“We’re full for three weeks. The problem is the billing.” That’s a practice owner reading a thin bank statement next to a packed schedule, and the second sentence is the one to test.
Revenue leakage detection starts earlier than billing. By the time a patient sits down with a clinician, the practice has already decided most of what that visit will pay. It has decided whether the plan is one the practice is in and whether the coverage is active today. It has also decided whether the patient’s share gets collected at the desk or chased by mail for months, and whether the next visit exists before this one ends.
Those decisions get made at the front door, and the front door isn’t where anyone looks for money. The billing team looks at claims, the bills sent to insurers, the owner looks at collections, and the front desk looks at the phone. The stretch between a patient’s first call and the moment they walk back to the exam room belongs to nobody, and it leaks at nine separate places.
This is the map of that stretch: what each step is, why it fails, why the failures don’t look like money, what they cost, and the morning check that closes them. A dozen shorter pieces on this site each take one step apart, and this is the one that puts them in order.
What is the front door of the money path?
The money path runs from the first time a patient contacts the practice to the day their balance is zero. Revenue oversight is the job of watching all of it, every day. The front door is the first stretch: everything that happens before care is delivered, plus the two things that happen on the way out.
Nine steps, in order.
A person calls or fills out a form and becomes a lead, meaning someone who wants an appointment and doesn’t have one yet. A referral arrives from another practice, which is the same thing with a fax attached. The lead becomes a booked appointment, and the appointment gets confirmed.
The patient’s insurance gets checked: is the plan one the practice is in, is the coverage active on the visit date, and what will the patient owe. The patient arrives, or doesn’t. The patient pays their share, or doesn’t. The next visit gets booked before they leave, or doesn’t. And the patient comes back next month, or quietly stops.
Every one of those nine has a count and a dollar figure. Almost no practice runs any of them, because the front desk is measured on whether the phone got answered and the chairs got filled. Both can be true while the money walks out.
Who owns the front door at your practice today?
Ask who is responsible for the patients who called and never booked, then ask who owns the patients who stopped coming. Watch how long the silence lasts.
The front desk owns the phone and the check-in, and intake, if the practice has one, owns the new-patient paperwork. The billing team owns everything from the claim onward and considers the visit to have started when the charge appears, and the clinician owns the room. Every one of those people is doing their job.
Nobody owns the lead who never booked, the referral that never turned into a call, or the patient booked under a plan the practice isn’t in. The same goes for the balance that could have been collected at the desk and the patient who was never asked to book the next visit. Each is a thing that didn’t happen. A thing that didn’t happen is nobody’s task, and the front door is made of them.
Why the front door decides the rest of the month
Everything downstream, meaning every later step, is arithmetic on what the front door lets through.
A visit booked under a plan the practice is out of network with produces a claim that will be denied, meaning refused by the insurer, no matter how well it is coded. A visit on a date the coverage had lapsed produces a claim to an insurer that doesn’t owe it.
A patient balance not collected at the desk becomes a statement, then a second statement, then a write-off.
Denials get created upstream, meaning early in the chain, and most of those refusals get created before the patient sits down.
And the front door is the only part of the money path that also decides revenue, not just collection of it. A lead that never booked is a visit that never existed, and a patient who stopped coming is a year of visits that never existed. The billing team can’t collect a dollar the front door never let in.
That’s the asymmetry. A denial costs the practice the work of fixing one claim. A lost patient costs every future claim that patient would have produced, and it doesn’t appear on any report, because there’s no claim to report.
The call that never became an appointment
Start at the beginning, because it’s the step with the least evidence.
Someone calls while the desk is busy, so the caller leaves a voicemail, or is told someone will call back, or is put on hold and hangs up. Someone fills out the website form, and the form goes to an inbox that gets read on Thursdays.
Each of those is a person who wanted care, at this practice, this week. Each one is a visit worth whatever a visit is worth at your practice, multiplied by however long that patient would have stayed.
None of them leave a record that looks like money: a voicemail that wasn’t returned isn’t a canceled appointment, because there was never an appointment to cancel. A form in an inbox isn’t a lost patient, because they were never a patient. The practice’s own numbers say the schedule is full. The schedule is full of the people who got through.
