A practice does not need forty KPIs. It needs a dozen figures, each with a definition, a source and a person who owns it — and 2026 ranges for what good looks like.
The dashboards sold to dentistry tend to fail in one of two ways: they show too much, so nobody looks; or they show numbers that were computed differently last month, so nobody trusts them. The cure for both is the same. Pick a short list. Write down the definition of each. Record the date the source data was pulled on every report. A figure with no source is not a low number — it is a missing one, and it should look missing rather than quietly show as zero.
The ranges below are 2026 industry estimates compiled from practice‑management benchmarks (Dentx and others). They are ranges, not targets: a fee‑for‑service practice in a wealthy suburb and a Medicaid‑heavy practice in a rural county will sit in different places for good reasons. Use them to know which of your twelve deserve a question this week.
Collections rate — collected ÷ net production over the same period. 95% is average; 98%+ is good; 99%+ is elite. Below 90% something structural is wrong, usually in claims follow‑up or patient balances. Watch that "net" means after contractual write‑offs; gross production makes everyone look worse than they are.
Overhead — total operating expense ÷ collections, excluding the owner‑dentist's compensation and any owner distributions. About 65% is average; 55–60% is good; under 55% is elite. This is the number everyone quotes and nobody defines: the same practice reads 55% or 68% depending on whether the owner's pay is inside the line. Decide once, write it down, and never compare your figure to someone else's until you've asked how they count.
Production per hour — net production ÷ scheduled provider hours. $450 average; $600+ good; $700+ elite. The denominator is the trap: count the hours the chair was scheduled, not the hours the building was open.
A/R over 90 days — the share of accounts receivable older than ninety days. 8% is average; under 5% is good; under 2% elite. The ninety‑day bucket is where write‑offs are born. A weekly cadence on the 60–90 bucket keeps it small.
Case acceptance — treatment scheduled or started ÷ treatment presented, over a window long enough to be fair. Measured on the day, 50–65% is typical; measured over 30–90 days, 75–85% is good and 90%+ is elite. Practices that quote 90% usually mean preventive care; major restorative sits nearer 55–70% on average. Track it by procedure class or it will mislead you.
No‑show rate — missed without notice ÷ scheduled. 8% average; under 5% good; under 2% elite. Late cancellations inside 24 hours belong in this number too, or it flatters you.
New patients per month — 25 average; 40+ good; 50+ elite — but read it against capacity. Forty new patients into a schedule that is already full at full fee is a marketing bill, not growth.
Retention — active patients seen in the last 18 months ÷ active patients a year earlier. 82% average; 90%+ good. This is the number that tells you whether the front desk's recall work is happening.
Hygiene share of production — 30–33% is the healthy band. Much lower and hygiene is under‑scheduled or under‑diagnosed; much higher and the doctor's chair has a problem.
Hygiene reappointment — patients who leave with their next hygiene visit booked ÷ hygiene visits. 85%+ is the target. Every point below it becomes recall work later, and recall work is the first thing a busy desk drops.
Perio share of hygiene — about a third of hygiene production coming from periodontal codes is typical of a practice that diagnoses what it sees. Far below that is usually under‑diagnosis, not a healthy population.
Production‑to‑wage ratio — hygiene production ÷ hygienist wages, around 3.5 to 1. Below 3 the department is subsidised by the doctor's chair.
Not all twelve every week. Pick the three that are furthest from their range and ask one question of each: what changed, when, and who owns it? Then leave the other nine alone until next month. A practice that looks at three numbers honestly beats one that glances at forty.
Print last month's production report. Next to each of the twelve, write the definition you used and the date the data was pulled. If you can't write either for a number, cross it out — it isn't a number yet.
Inside Vella OS every figure is labelled fact, forecast or realised, with the date its source data was recorded, and a missing source shows as missing — never as zero. Run your own numbers →
Back to the curriculum — pick the seat you're sitting in and read five, in order.
Canopy University →Ask her. She answers from the lessons here, and says so when the honest answer is that it isn't written yet.
Ask about this lesson →
Every lesson here is written for the person doing the work, whether or not they ever buy from us. Inside Vella OS the same questions get answered from your own record.
The scalable operating system that actually improves financial performance on its own.