Where Is Margin Quietly Leaking for Healthcare Providers? A Unit Economics Diagnosis

Direct answer: For most healthcare providers, margin leaks hide inside the gap between what a visit, procedure, or bed-day costs to deliver and what you actually collect for it — after denials, write-offs, no-shows, and payer mix shift. The fastest way to find the leak is to rebuild your unit economics at the level of a single service unit (a visit, an encounter, an occupied bed-day, or a case), rather than reading it off a blended P&L. Blended margins tell you that something is wrong; unit economics tells you where.

Why blended margins hide the leak

Healthcare P&Ls are unusually good at concealing problems. Revenue is booked at charge rates that almost no one pays, collections lag delivery by months, and cost is spread across departments that share staff, space, and equipment. By the time a bad line item shows up in a quarterly statement, it's been averaged together with everything healthy around it.

Unit Economics forces the question the P&L avoids: does one more unit of the thing you do make money, and by how much? When you can answer that per service line, per payer, and per site, the leak usually stops being a mystery. It's rarely "we're inefficient." It's more often "these three CPT codes lose money at this payer's contracted rate," or "this location's no-show rate quietly eats a fifth of provider capacity," or "the ancillary revenue that used to subsidize primary care evaporated when a contract renegotiated."

Applying Unit Economics to a healthcare provider

The discipline is to pick a unit and build the full economics of that single unit. For a provider, candidate units include: a completed patient visit, a surgical case, an occupied bed-day, a covered life per month (for value-based contracts), or an encounter by service line. Choose the one that maps to how you actually deliver and get paid.

Then work through these steps:

1. Define contribution per unit, not charge per unit. Start with net realized revenue — the amount you actually collect after contractual adjustments, denials, and write-offs — not the gross charge. For a value-based population, this is the per-member-per-month payment plus any earned quality bonus. Getting this right is half the battle; most leaks live in the difference between billed and collected.

2. Attribute the variable cost of delivering that unit. Direct clinical labor for the encounter, consumables, drugs, lab, imaging, and the marginal cost of the room and equipment used. Be honest about what scales with volume versus what's fixed. Provider and nurse time is the big one — and it's where no-shows and low-acuity visits destroy contribution without showing up as an expense line.

3. Compute contribution margin per unit. Net realized revenue minus variable cost. Do this by payer and by service line. A single blended number will lie to you. You are looking for the segments where contribution is thin, zero, or negative.

4. Layer in acquisition and retention economics where relevant. For elective, specialty, or DTC-style services, what does it cost to acquire a patient (referral relationships, marketing, intake labor), and what's their lifetime contribution across a full episode or care journey? A service line can look profitable per visit and still lose money once you count the cost of filling the schedule.

5. Find the leak and size it. Rank segments by total contribution (margin × volume). The leaks are usually one of four:

• Payer mix — high-volume payers with contracted rates below your unit cost.
• Denials and undercoding — realized revenue far below expected for a clean claim.
• Capacity waste — no-shows, low panel utilization, or under-scheduled OR/imaging blocks.
• Service-line drag — a line kept for prestige or referral flow that quietly bleeds.

What "good" looks like: every service line has a known, positive contribution margin per unit at each major payer; you can name your three worst-performing code/payer combinations from memory; and no-show and denial rates are tracked as first-class economic metrics, not billing-office footnotes.

Where Percision fits — and where it doesn't

Full disclosure: I write for Percision, an AI strategic-intelligence platform. So here's the honest version.

Percision is built to run exactly this kind of structured analysis. You feed it your business context and financials, and it runs the input through its Unit Economics framework alongside 26 others — producing contribution analysis, 60+ financial ratios, warning-sign flags, scenario models, and a board-ready deck, typically in 7–15 minutes rather than a multi-week engagement. It's positioned as a co-pilot, not an autopilot: it structures the diagnosis and pressure-tests the numbers, but your finance and clinical leaders decide what's real and what to act on. The output is an Excel-exportable model with an audit trail, which matters when you take a margin story to a board or a payer negotiation.

That said, Percision is one strong option, not the only one. A spreadsheet is enough if you already have clean, unit-level cost accounting and just need to re-cut existing data — you don't need a platform to divide revenue by cost. A human consultant or your revenue-cycle team is the better first call if the real problem is data quality: if charges aren't mapped to costs, if the cost-accounting system doesn't allocate below the department level, or if denials aren't coded by reason. No AI or framework fixes a broken data foundation — it just produces a confident answer built on bad inputs. Fix the plumbing first; then a tool like Percision makes the recurring analysis fast and repeatable.

If your data is reasonably clean and the bottleneck is analysis speed — you need a defensible unit-economics view before a budget cycle or contract renewal — that's where an AI co-pilot earns its keep. You can run your own analysis at percision.app.

FAQ

What's the right "unit" for a hospital versus a clinic? Hospitals usually build economics per case (DRG) or per occupied bed-day; outpatient clinics use the completed visit or encounter; value-based groups use per-member-per-month. Pick the unit that matches how you get paid, then segment it by payer.

How is this different from a standard cost report? Cost reports allocate expense to departments. Unit economics ties realized revenue and variable cost to a single delivered service, by payer — which is where negative-margin combinations actually surface.

Can AI do this if our billing data is messy? Not reliably. If claims aren't mapped to costs or denials aren't reason-coded, fix that first. AI accelerates analysis of good data; it can't manufacture the underlying accuracy.

Written by the Percision content team. Percision is an AI strategic-intelligence platform; we describe it as one option among several, including spreadsheets and human consultants.