The count is simple: leads this week, appointments booked from them, and the gap. A practice that runs it will find the gap isn’t zero, and the pattern in it, which desk, which day, which hour, is where the fix goes.
The referral that died between two records
A referral is a lead with a paper trail, and it dies the same way.
The referring practice sends a fax or a message, and it lands somewhere: a queue, a folder, a person’s desk. Someone is supposed to call the patient, and on a good week they do. On a busy week the referral sits, the patient doesn’t hear from anyone, and after a while they go somewhere else or nowhere at all.
Every stage in the practice ends with a record, and a referral that dies between two of them ends with nothing. The referring practice thinks it sent a patient, the receiving practice thinks it never got one, and the patient thinks nobody called. All three are right, and the visit is gone.
The fix is a count with a clock on it: referrals received, referrals contacted within two business days, referrals booked. The step that’s failing shows up as the gap between two of those numbers.
The patient on the schedule whose plan you are not in
Now the appointment exists, and the next question is whether the practice can be paid for it at all.
Every practice is in network with some plans and not others. In network means the practice has a contract with that insurer and will be paid its rate. Out of network means no contract, and depending on the plan, a much smaller payment, a denial, or a bill the patient didn’t expect.
A patient books, gives the name of their insurer, and gets a slot. The name of the insurer isn’t the plan: one insurer sells dozens of plans, and the practice is in some of them. The desk hears a familiar name and books the visit.
The claim goes out, comes back denied for being out of network, and the practice either eats it or sends the patient a bill that ends the relationship.
The check happens at booking or it doesn’t happen. A list of plans the practice isn’t in, read against tomorrow’s schedule every afternoon, catches every one of these a day early. There’s still time to call the patient, move the visit, or explain the cost, but after the visit there’s nothing to do but absorb it.
The insurance check that said approved
Most practices run an automated eligibility check, meaning a query to the insurer that asks whether the patient’s coverage is active. It comes back with a flag: approved, or not.
The flag isn’t the answer: it says a policy exists. It doesn’t say the plan on that policy is one the practice is in, or that the coverage is active on the visit date rather than the day the check ran. It doesn’t say the service being scheduled is covered, or that the patient hasn’t switched plans since the card was scanned.
At practices we have worked with, the check reads approved on most of the visits later refused for coverage. The flag was accurate about a question nobody had meant to ask.
The check that works reads the response, not the flag. The plan on the response matched against the plans the practice bills, the coverage dates against the visit date, the service against what the plan covers. That takes a few seconds per patient when it runs automatically the day before, and it never runs when a person has to remember to open each response.
Coverage also expires quietly. Nothing happens the day a card expires, which is exactly why nobody catches it. A patient whose coverage lapsed last month keeps their appointments, and every one of them produces a claim to an insurer that no longer owes it.
What the patient owes, and when you find out
Insurance pays part of most visits. The patient pays the rest: a copay, a fixed amount per visit; a deductible, the amount they pay in full before the insurer starts paying; and coinsurance, their percentage after that.
Together those are the patient’s share, and the practice finds out what it is at one of two moments: before the visit, from the eligibility response. Or weeks later, from the insurer’s payment report, the notice of what it paid and what it didn’t, after the patient has gone home.
The timing decides whether the money arrives. A large share of the patient dollars owed at the visit walks out uncollected and gets chased for months by statement. Patient balances decay with every day that passes. And the desk can’t ask for an amount it doesn’t know.
The desk also skips amounts it does know when asking feels awkward. Your real collections policy is a thousand half-second mercies at a busy desk. A system that asks before the visit doesn’t get busy.
There’s one more silent piece. At one practice, the billing system already knew which 171 patients’ cards on file had expired, and the balances those cards were supposed to cover had been sitting uncollected. The card expired, the automatic charge failed, and nothing told anyone.
The no-show, and the fee you wrote a policy for
The patient doesn’t come, so the chair sits empty for the hour, and the clinician’s day has a hole that can’t be sold on twenty minutes’ notice.
No-shows are a system output: they come from how the practice confirms, reminds, and rebooks. A confirmation that demands a reply beats a reminder that asks nothing. A same-hour rebooking call beats a letter, and a waitlist that fills the slot beats an empty hour.
Most practices also have a no-show fee in the policy. A no-show fee the practice wrote and never billed is money left at the front door when coverage allows the fee at all.
