the owner is in the building or not, and that is what the day-90 test checks. Say the whole cost out loud, not just the owner’s. The office manager carries about two and a half hours a week: fifteen minutes on Friday updating the Snap, ten minutes a day on the huddle sheet, plus both meetings. The hygiene lead carries about an hour. A promise that is precise about one person and silent about another gets discovered in week three and resented.

12 · Every quarter, start again

Score, revisit the destination, re-rank, choose three, decide one deferred thing. The destination comes back on the table before anything else, every quarter. A changed destination is not a failure. An owner who decides in month seven that they are selling rather than holding has just saved two years of building the wrong thing.

How the tool is built

How the tool is put together, what the formulas do, and the pieces that still need building.

Tabs

Your DPS — destination display and the practice plan. Mostly owner-written.
Practice Snap — thirteen metrics, attention block, three priorities. Weekly input.
Patient · Process · Team · Money — area metrics, six best practices, five assessment statements each. Team also carries the seat list.
How we do it — this manual. Nothing to fill in; it is here to be read.

Data model

Destination — one of four keys, plus a date string.
Values — one key per metric, each holding [current, prior].
Scores — one key per assessment item, 0 to 5, zero meaning unanswered.
Seats — function, name, departure risk.
Priorities — text, owner, status.
Plan — free-text fields.
Whole state serialises to a single JSON file. Save and load are manual by design — no account, no server, no data leaving the practice.

Formulas

Target = destination override if present, else the metric’s base target.
Weight = destination override if present, else the metric’s base weight.
Deviation in bands = (target − actual) ÷ band for higher-is-better; (actual − target) ÷ band for lower-is-better.
Score = deviation × weight, computed where deviation > 0 and weight > 0.
Net production = gross production × (1 − adjustments ÷ 100).
Variance bar = deviation clamped to ±3, rendered as a proportion of half the track.
Area score = sum of its five item scores. Whole-practice score = all twenty added up.
Vacant seats = count of seats with an empty name, recomputed on every edit and fed to the calculation.
Best practices are a fixed list of six per area. No ranking, no weighting.

Data flow

Changing the destination re-reads every target and weight and recomputes the attention block. Nothing is cached.
Editing a metric anywhere updates it everywhere — the Snap and the area tabs read the same store.
Editing the seat list updates the vacancy count, the practice plan’s ownership panel and the calculation in the same pass.
Adopting a priority from an area tab writes it into the first empty priority slot and marks the candidate as on plan.

Still to build

The complexity budget. All four destinations currently receive the same seventeen metrics and four meetings. Group and sale should receive more.
Three-band status. The Snap renders binary on-target / off-target; green, amber, red and severe are specified above.
Trend rules. The store holds a prior value, not a series, so the two-period improvement and three-period deterioration rules cannot run.
Raising a number automatically. Three-consecutive-weeks-red requires history the tool does not retain.
The driver trace, as an on-screen path rather than a rule in this manual.
Multi-rater assessment and divergence reporting.
Printable huddle sheet and priority planner as standalone pages.
The date field is display-only. Nothing computes from the destination date — a sale fourteen months out should weight differently from one thirty-four months out.

What we deliberately leave out

If a practice needs any of these, it has outgrown this version. Needing two of them is a better sign they have moved up a level than any revenue number.

Associate pay structures and partnership tracks
Multi-location scorecards and common definitions across sites
Function reviews by discipline rather than by location
Regional or lead-doctor accountability layers
Bringing an acquired practice into the group
De novo ramp curves
Thirteen-week cash
Board or sponsor reporting
Diligence prep running alongside everything else