Here is what to do when a cash forecast keeps missing: compare each forecast week against the bank activity and ask whether the money arrived late or in a different amount. In many growing companies, repeated misses come from timing gaps rather than from wrong revenue totals. The total may be right while the calendar is wrong, and that difference decides which fix you need.
Start by separating timing gaps from amount gaps
A forecast that keeps missing is a symptom, not a diagnosis. The forecast says cash will sit at a certain level on a certain date, and the bank statement says something else. Before you change any assumption, decide which kind of difference you are looking at. A timing gap means the right amount moved on a different day. An amount gap means the money that moved was larger or smaller than the plan assumed, sometimes for reasons unrelated to the calendar.
The distinction matters because the fixes do not overlap. If you treat a timing gap as an amount gap, you may raise a revenue assumption that was never wrong, and the forecast will look optimistic each time a payment lands late. If you treat an amount gap as a timing gap, you may spend months moving dates while a real drop in sales goes unexamined. In both cases the same miss returns, and the team starts to distrust the model.
A useful first check is plain. Take the most recent forecast cycles and, for each missed line, write down the expected date, the actual date, the expected amount and the actual amount. If only the dates differ, you have a timing problem. If the dates match and the amounts differ, look at pricing, volume, credits, refunds or customers who paid less than they were invoiced. Many teams skip this step and argue about the total, which tells them the size of the miss but not its cause.
The person who prepares this comparison should ideally not be the person who built the original forecast. The builder remembers why each assumption was made and tends to read the gap through that memory. Someone who reads the bank activity without that history sees the movement more plainly. This does not require a new hire. A bookkeeper, a controller or an outside advisor can review the bank feed once each cycle and write the comparison in a short working session.
Keep the output short. One page that lists each missed line, its type of gap and the owner of the correction is enough to start. A long analysis that nobody reads before the next forecast is due does not help anyone. A report is useful only if people can understand it in time to act, so build the first version for the next decision, not for an archive.
Once a gap is named, write one plain sentence that says what happened, for example that the receipts from one customer arrived a week after the forecast date and the amount matched. That sentence becomes the agenda for the review meeting. It also gives the next reader something to check against the bank, which is faster than reconstructing the story from memory.
How to read a missed 13-week cash forecast week by week
A 13-week cash forecast is a rolling view of expected receipts and payments. Each week the first column closes, the actual numbers replace the estimate, and a new week is added at the end. It is most useful when actual bank activity is compared with it at the same level of detail. The common mistake is to compare only the ending balance. Two very different problems can produce the same ending balance, and the forecast looks equally wrong in both.
Start with the receipts. Each expected customer payment should carry a date, an amount and an invoice reference. When a receipt is missing from the bank in the expected week, check three things: whether the invoice went out late, whether the customer paid on terms but the payment was applied to a different invoice, and whether the invoice was ever issued at all. Each looks like the same miss in the forecast, yet each needs a different correction.
Then review the disbursements. Forecasts often use the invoice date or due date image_comment thanks for laying out how fractional cfo support can fit into a growing business. reliable reporting for the board is easier to build early. looking forward to a follow-up on this topic. vendor bills, while the company actually pays through a payment run on set days. A bill due on a Friday may leave the account on the following Tuesday, which moves cash from one week into another. The forecast then shows a shortfall that never happens, or a cushion that disappears when the run finally posts.
Here is a hypothetical example, built only to show the mechanics. Suppose a forecast expects a customer payment in one week and an equal vendor payment in the next. The customer pays on schedule, but the vendor payment is moved earlier because the payment run changed. Both lines are right in total, yet the balance dips in the first week and recovers in the second, which is the reverse of what the forecast shows. A reader who looks only at week-end balances would miss this completely.