The next visit that never got booked
Now the patient was seen. Two steps left, and they decide next month’s revenue.
The first is the next appointment: a patient who needs a follow-up and leaves without one has to remember to call. In telehealth there is no desk to walk past, so the next visit never gets booked. A follow-up booked before the patient leaves is a visit that exists. The count is the share of visits that ended with the next one on the calendar, by clinician, and it’s the single best predictor of next month’s schedule.
The patient who left without telling you
The second is the patient who simply stops. They didn’t cancel. There was nothing to cancel, because no next visit was booked. The schedule filled behind them, and the visit count stayed flat while the patient base turned over underneath it.
A no-show creates a record. A patient who stops coming creates nothing at all. Compare each patient’s expected return against the calendar and count the ones past due with nothing booked. At one practice, the slow month started three months earlier, when patients left through a door nobody was watching.
The pattern is rarely practice-wide. Split it by clinician and the comfortable average falls apart. At practices we have worked with, the fix was three conversations, not a marketing budget.
The chairs that sat empty while patients waited
One more front-door failure belongs here because it’s two failures that cancel each other’s evidence.
Intake knows who is waiting for an appointment, and operations knows which hours are open. Nobody in the building has ever put those two numbers on the same page. So a practice can carry a waitlist and empty chairs in the same week, each department certain the other is the bottleneck, and the visits that could have happened don’t.
The count is unfilled bookable hours inside the next ten business days, by clinician, next to the number of patients waiting. When both are above zero, the fix is a phone call, and it pays for itself by lunch.
What the front door costs
The cost lands in three places, and only one of them ever gets a name.
The first is claims that were never going to pay: out-of-network visits, lapsed coverage, wrong plans. The billing team works the denial and gets blamed for a rate that was set at the front desk.
The second is patient money that leaves the building: the copay not asked for, the deductible nobody knew about until the payment report, the expired card, the no-show fee never raised. Together they’re why fewer than one in four patient appointments at one practice reached paid in full.
The third is revenue that never existed: the lead who never booked, the referral nobody called, the patient never asked to rebook, the one who stopped coming. None of these produce a claim or a report. Every stage a patient touches sends a signal. Every stage that’s entirely yours sends nothing. The front door is almost entirely yours.
Real situations, and what the front door hid
At practices we have worked with, a denial review that starts in billing ends at the front desk, where the check reads approved on plans the practice doesn’t bill. The check was running, but nobody was reading the response, and the plan on it wasn’t one the practice billed.
At another, the desk collected copays and nothing else, because the deductible amount wasn’t on the screen when the patient stood there. It arrived weeks later on the insurer’s payment report, and by then the patient was a statement, then two, then a write-off. The money had been knowable on the day of the visit.
At a third, a practice with a full schedule was losing its patient base under the schedule. Nearly half the patients who should have returned didn’t, and no visit had been canceled, because none had been booked. The schedule was full of first visits.
What the daily check looks like
The practices that close the front door run a short list every afternoon for tomorrow and every morning for yesterday. Each line has a count, a dollar figure where there’s one, and a name.
- Leads and referrals received this week with no appointment booked, and how long since contact. Owner: intake. Fix: call today.
- Tomorrow’s schedule read against the plans the practice isn’t in, and against coverage dates. Owner: the front desk lead. Fix: call the patient before the visit.
- Tomorrow’s schedule with each patient’s share next to the name, from the eligibility response. Owner: the front desk. Fix: ask for it at check-in.
- Yesterday’s no-shows, each with a rebooking call logged and the fee raised. Owner: the front desk. Fix: same day.
- Yesterday’s completed visits that ended with no next appointment booked, by clinician. Owner: each clinician’s desk. Fix: call and book.
- Patients past their expected return with nothing on the calendar, by clinician. Owner: intake. Fix: outreach this week.
- Open bookable hours in the next ten days next to the waitlist count. Owner: scheduling. Fix: fill from the list.
Each list is short at a practice that runs it daily, but the first week, every list is long, because none of them has ever been run. That first week is the finding.
And every list carries a do-not. Don’t book a patient under a plan you aren’t in and hope. Don’t send a statement for an amount the desk could have asked for. Don’t let a no-show go without a rebooking call because the desk is busy, because the desk is always busy.
