The few metrics that actually matter for running a healthy therapy practice

You probably already track plenty of metrics. The harder question is what they’re telling you.

You can pull a dozen reports out of your EHR and still not know whether your practice is doing well or starting to slip. The numbers are there. Knowing what to do with them is the hard part.

The few metrics that actually matter aren’t a longer list. They’re the ones you’ve spent enough time with to understand in context: how they compare to your own history, when a weird week occurs, and what the beginning of a problem forming looks like. The blog post you’re currently reading is about getting more out of the metrics you already have, not adding more. If you want a primer on which numbers to start with, the 5 KPIs every solo practice should track is a good place to begin before coming back here.

Why more metrics rarely fix anything

Most practice owners can rattle off a dozen key performance indicators (KPIs) they’re keeping an eye on. A lot fewer can tell you which ones are getting better, which are getting worse, and which haven’t budged in months. they’re keeping an eye on. A lot fewer can tell you which ones are getting better, which are getting worse, and which haven’t budged in months.

That gap is where most practice problems live. A dashboard with twelve indicators can feel like you’re on top of things. It can also be the reason you miss the one thing that’s actually going wrong, because everything looks similar.

Asking fewer questions more carefully usually works better than tracking more numbers.

The four questions a healthy practice can answer

Whether you’re solo or running a team of fifteen, a healthy practice can usually answer these four questions without much digging.

1. Is the schedule full enough to be sustainable?

It’s not your no-show rate alone, though that matters. The fuller picture is capacity utilization: what percentage of your available clinical hours actually turn into completed, billable sessions.

A 22% no-show rate sounds bad on its own. But if your schedule is 95% booked and your no-shows tend to refill from a waitlist, you’re probably fine. A 15% no-show rate doesn’t mean much if you’re only filling 60% of your available slots in the first place.

If you’re solo, this number swings hard week to week. One sick week, one slow start to fall, and your chart can look scary. What’s worth looking at is the trend over 8 to 12 weeks, not the number on any given Monday.

If you’re running a group, the average can hide what’s happening with individual clinicians. Three people at 90% utilization and one at 50% averages out to a number that looks healthy, but it’s telling you someone on your team needs support, marketing help, or a referral conversation.

2. Are you getting paid for the work you’re doing?

Days in accounts receivable (A/R) is the number most people track. It’s useful, but it doesn’t tell the whole story by itself.

Look at it alongside two other data points: the percentage of your A/R sitting in the 90+ day bucket, and your net collection rate (what you actually receive divided by what you’re contractually owed). Those three data points together tell you whether money is moving fast enough and whether you’re actually collecting what you billed.

One of these can look fine while the other is leaking. A 35-day average in A/R looks great until you notice the 90+ bucket has doubled in two months, and your net collection rate has quietly dropped from 96% to 91%. The average lagged because it was getting watered down by faster-paying claims.

When you start seeing that kind of pattern, the next step is figuring out where the leak is coming from. The billing health check guide walks through what to look at first, including how to recognize denial patterns and where payment leaks tend to hide.

3. Are clients staying long enough to benefit?

Most practices look at retention as a single number: what percentage of clients are still active after some period.

That’s a snapshot. What’s actually useful to know is where people are falling off.

If most of your dropouts cluster after intake but before session three, that’s usually a fit, expectations, or onboarding issue. If they cluster around session eight to ten, that’s often where natural ending points show up such as clients feeling ‘good enough,’ authorization renewals coming due, or life circumstances shifting, and where clients may not be getting supported through what comes next. If drop-offs spike around insurance verification or copay changes, the problem may not be clinical at all.

A flat retention number hides all of this. The same 65% three-month retention number could come from a healthy curve or a broken one. Where the drop-offs happen tells you more than the percentage does.

4. Is the work sustainable for the people doing it?

This question gets the least attention, and it’s often the earliest warning that something’s going to show up later in your revenue and retention.

If you’re solo, what to watch is how much of your time is going to admin versus clinical work. When admin starts creeping up week over week, that’s the early sign: notes get rushed, claims go in late, follow-ups slip, and a couple months later one of the other three questions starts looking off.

If you’re running a group, watch documentation lag (how long it takes for notes to get done after a session), supervision load, and how much clinician schedules are bouncing around week to week. Notes consistently completed more than 48 hours after the session are something to pay attention to. Many payers, including Medicare and Medicaid, expect notes within 24 to 48 hours of session. Notes that are routinely a week late cause real problems, both for compliance reasons and because it usually means people are running too hot.

A practice can look financially healthy while its team is burning out. Six months later, those numbers catch up.

Watching for patterns, not snapshots

Once you know which questions you’re answering, the work is paying attention to how the numbers move over time, not just where they land in any one week.

A single bad week is usually just noise. If three weeks are moving in the same direction, it’s worth a closer look. If this continues for a whole quarter, it’s worth doing something about.

The most useful comparison, then, is your practice against its own past, not your practice against an industry benchmark. Benchmarks are okay as a rough starting point, but they can’t account for your payer mix, your modality, where you practice, your team, or your stage of growth. The benchmark for “good” looks different for a CBT-focused solo practice in a high-density urban area than for a rural group doing trauma work with a large Medicaid panel.

The question that beats “are we hitting the benchmark” is “is this better or worse than where we were last quarter.”

How this looks different in solo vs. group practice

The questions are the same. How you look at them is different.

If you’re solo, you’re working with smaller numbers that swing harder. One canceled week can really move your monthly numbers. So, pattern length matters more than any specific threshold, and it’s easy to overreact to a bad month.

If you’re running a group, the numbers are bigger and smoother, which has its own problem: the averages look fine, but they can hide individual clinicians, payers, or service lines that are quietly underperforming. To get something useful out of the numbers, you have to break the averages apart by clinician and by payer. It’s easy to miss real problems because the average makes everything look okay.

Either way, the goal is the same: a clearer answer to those four questions, checked often enough that small problems don’t have time to become bigger ones.

A simple monthly review

If you don’t have a monthly review habit yet, here’s the smallest version that works. 

Block off thirty minutes once a month, at the same time every month if you can. Pull the same handful of numbers each time: capacity utilization, days in A/R along with 90+ day aging and net collection rate, your retention curve, and a sustainability number (admin hours if you’re solo, documentation lag if you’re running a group). 

Compare each one to last month and to the same month a year ago. Look at direction first, the actual value second. Then ask yourself: what changed, and what would I want to be different next month. 

You’re not trying to find every problem. You’re just trying to make sure no problem stays hidden for too long.