Most cash flow forecasts fail for a reason nobody writes about. Not the model. The meeting. A forecast is a claim about the future that depends on inputs held by people who do not report to finance, and if nobody owns the job of collecting those inputs on a fixed day, the model degrades quietly until it is wrong and nobody notices. The weekly cash meeting is the operating rhythm that stops that happening. Here is what it contains, who sits in it, and what it takes 30 minutes to cover.
What a weekly cash meeting is and who attends
A weekly cash meeting is a 30 minute standing session where the person who owns the forecast reconciles last week’s actuals against last week’s prediction, collects the three or four inputs that change the next 13 weeks, and confirms what decisions are now unblocked or newly blocked. It is not a financial review and it is not a board update.
Attendance is small on purpose. Four people is workable, six is the limit.
| Role | Brings | Leaves with |
|---|---|---|
| Forecast owner (finance lead, bookkeeper, or fractional CFO) | Updated 13 week view, variance against last week | Confirmed inputs, escalations |
| Sales or revenue lead | Deals expected to close and expected payment timing, not just pipeline value | Which deals need payment terms negotiated |
| Operations or delivery | What will be invoiceable this week, what has slipped | Which milestones to prioritise for billing |
| Owner or CEO | Pending commitments, decisions in flight | A go, wait, or conditional on each one |
| Collections or AR (if separate) | Status on accounts past 30 days | Named escalations for the week |
Note what the sales lead brings. Pipeline value is the wrong input. A $60,000 deal closing next week that pays on net 45 terms is a Q4 cash event, not a September one. Collecting expected payment date alongside expected close date is the single highest value change most teams can make to their forecast quality.
The 30 minute agenda
Fixed order, every week. The discipline is in not letting it become a general business discussion.
- Variance, five minutes. What did we predict for last week, what actually happened, and which line was wrong. Do not explain it yet. Just name it. Persistent bias in one line, always receipts or always payroll timing, is the signal worth chasing.
- Receipts, ten minutes. Walk the accounts due in the next two weeks. Confirm or move each expected date based on what happened when someone actually contacted the customer. An unchanged expected date on an invoice nobody has chased is a guess wearing a number.
- Outflows, five minutes. Confirm payroll dates, tax payments, debt service, and anything unusual. This is the least interesting section and the one that causes the most damage when skipped, because these are the payments you cannot defer.
- Decisions, eight minutes. Every pending commitment gets a go, wait, or conditional. Conditional means naming the condition, for example proceed if the two largest open invoices clear by the 20th. Write the condition down.
- Lowest point, two minutes. State the lowest projected balance in the next 13 weeks and the week it occurs. If that number moved materially since last week, say why. This is the one number everyone in the room should remember on the way out.
If the meeting runs long, the problem is almost always that someone is explaining a variance instead of naming it. Explanation happens after the meeting, with two people, not five.
The inputs that actually need chasing
Forecasts do not usually break on the accounting side. The ledger updates itself. They break where finance depends on someone else remembering to tell them something.
| Input | Who holds it | What happens when it is missed |
|---|---|---|
| Expected payment date on a closing deal | Sales | Cash lands 30 to 60 days later than modelled |
| Milestone slipped so invoice not raised | Delivery | A receipt disappears from the week with no warning |
| Start date agreed with a new hire | Owner or hiring manager | Payroll steps up a period early |
| Supplier terms changed at renewal | Operations or procurement | Outflow timing shifts across a month boundary |
| Deposit taken that is not yet earned revenue | Sales or admin | Cash looks available when it is a liability |
Every row is a person, not a system. That is why the meeting exists. Automating the forecast, which a tool like Noya does well, removes the arithmetic but not the requirement to ask a human what they know that the ledger does not.
Running it across a client book
Accountants and fractional CFOs face a version of this problem at a different scale. You cannot sit in a weekly meeting for 25 clients. What works instead is a tiered cadence: weekly for clients in or near a squeeze, fortnightly for clients with a decision pending, monthly for stable ones, with an exception alert that pulls any client into the weekly tier when their lowest projected balance drops below an agreed threshold.
That structure only works if the underlying view is standardised across clients, which is the practical argument for a single cash flow health view rather than 25 spreadsheets. Our post on how fractional CFOs manage more clients covers the tiering in more detail, and the fractional CFO page covers the workflow side.
Give the Meeting Something Worth Discussing
A weekly cash meeting is only as good as the forecast it opens with. If yours takes two hours to rebuild each week, the meeting will not survive the month. Connect your QuickBooks or Xero account and get a rolling forecast that refreshes itself in under 60 seconds.
Create your free Finoya account and walk into Monday knowing your lowest projected balance.