Summarize the reason for each miss in a small table. The table below is a template you can copy into a spreadsheet. It describes the kinds of cause to look for, not a finding about any particular company.
| Line type | Typical symptom | Likely cause | Correction |
|---|---|---|---|
| Customer receipts | Arrives one or more weeks late | Invoice sent late, or payment matched to the wrong invoice | Confirm invoice dates and terms with the person who issues invoices |
| Vendor payments | Leaves in an earlier or later week | Payment run date differs from the model | Align the forecast with the actual payment calendar |
| Payroll | Lands in a different week | Pay date or tax deposit date not updated | Load the full payroll and tax calendar, and check it at every cycle |
| Financing draws | Missing or early | Lender timing assumed, not confirmed | Confirm the funding date in writing before entering it |
Patterns in this table matter more than any single miss. If the same line type misses in every cycle, the fix is a process change, such as moving the payment run date in the model or changing invoice timing, not a one-time adjustment to the balance. A single late receipt is noise. The same late receipt every month is information.
Why revenue recognition can mislead a cash forecast
Revenue and cash run on different clocks. Under ASC 606, revenue is recognized as the company satisfies its performance obligations, which can happen before, during or after the customer pays. A forecast built directly from revenue reports therefore shows cash arriving when the work is delivered, even when the customer is on terms that delay payment for weeks. That gap is not an error in the revenue report. It is a mismatch between the question the report answers and the question the forecast asks.
This is where accrual and cash diverge. An accrual records an economic event before money moves, which is correct for financial statements and misleading when copied straight into weekly lines. Many growing companies build forecasts from monthly revenue reports because those reports already exist and the team knows them. The shortcut is understandable, but it hides the collection timing the forecast needs, and it often hides it in the same direction each month.
Deferred revenue creates a second trap. When a customer pays in advance, cash arrives before the revenue is earned. A forecast that looks only at revenue may miss the cash when it actually arrives, then show it again when revenue is recognized, which overstates the later weeks. Check whether the forecast is built from billings, collections or revenue, and label the line so nobody assumes the wrong thing. A label takes a minute and prevents a long argument.
A question a board member would ask is simple: which number is the forecast built from, and does that number tell you when money moves? If the answer is revenue, you have a bridge to build. The bridge starts from billings, adjusts for changes in receivables and deferred revenue, and produces expected collections by week. You do not need to rebuild the whole model to add it. A single tab with a clear starting figure and a few adjusting lines is often enough.
Do not ask the forecast to carry full accounting detail. Ask it to show when cash is expected. The people who own revenue reporting and the people who own cash forecasting should agree on the bridge, and the bridge should be written down so a new reviewer can follow it without a meeting. If the bridge lives only in one person's head, the forecast will miss again the moment that person is busy or away.
If the company later shares its financials with lenders or investors, both views will be requested: the accrual view for statements and the cash view for liquidity. Keep them connected through the bridge. A reader who can move from one to the other trusts both more, and a disconnect between them is one of the first things a careful reader will question.
Collections: when customers pay later than the forecast assumes
The symptom is easy to spot. Receipts arrive later than the forecast says, cycle after cycle, from the same group of customers. The cause is usually an assumption that payment happens on the due date, when those customers actually pay on a different schedule. The correction is to change the assumption, but only after you confirm the pattern in the records, not in someone's recollection.
Begin with the aging report. It groups open invoices by how long they have been outstanding. If the forecast assumes payment on the due date and the aging shows a steady share of invoices moving into the next bucket each month, the timing assumption is wrong. The forecast will keep missing until the assumption matches what customers actually do.
Then ask why customers pay late. Sometimes the invoice went out days after delivery. Sometimes a purchase order number was missing, so the customer's accounts payable team could not match the invoice to an approved order. Sometimes the customer's own approval chain is slow, or the payment waits for a monthly signature. Each cause has a different correction, and only the first is fully in the company's control.
The corrections can be practical. Invoice on the day the work is accepted, not at month end. Put the purchase order number and the payment terms at the top of the invoice. Confirm, before work starts, who approves payment on the customer side. For larger projects, deposits or milestone billing can change the timing of cash without changing the price, though they require the customer to agree. Test each change with a small group of customers before it becomes a rule.

