13-Week Cash Flow Forecasting for Restaurant Operators

The most uncomfortable moment in restaurant operating life is realizing on a Friday morning that payroll runs on Tuesday, the food order is being delivered Monday, and the bank balance is not going to cover both. The restaurant is profitable. The P&L is fine. The cash position is not.

This is the gap that destroys restaurants that have no business being destroyed. The P&L says you made money. The bank account says you can't make payroll. Both are accurate, and the bridge between them is cash flow forecasting.

A 13-week rolling cash flow forecast is the single most useful piece of financial discipline an independent operator can install. It costs nothing, takes about four hours to build the first time, and turns the most stressful week of every quarter into a calm one.

Why profitable restaurants run out of cash

Three structural mismatches sit between the P&L and the bank account.

Accrual versus cash. Your P&L books revenue when you serve it and books expense when you incur it. Your bank account moves when the money actually changes hands. A catering deposit hits the bank in March and books to revenue in May. A vendor invoice books to expense on the 15th and gets paid on the 30th. Over a year these flatten out. In any given week they don't.

Lumpy outflows. Rent prints on the 1st. Payroll prints every other Tuesday. Quarterly sales tax prints four times a year and feels like a punch. Equipment leases, insurance premiums, payment-processor reserves — all hit on schedules that have nothing to do with when sales come in.

Lagging inflows. Credit card settlement runs two to three business days behind. Third-party delivery runs five to seven. Catering deposits hit early but final invoices clear 30–60 days later. Net-30 corporate accounts run 35–55 days in practice.

A typical full-service independent in the DMV is running 6–9 days of operating cash on hand. That is enough margin that one badly-timed quarterly tax payment or one slow week colliding with a payroll Tuesday can leave the operator wiring personal funds in to cover the gap. The forecast is what tells you that wire is coming three weeks before it does.

What a 13-week forecast actually looks like

It is a spreadsheet. Thirteen columns across (one per week, starting next Monday), six categories down. The categories are:

1. Opening cash position (the bank balance at the start of the week)
2. Cash in (deposits actually expected to clear in that week)
3. Cash out — fixed (rent, debt service, insurance, equipment leases, scheduled tax payments)
4. Cash out — payroll (gross payroll plus employer taxes on payroll weeks, zero on off-weeks)
5. Cash out — variable (food, beverage, supplies, utilities, marketing, repairs)
6. Closing cash position (#1 + #2 − #3 − #4 − #5)

Closing cash position from week one becomes opening cash position for week two. The forecast rolls forward weekly — every Monday you delete the column that just closed and add a new week 13 at the right.

The key is honesty about when cash moves, not when the P&L books the transaction. A net-30 vendor invoice for $8,400 received this Friday hits cash out four to five weeks from now, not today. Credit card sales from Saturday night land in cash in on Tuesday or Wednesday, not Sunday. Catering deposits hit immediately; the balance hits 30 days after the event.

The 13-week horizon is deliberate. It is long enough to see a quarterly sales tax payment coming three months out, and short enough that the line-by-line assumptions are still real instead of guesses.

Building the first version

Block four hours. Pull the last 13 weeks of bank statements and the last 13 weeks of the P&L. Then build the categories in this order.

Step 1: Fixed outflows

Fixed outflows are the easiest to model and the most painful to get wrong. List every recurring payment with a known amount and a known date:

• Rent and CAM
• Mortgage or equipment debt service
• Insurance premiums (monthly or quarterly)
• Equipment leases
• Software subscriptions over $200/month
• Quarterly estimated taxes
• Annual liquor license renewal
• Annual workers comp audit true-up
• Annual property tax (where applicable)

The annual ones are the killers. A $14,000 liquor license renewal that hits once a year wrecks the cash position of operators who do not see it coming. Put it on the 13-week forecast on the right week, every time it falls in the window.

Step 2: Payroll

Map the payroll calendar. Most DMV operators run biweekly payroll. Mark each payroll Tuesday. The gross payroll number is roughly your historical payroll plus employer taxes — for forecasting purposes, multiply scheduled payroll by 1.10 to capture the employer side (FICA, FUTA, SUTA, workers comp accrual). The forecasted payroll number on payroll weeks is large enough that the cash trough almost always falls on those Tuesdays.

