USE CASE: How to Choose the Marketing KPIs That Actually Fit Your Business Model for a Dental Clinic

A real-world GenAI marketing use case: how a dental clinic stopped tracking impressions and clicks that told it nothing, and used GenAI to derive the handful of marketing KPIs that genuinely reflect how a dental practice makes money.

The marketing KPIs that matter are not universal; they’re decided by your business model. This is what putting that principle into practice looks like: how a dental clinic replaced a dashboard of generic vanity metrics with the few numbers a dental practice actually runs on, by deriving them from how the business makes money.

The Context: a Dashboard Full of Numbers, None of Them the Business

A large regional dental clinic with a marketing dashboard that looked healthy: impressions up, reach up, likes up, website visits up, ad clicks up. Every meeting had numbers to point at. What nobody could answer from that dashboard was the only question that mattered – is the practice actually growing?

The Challenge: Measuring Activity, Not the Business

The dashboard measured activity, not the business. Impressions, reach, likes, clicks – those are generic metrics; they would look the same for a shoe shop or a software company, and precisely because they’re generic, they say nothing about the health of this practice. A dental clinic makes money in a specific way: it acquires patients affordably, gets them to accept the treatment they need, and keeps them coming back for years of recall and care, all within the hard limit of its chairs. None of that showed up on the dashboard. And the trap of vanity metrics isn’t only that they’re shallow; it’s that they let a practice feel measured while flying blind on the things that decide whether it grows: whether a new patient costs more than they’re worth, whether treatment plans are being accepted, whether patients ever come back. Activity was being counted. The business was not.

Generic metrics measure a generic business: There is no universal set of marketing metrics. The ones that matter are decided by how your business actually makes money; so if your dashboard would look identical for a shoe shop and a software firm, it isn’t measuring your business. The right KPIs aren’t borrowed; they’re derived from the model.

The GenAI Workflow: Derive the KPIs from the Business Model

The fix was to stop borrowing metrics and start deriving them, to begin from how the practice makes money and let that dictate what to measure. The team used GenAI as a translation tool: it described the clinic’s real business model (how patients are acquired, how treatment is recommended and accepted, how recall brings them back, how chair capacity caps the whole thing), and asked GenAI to work out which handful of metrics actually reflect that model, and which of the current metrics were vanity. What came back wasn’t clicks and likes; it was the economic vital signs of a dental practice: what a new patient costs to acquire, how much of recommended treatment gets accepted, how many patients return for recall, what a patient is worth over years, how full the chairs run. GenAI proposed the set; the team tailored it to what this clinic actually was and could measure, and retired the vanity metrics. The dashboard stopped measuring activity and started measuring the business.

🖥️

The GenAI prompt:

You are a marketing consultant helping a dental clinic choose the marketing KPIs that actually fit its business model.

Here is how this practice really makes money: [how patients are acquired, how treatment is recommended and accepted, how recall works, the private/insurance mix, chair capacity, our goals].

1. From that model, derive the handful of KPIs that genuinely reflect how this practice makes and keeps money — with a one-line definition and formula for each.
2. Then list the metrics we currently track that are vanity for this model, and say what each should be replaced by.

Keep it to the few that matter, not a long dashboard. Do NOT invent our numbers; define what to measure, not the values. Flag any KPI that could push us to pressure patients or measure something unethical, so we don’t optimise the wrong thing. And mark anything you’re assuming about our model as CONFIRM WITH ME.

The caveat that decides whether this works: The right metric set is derived from a business model, and GenAI does not know yours; it knows the textbook dental practice, not this clinic’s real mix of private and insurance work, high-value and routine treatment, one site or several. So what it produces is a strong template to tailor, not a dashboard to adopt whole.

Three more boundaries: GenAI helps you choose what to measure, not measure it, it has none of your numbers, and if asked for figures it will invent plausible ones, so it defines the KPI and you supply the data. The point is fewer, truer metrics, not a longer list, GenAI will happily generate thirty, and a dashboard of thirty is just vanity metrics with better names; insist on the handful that matter. And a KPI changes behaviour, which in healthcare cuts both ways: a metric like treatment acceptance, pushed carelessly, can become pressure on patients – so the metrics have to serve the patient and the practice, never distort care. GenAI derives the candidates; the fit to your real model, the honesty of the numbers, and the ethics of what you reward stay human.

The Result: a Dashboard that Measures the Business

The dashboard changed from activity to business. The clinic stopped watching impressions and likes and started watching the metrics its business model actually runs on: what a new patient costs to acquire, how much recommended treatment gets accepted, how many patients come back for recall, what a patient is worth over years, how full the chairs run. Because those KPIs were derived from how the practice makes money – not borrowed from a generic template – every one of them connected to a decision the clinic could act on. And because the set was kept deliberately small, the team could actually see the business rather than drown in numbers. No invented figures here: the change is that the clinic stopped mistaking a rising activity chart for a healthy practice, and started measuring the few things that decide whether it is.

Recommended KPIs to Follow

This case is itself a KPI-selection exercise, so the recommendation is the payoff: the handful of metrics a dental practice’s business model actually runs on, a world away from impressions and likes. Here’s where the evidence sits and the direction each should push. The point is the direction of travel, not a promised number.

Treatment Plan Acceptance Rate

The share of recommended treatment (by value) that patients actually accept and schedule, the biggest revenue lever a practice controls without a single extra new patient. A gap here is money already diagnosed and then lost.

Benchmark: The ADA target is roughly 75-80% acceptance, yet real-world acceptance for elective and major work often sits at just 50-65%, a costly gap between treatment diagnosed and treatment scheduled (CareCredit / ADA; Dental Economics).

Patient Lifetime Value (via recall & retention)

What a patient is worth over years of recall and care, not a single visit. This is the metric that reframes acquisition: a costly new patient can be a bargain if recall keeps them for a decade, so track the recall/reappointment rate that drives it.

Benchmark: No clean cross-market LTV figure, an internal metric; recall/reappointment rate is the engine to baseline, and a patient who completes treatment is far more likely to return for recall and refer (Teero). Set your own baseline.

Chair / Appointment Utilisation

How much of your available chair time is actually booked and delivering care, the capacity constraint that caps the whole model. Empty chairs are revenue that can never be recovered; this is where marketing meets operations.

Benchmark: Utilisation is measured as patient chair-time ÷ available chair-time; underutilisation points straight to lost revenue (a practice with chairs open 160 hours, but used 120 runs at ~75%) (Teero). Baseline yours and watch the gap.

Acquisition cost belongs here too (cost per acquired patient, not per click) and it’s the natural bridge to the case on paid search. But note the framework’s own point in miniature: even within dentistry, a paediatric practice may prioritise recall and new-patient numbers while a specialist prioritises case acceptance, so treat this as the dental starting set and derive yours from your model. Track your own trend; the benchmarks are context.

Why this Transfers

Every business can find a dashboard of metrics that go up and to the right while the business itself stalls, because generic metrics measure a generic business, and you don’t run one. The transferable move is to start from how you actually make money and derive your KPIs from that, keeping only the handful the model turns into revenue, then treat every metric that would look identical for any other business as the vanity metric it is.