Use history rather than hope. Forecast each customer's receipts from its actual payment behavior over recent cycles, not only from the stated terms. This is a judgment call, and the forecast note should say so. A customer who always pays a few weeks after the due date should not appear in the forecast on the due date, even if the contract says otherwise. The contract tells you what was agreed, and the bank tells you what happens.
Set a rule for when the assumption changes. If a customer's late pattern holds for consecutive cycles, move the expected date in the forecast. Do not wait for a quarterly review to do it. Each week the forecast carries the wrong date, the decisions built on it, such as whether to draw on a credit line or delay a purchase, are made with the wrong information.
Payables and payroll dates that run ahead of the forecast
Payables tend to surprise forecasts in the other direction. Cash leaves earlier than the model shows, and the balance dips at a point where the plan showed room. The usual cause is a gap between the date a bill is approved and the date it is paid, or between the date a card statement closes and the date the card is charged. The forecast may use the invoice date, while the accounts payable team pays on a weekly run.
Payroll is more regular, which makes it easier to model and easier to get wrong. A forecast can be right about the amount and still wrong about the week if the pay calendar shifts across a holiday, or if a bank processing day moves the debit by one business day. Tax deposits follow schedules set by federal and state agencies, and the payroll provider usually handles them. The forecast still needs the actual dates, not the typical ones.
A practical step is to build a payment calendar that lists each recurring outflow, its owner, its trigger, and the day it actually leaves the bank. Recurring items include rent, software subscriptions, insurance, loan payments and payroll taxes. Once the calendar exists, the forecast can pull dates from it rather than from memory. The calendar also shows who to ask when a date changes.
Approval thresholds deserve a look too. If bills above a certain size need an extra signature, the payment date depends on who is available that week. The forecast cannot know the signer's calendar, but it can assume a delay for large items and note that assumption. A conservative date for large payments is often more useful than an optimistic one, because a payment that goes out later than planned is usually a small surprise, while one that goes out earlier can force a draw on a credit line.
Check the calendar against the bank for at least one full cycle before relying on it. Compare each scheduled outflow with its actual posting date. Any line that posted on a different day becomes a correction to the calendar, and the reason goes next to it. After one clean cycle, the calendar is ready for the forecast to read from.
The payment run itself should be a documented decision. If the company pays vendors weekly on a particular day, that day belongs in the forecast. If the team moves the run during a busy close, the forecast should reflect the move in advance, not after the balance has already dipped. A calendar that is planned, rather than discovered, is what stops the same miss from coming back.
Who owns the forecast and who can update it when facts change
A forecast that keeps missing often has an ownership problem. The model lives in one person's spreadsheet, the bank data arrives in another person's inbox, Image_Comment Thanks for laying out how fractional CFO support can fit into a growing business. Reliable reporting for the board is easier to build early. Looking forward to a follow-up on this topic. and the sales team's expected receipts come from a pipeline report that nobody reconciles. Each person has part of the answer and nobody has the whole. The fix begins with a plain question: who is accountable for the accuracy of each line, and who can change it when the facts change?
Write the answer down as a simple responsibility list. Assign an owner for receipts, disbursements, payroll, financing and the bridge to the accrual numbers. Each owner should know which bank account, system or report feeds their line, and how often they check it. A forecast with named owners is easier to correct than one with a shared file that everyone edits and nobody maintains.
Coordination with the existing CPA or controller deserves particular attention. Those people hold the books, close the periods and sign off on the numbers that external parties see. A strategic finance review that builds a separate view of cash without talking to them can create two versions of the truth. Agree early on which system is the source of record for each figure, and how differences between the two views are resolved.
Some companies bring in fractional or interim financial leadership to own this coordination without adding a full-time executive. That is one option among several. It can be useful when the company is growing faster than its finance function, and when the owner of the forecast needs both the time and the authority to challenge assumptions. Whether it fits depends on the company's size, systems, staffing and goals, and no single arrangement suits every business.
Set a meeting cadence that matches the decisions the forecast supports. A weekly review of the next few weeks is often enough for operating decisions, while a longer monthly review covers the full horizon and the bridge to financial statements. Without a cadence, updates happen only when someone notices a problem, which is usually after the balance has already moved.