Step 3: Variable outflows

Variable outflows are the noisiest line, but they have a stable pattern. Pull the last 13 weeks of variable spend and compute a weekly average. Then adjust for known anomalies — a slow January, a holiday-driven December, a vendor invoice batch that always lands on the 15th and the 30th. The point is not perfect precision; the point is the right order of magnitude.

Step 4: Cash in

Cash in is the line operators most often overestimate. Take the last 13 weeks of credit card deposits, third-party delivery deposits, catering deposits, and corporate AR collections. Map them to the week they actually cleared, not the week the sale happened. Most independent operators discover that their average "cash in" week is 6–9% lower than they thought, because they were mentally counting the sale and forgetting the settlement lag.

Step 5: Roll it forward

Once you have a working week one, copy the structure 12 more times. Adjust each week's known events (which Tuesdays are payroll Tuesdays, which week has rent, which week has the quarterly sales tax). The recurring stuff fills itself in.

The first build will take three to four hours. Every weekly update after that should take 20 minutes.

What the forecast tells you that the P&L cannot

Three things show up in the cash forecast that the P&L hides.

The trough week

Every 13-week window has one or two trough weeks — the weeks where opening cash, planned outflows, and the timing of inflows align badly. The forecast will tell you exactly which Tuesday is going to be the painful one, weeks before it arrives. With that warning you have options. You can negotiate a vendor invoice to net-45 instead of net-30. You can move a discretionary marketing spend two weeks later. You can pull a catering deposit forward by emailing the client.

Without the forecast, the trough week arrives unannounced and the only available move is panic.

The structural deficit

If you build a clean 13-week forecast and the closing cash position trends down week after week even when the trough weeks are flat, you have a structural problem. Your weekly cash burn exceeds your weekly cash generation. That can be hidden on a P&L for months by accruals, prepaid expenses, and slow-pay vendors absorbing the gap. On the cash forecast it shows up immediately.

A structural deficit is a different category of problem than a timing problem. It requires either revenue, pricing, or fixed-cost work — not a cash management tweak. The most common cause of a structural deficit in the DMV market is a labor or COGS line that has drifted 3+ points without a corresponding menu price adjustment. See prime cost benchmarks for what your prime cost should actually look like.

The investment window

The other thing the forecast surfaces is the opposite of a deficit — a stretch of weeks where cash is healthy and there is room for discretionary investment without risk. Without the forecast, operators tend to either spend reactively (when the bank looks fat one Friday) or never (because they are nervous all the time). A 13-week view lets you say "we have a clean window in weeks 6–10; that is when we replace the walk-in compressor or fund the marketing push."

What to do when the forecast shows a problem

The forecast surfaces a problem 6–10 weeks early. That is enough lead time for any of the following to work:

• Stretch a vendor from net-30 to net-45 on the largest two or three invoices. Most DMV vendors will agree if the conversation is direct and one-time.
• Pull a catering deposit forward. A 50% deposit policy with a clean email reminder gets cash in two weeks earlier on average.
• Adjust the payment schedule on an equipment lease by one billing cycle. Most lessors will do this once without penalty.
• Move a discretionary spend — marketing, repair, renovation — two to four weeks later.
• Draw on a line of credit intentionally, with a documented payback in week 8 or 9, rather than discovering you needed it the Friday before payroll.

The point of the forecast is optionality. Every one of these moves is easy when you have six weeks of warning. None of them work when you have 36 hours.

Operators who run a 13-week forecast for a year almost never call us in a panic. The forecast does not eliminate problems. It eliminates surprises. The two are very different.

Wiring this into your operating rhythm

Once the forecast exists, it has to be updated every Monday. The cadence:

1. Pull the prior week's actual closing cash from the bank
2. Replace the column that just closed with actuals
3. Add a new week 13 at the right
4. Adjust the next 4–6 weeks based on anything new you know (a vendor invoice that came in higher, a catering deposit that cleared early)
5. Read the bottom row — closing cash by week — and identify the trough

The Monday update is a 20-minute meeting between the operator and whoever owns the bookkeeping. It should be on the calendar. Operators who run it formally never miss a payroll. Operators who run it informally drift back to monthly within 90 days.

If you want help building the first version of this forecast against your actual operating data, book a discovery call. Bring three months of bank statements and your P&L. We will build the model on the call and leave you with a working 13-week view that updates in 20 minutes a week.

The forecast does not make a struggling restaurant profitable. It makes a healthy restaurant unkillable.