The meeting exists to decide, not to recap
A 13-week cash meeting that does not produce a kill-or-keep decision is a status call. Status calls burn 45 minutes and still miss payroll.
The test is simple. When the meeting ends, did someone keep a spend, kill a spend, pull a collection, or change a timing decision? If the only output is we are watching it, you did not have a cash meeting. You had a reading of the bank balance.
I run these meetings as a fractional COO and CFO. The companies that stay out of the ditch treat cash like a weekly operating rhythm, not a monthly surprise. The 13-week forecast is the instrument. The meeting is where the instrument gets used. Most teams build the spreadsheet and skip the meeting design. Then they are shocked when the model is stale and the founder still asks, are we okay?
You already know cash is not profit. That gap is one of the hidden financial risks that sink growing companies. This article is the weekly practice that catches the gap while you can still act.
One owner, not a committee
Name one owner. Not finance. Not the leadership team. One person.
The owner builds the 13-week view, sends the packet, and presents actual versus forecast versus variance. Everyone else arrives having read. If two people own it, nobody updates week seven when a customer slips.
In a company with a controller, the controller owns the packet. The CEO owns the collection calls and the spend kills the numbers imply. In a company without a controller, a fractional CFO or the operator who already owns AP and AR owns the packet. Do not give it to the founder by default. Founders edit hope into week six.
The owner is allowed to be wrong. The owner is not allowed to be late. A packet that lands at 8:50 for a 9:00 meeting is an ambush. People cannot decide on numbers they just saw.
Put the name on the accountability chart next to the cash floor. If your operating system uses a scorecard, cash on hand, DSO, and the week-seven gap belong on it. The meeting then works the issues those numbers raise. That is the same discipline as installing a business operating system, applied to one number that can end you.
The packet that has to land 24 hours early
Send the packet by the same hour every week, at least 24 hours before the meeting. PDF or a locked sheet. Not a live tab someone is still typing.
Five pages. Not fifteen.
Page 1 is the cash floor and the next 13 weeks, inflows and outflows by week, with a running balance. Circle the first week the balance would break the floor. If no week breaks it, say so in one line.
Page 2 is last week. Actual versus forecast versus variance for receipts, disbursements, and ending cash. Then one sentence of cause. Then one sentence of risk. Receipts were $312K versus a $410K forecast, a $98K miss, because two invoices totaling $120K slipped from Friday to Tuesday. The risk is week 5 now sits $70K under the floor if those two customers pay on their new verbal date.
Page 3 is the aging. Top ten past-due accounts, dollars, promised date, named collector. If a promise is broken twice, the next move is on the page: withhold work, convert to ACH, or send to counsel.
Page 4 is the payables and commitments. What must be paid. What can move. What is already late. Include payroll, tax, rent, and any vendor who can stop a job.
Page 5 is the kill-or-keep list. Discretionary spend, contractor lines, software, inventory buys, capex POs, and owner draws that have not yet gone out. Each row is keep, kill, or delay, with a dollar amount and a date.
If a page is missing, cancel the meeting and send the packet. Meeting without paper trains the team to wing it.
The agenda is kill or keep
Hold the meeting to 30 minutes. Same day, same time, every week. Standing attendees are the owner, the CEO, and whoever can actually stop a payment or make a collection call. Sales attends when receipts missed. Ops attends when a job is about to be starved of materials.
| Minute | Item | Decision required |
|---|---|---|
| 0-5 | Floor and 13-week view | Confirm the first week that breaks the floor, or confirm none does |
| 5-12 | Last week actual vs forecast vs variance | Accept the cause, or send it back for a real cause |
| 12-20 | Aging and collections | One action per broken promise |
| 20-26 | Payables and the kill-or-keep list | Keep, kill, or delay, with dollars |
| 26-30 | Recast week 1-13 | Owner updates the model the same day |
No new slides. If it is not in the packet, it waits a week unless payroll is at risk today.
No narrative without a variance. We had a good week is not an input. Receipts beat forecast by $40K because three prepayments landed is an input.
No parking-lot for money. Delay is allowed once, with a date. A second delay is a keep you will not admit.
I will make the call. If ending cash is inside two weeks of the floor, freeze discretionary spend in the meeting, not after a side chat. If you think that is too sharp, bring a customer check that has cleared. Until it has cleared, it is not cash.
A worked example: the recap that almost missed payroll
An $8.4M distributor, 31 people, seasonal receipts. Budgeted cash floor was $450K, about six weeks of fixed cost. Actual cash on a Monday in April was $610K, so the room felt fine.
The weekly meeting ran 40 minutes. The controller shared a bank screenshot. The founder asked if anything looked weird. The sales lead said a big customer would pay this week. No packet. No variance. No kill list. They adjourned.
The 13-week rebuild that afternoon told a different story. Last week receipts were $188K versus a $340K forecast, a $152K miss. Cause was one customer at 22% of revenue pushing a $140K invoice from net 30 to net 60. Week 6 payroll plus a $220K inventory receipt put the running balance at $290K, $160K under the floor. Week 8, without a change, went negative.
We installed the meeting the next day. Owner was the controller. Packet due Tuesdays at noon. Meeting Wednesdays at 9:00, 30 minutes. Page 5 that first week had four rows.
A $22K/month contractor bench for overflow warehouse labor. Kill. A $48K software renewal set to autopay in 11 days. Delay 30 days and cut unused seats. A $180K inventory PO for a slow SKU. Kill. Owner draw of $25K scheduled for week 3. Delay to week 10.
Collections: CEO called the 22% customer the same morning. Got $70K by Friday and a dated ACH plan for the rest. Top-customer revenue share went on the scorecard.
Recast ending cash in week 6 sat at $505K. Floor held. The old recap would have discovered the hole when the inventory receipt hit.
What had to go right was the CEO making the call and not reopening the killed PO. We treated a second customer slip as a credit-hold trigger, written in the packet.
This is the same cash discipline that shows up when you inflation-proof a business. Inflation makes customers pay slower while your payables grow. The 13-week meeting is how you see that collision in week numbers instead of in a returned ACH.
Common mistakes
Letting the founder own the model. Founders put verbal promises into week 2 as if they had cleared. The packet owner should be the person who can be fired for a wrong cash number.
Sending the live workbook. People scroll and argue about tabs. Send a packet.
Inviting eight spectators. Three to five people who can act produces decisions. Recap does not.
Treating the floor as a wish. If you have not named the floor in dollars, you cannot know when you have broken it. Three months of fixed cost is a sound starting point.
Skipping a week because cash feels high. The week you skip is the week a customer stretches terms.
Forecasting revenue instead of receipts. Booked sales in week 3 are not cash in week 3.
FAQ
Who should own the 13-week cash meeting in a small company? One named owner of the packet, usually the controller or the operator who already runs AP and AR. The CEO owns the collection calls and the spend kills. Do not default ownership to the founder. Founders edit hope into the near weeks. If you do not have a controller yet, a fractional CFO can own the packet until the seat exists.
What belongs in the packet, and when must it go out? Five pages, 24 hours before the meeting: the 13-week running balance, last week actual versus forecast versus variance, the aging with named collectors, the payables that can move, and the kill-or-keep list. If the packet is late, cancel the meeting. Do not sit down to discover the numbers live.
How is this different from a monthly finance review? The monthly review explains the P&L that already happened. The 13-week cash meeting decides the next 90 days of receipts, disbursements, and discretionary spend. One is after-the-fact. The other is a kill-or-keep operating rhythm. You need both. Only one of them can stop a hole in week seven.