Finally, write a short rule about who can change an assumption without a meeting. Small date changes supported by bank evidence might be made by the owner alone, with a note. Changes to revenue, large payment timing or financing assumptions should go to the accountable executive. The rule keeps the model moving without letting it drift away from the people who must act on it.
Systems and reporting deadlines that hide the gap
Systems can hide timing gaps. A financial system may post a transaction on the date it was entered rather than the date the bank cleared it. Bank feeds can lag, and a manual import may skip a day. An ERP is a useful place to hold this data, but only if its dates are defined the same way the forecast defines them. If the team argues about whether a transaction belongs to May or June, check the date field before checking the amount.
Reconciliation timing matters as well. If bank reconciliations happen monthly, the forecast team may not see a posting problem until weeks after it occurred. A weekly bank reconciliation of the operating accounts is a modest change that shortens the time between a miss and its explanation. It does not need to be complete. It needs to be current enough to show what moved in the last few days.
Reporting deadlines create their own pressure. A lender may ask for a liquidity figure on a specific date, and a board package may be due before the monthly close is finished. When the close is late, Image_Comment Thanks for laying out how fractional CFO support can fit into a growing business. Reliable reporting for the board is easier to build early. Looking forward to a follow-up on this topic. the team may use the forecast as a stand-in for actuals, which mixes estimates with facts. Label each figure as actual, estimated or forecast so the reader knows which is which.
Cutoff dates are a common source of confusion. If the close uses one cutoff and the bank uses another, a payment made on the last day of the month may belong to different periods in each view. The forecast does not need to resolve every cutoff question, but it should state the convention it uses, and that convention should match the one in the statements the company shares.
Use the system's own reports for the checks you already need. An aging report, a payment file or a bank exception list can answer many questions the forecast raises. Ask the system for the report rather than building a manual version each cycle. Manual versions drift, and a drifted report can look authoritative while it is wrong.
Upgrading a system is a larger decision and should follow the cash process, not lead it. Fix the definitions, owners and cadence first. Then decide whether a new ERP module, a better bank integration or Image_Comment Thanks for laying out how fractional CFO support can fit into a growing business. Reliable reporting for the board is easier to build early. Looking forward to a follow-up on this Topic. a reporting tool would remove a repeated step. A tool that automates a process nobody understands tends to automate the same miss faster.
What to verify before the next forecast goes out
Before the next forecast leaves the finance team, a short check can catch the causes covered above. The items below do not replace a full review. They confirm that the common timing problems have a named owner, a date and a source.
- Confirm that every receipt line has an invoice reference, a due date and an expected payment date based on recent customer behavior.
- Compare vendor payment dates in the model with the payment run calendar and the most recent actual postings.
- Check that payroll, tax deposit and financing dates come from the calendar or from written confirmation, not from memory.
- Confirm that the bridge from billings or revenue to expected collections is documented and that one person can explain each adjustment.
- Verify that every missed line from the last cycle has a named cause, a correction and an owner, and that the CPA or controller has reviewed the differences.
If any item cannot be checked in an hour, that is information too. It usually means the process depends on someone's memory or on a file nobody else can open. Record it as a risk, not as a footnote, because it is the most likely place for the next miss to start.
Keep the first version modest. A forecast that is accurate about timing for a few lines is more useful than one that covers every line and is wrong about the dates. Add detail as the team earns confidence in each part, and let each correction stand on the bank evidence that supports it.
Avoid a common trap at this stage: adding a new assumption to explain a miss you have not yet understood. A new assumption can make the next forecast look better without making it more correct. If the cause is still unclear, leave the line as a flagged estimate and keep investigating until the bank confirms what happened.
A forecast should help a decision, not decorate a meeting. Before it goes out, ask the person who will act on it which decision it supports and which date matters most to that decision. Then check that date first. A timing gap caught before the decision is far more useful than an accurate total that arrives after the money has already moved.