The cash flow trap unique to ABA
Your RBTs get paid every two weeks. Your payers pay in forty-five to ninety days, sometimes longer, and only after a claim clears.
That gap is the entire problem. You deliver a session on Monday, pay for it two weeks later, and collect for it two months after that. In between, you are financing your own payroll out of the balance sheet.
For a single clinic, that is manageable. Across four or six clinics, with a de novo ramping and an acquisition integrating, the gap becomes the thing that can end an otherwise healthy company. I have seen profitable ABA practices come within one payroll cycle of failing — not because the business did not work, but because nobody was looking thirteen weeks out.
The uncomfortable part: growth makes it worse. Every new client you add consumes cash before it produces any. Every new clinic burns for a year. A practice growing at forty percent has a harder cash problem than one growing at ten, and owners consistently discover this at the worst possible moment.
Why the P&L will not warn you
Profit and cash are different, and in ABA they can diverge dramatically for months at a time.
Your income statement records revenue when the service is delivered. Your bank account records it when the payer pays. In a growing practice, the difference accumulates on the balance sheet as accounts receivable — which is to say, as a number that looks like an asset and behaves like a hole.
A practice that grows revenue thirty percent in a year can show excellent EBITDA and still have less cash in December than it had in January, because the growth is sitting in AR. The P&L is not wrong. It is just answering a different question than the one that matters on payroll Friday.
Profit is an opinion about the period. Cash is a fact about the day. You can only be wrong about cash once.
The thirteen-week forecast
The tool that fixes this is a rolling thirteen-week cash forecast. Thirteen weeks is one quarter — long enough to see a problem while you can still act on it, short enough that the estimates mean something.
It is not a budget. A budget is monthly, annual, and about targets. This is weekly, rolling, and about survival. It answers one question: what is the cash balance at the end of each of the next thirteen weeks?
The structure is simple. For each week: opening cash, plus expected receipts, minus expected disbursements, equals closing cash — which becomes next week's opening balance. Every week you drop the week that finished, add a new week thirteen, and update the estimates.
The discipline of updating it weekly matters more than the sophistication of the model. A rough forecast maintained every Monday beats an elegant one built once in March.
Building the receipts side
This is the hard half, and it is where most ABA forecasts fail. Owners estimate collections as a percentage of billings, which is far too blunt.
What you need is a collections curve by payer. For each significant payer, look at your last twelve months and answer: of the dollars billed in a given week, what percentage arrived in week two, week four, week six, week eight, week twelve?
The curves differ more than you would expect. One commercial plan may pay eighty percent within thirty days. Another may center on sixty. Medicaid in your state may run differently again. Applying a single blended assumption across all of them produces a forecast that is wrong in both directions and useless for planning.
Once you have the curves, forecasting receipts becomes mechanical: take billings by payer by week, apply that payer's curve, and sum. Layer in your existing AR aging on top, because those dollars are already in flight.
Three things to build in:
Denials and rework. A portion of what you bill will not be paid on the first pass. If your first-pass rate is ninety-three percent, do not forecast a hundred. Model the denied portion as arriving much later, or not at all.
Seasonality. ABA volume moves with school calendars, summer schedules, and holiday closures. December and late summer look different from March. Your own history will show you the pattern.
Payer behavior changes. A payer that shifts its processing cycle can move a million dollars of expected receipts by three weeks with no notice. This is why the forecast is rolling — you catch the shift in the variance and reforecast, rather than discovering it when the balance runs low.
Building the disbursements side
Easier, because you control most of it. But precision on timing matters.
Payroll is the dominant line. Get the actual dates on the calendar, and note that a biweekly cycle produces three payrolls in some months. That third payroll has caused more cash surprises than any other single item I have seen. It is entirely predictable and routinely forgotten.
Payroll taxes follow their own deposit schedule, often separate from the payroll run itself.
Rent on the first, across every location.
Then the rest — insurance, software, supplies, professional fees — most of which are predictable within a few hundred dollars.
Debt service, including any line of credit interest, on its actual dates.
Capital expenditure and one-time items. A clinic build-out, a deposit, an earnout payment, a tax distribution to owners. These are the lines that blow up an otherwise fine forecast, because they are large and irregular. Put them in as far ahead as you know about them.
The de novo problem
If you are opening clinics, this deserves its own line in the model, because a new clinic is a cash consumption event with a long tail.
Before it sees a single client, a de novo costs you a lease deposit, build-out, furniture and materials, and credentialing time. Then you hire clinicians who are paid in full from day one while the caseload fills over months. Then those first sessions bill, and the cash arrives another sixty days after that.
Realistically, a new ABA clinic consumes cash for somewhere between nine and eighteen months depending on your market and ramp speed. Two clinics opened in the same quarter can consume more cash than a comfortable-looking business generates.
This is why the pattern of a fast-growing, profitable practice running out of money is so common. Nobody did anything wrong. They just stacked de novo burn without modeling it week by week.
Model each clinic separately with its own ramp, and use your own historical curve if you have one. If you do not have one yet, be conservative — first clinics almost always ramp slower than the plan.
What good looks like
A few practices separate the owners who sleep well from the ones who do not.
Know your floor. Decide the minimum cash balance you will operate above — I would generally want at least one full payroll cycle plus rent, and more if you are mid-expansion. The forecast is not there to tell you the balance. It is there to tell you when you will cross the floor.
Forecast a range. Run a base case and a downside where collections slow by two weeks and cancellations rise a few points. If the downside breaches the floor in week nine, you have nine weeks to act — which is a manageable problem. Discovering it in week two is not.
Track variance every week. Forecast versus actual, by line. This is what makes the model get better. If receipts consistently come in five percent under forecast, your curves need adjusting, and you would rather learn that from a pattern than from an overdraft.
Get the line of credit before you need it. Banks lend to practices that can demonstrate they understand their own cash cycle. A thirteen-week forecast is the single most persuasive document you can put in front of a lender. It is also much harder to get facility approved during the quarter you actually need it — arrange it while the numbers look calm.
Attack the receipts side, not just the model. The forecast is diagnostic. The cure is usually operational: clean claim rate, denial rework speed, aged AR follow-up, authorization tracking. Shortening your collection cycle by ten days is permanently worth more than any financing arrangement, and it costs nothing but attention.
Who should own it
In a practice under roughly $5M, the owner can run this personally in a spreadsheet. It takes a few hours to build and perhaps thirty minutes a week to maintain.
Past that, it should belong to a controller or finance lead, with the owner reviewing weekly. What should not happen is what I most often find: the forecast exists, someone built it once, and nobody has updated it since the quarter it was created. A stale cash forecast is worse than none, because it produces confidence without information.
The practices that handle growth well are not the ones with the best margins. They are the ones that knew, in March, what their cash position would be in June — and had already decided what to do about it.
If you are opening clinics this year and have not modeled the cash burn week by week, that is worth doing before the lease is signed rather than after.