Why the reports cannot show you this
Every standard report a practice runs starts at the charge or the claim: collections, aging (unpaid bills sorted by how old they are), denials, days to payment. All of them describe what happened to visits that happened.
The front door is the visits that didn’t happen, and the money that left before a charge existed. A lead isn’t a claim, and a lapsed patient isn’t a denial. A copay not asked for isn’t on the aging report, the list of unpaid bills by age, until it’s a statement, and by then it’s a different problem with a different name.
So the numbers that describe the front door have to be built from records the practice already keeps and has never combined. The pairs are the phone system and the schedule, the schedule and the plan list, the eligibility response and the check-in screen, and this visit and the next. Each pair is an afternoon to build once and a minute to read every morning.
Where does the practice start?
With three counts that take an afternoon.
Pull tomorrow’s schedule and read every patient’s plan against the list of plans the practice isn’t in. The number of matches is the first finding. At most practices it isn’t zero, and every one is a denial that hasn’t happened yet.
Pull last month’s completed visits and count the ones that ended with no next appointment booked, by clinician. That number, and the spread between clinicians, is the second finding.
Pull the no-shows from the same month and count how many have a fee raised and how many got a rebooking call. The gap between the policy and the count is the third finding.
Then pick the first step to instrument. For most practices it is the rebooking, because the next visit booked before the patient leaves is the number that moves next month’s schedule. It moves in a week once a clinician sees it with their name on it.
For a practice with a climbing denial rate, it’s the plan check on tomorrow’s schedule. For a practice with a statement backlog, it’s the patient’s share on the screen before check-in.
The rest of the money path starts where the front door ends. Once the visit happens, the stretch between the note and the claim takes over, and it leaks the same way for the same reason.
What this means for you
Your full schedule and your thin bank statement are the same practice, seen from the front door and the back. Before a patient sits down, the practice has already decided whether the plan pays, whether the coverage is active, whether the patient’s share gets collected, and whether there’s a next visit.
After they leave, it has decided whether they come back. None of those decisions appear on a report, and all of them are yours.
The fix is a short list, every afternoon for tomorrow and every morning for yesterday, of the things that didn’t happen at the front door, with a number and a name on each line.
Grab 30 minutes with us. Prep nothing. You will see how many of next week’s scheduled patients are booked under a plan you aren’t in, and how many of last month’s visits ended without a next one.
Questions practice owners ask
What’s the front door of a medical practice’s revenue?
The front door is everything that decides what a visit will pay before care is delivered, plus two steps on the way out. It covers calls and referrals, booking, confirmation, the insurance check, the patient’s share, and the no-show. On the way out, it covers booking the next visit and whether the patient comes back.
Why doesn’t a full schedule mean my practice’s money is fine?
Because the schedule counts visits and the bank counts visits that paid. A visit under a plan the practice isn’t in, on a date the coverage had lapsed, or with a patient share nobody asked for fills a chair and pays a fraction. And a full schedule can hide a patient base that’s turning over underneath it.
What’s the difference between an approved eligibility result and a real eligibility check?
An approved result only says a policy exists, while a real check reads the insurer’s full response. An eligibility check asks the insurer whether the patient’s coverage is active. Reading the response means matching the plan against the plans you bill, the coverage dates against the visit date, and the service against what’s covered. Approved can show on a visit the insurer will refuse.
When should a medical practice find out what the patient owes?
Before the visit, from the insurer’s answer to the coverage check, so the desk can ask for it at check-in. Finding out weeks later from the insurer’s payment report turns a desk collection into a mailed statement. Patient balances decay with every day that passes. And the desk can’t ask for an amount it doesn’t know.
How do I know if patients are leaving?
Compare each patient’s expected return against the calendar and count the ones past due with nothing booked. A patient who stops coming creates no record, so the count has to be built. Split it by clinician, because the pattern is rarely practice-wide.
Which front-door step should my practice fix first?
Start with booking the next appointment before the patient leaves, because it predicts next month’s schedule. It moves within a week once clinicians see their own number. If denials, the insurer’s refusals to pay, are climbing, start with the plan check on tomorrow’s schedule instead. If unpaid patient statements are piling up, put the patient’s share on the check-in screen first.