Blog
Spreadsheet cash forecasting vs two-sided cash forecast: why a single AR column hides the week that matters
For most SaaS finance teams the cash forecast lives in one of two shapes: a one-column AR ledger on a spreadsheet, reconciled by hand into a net position every Monday — or the two-sided weekly inflow-vs-outflow methodology on /forecast, where scheduled AP and scheduled AR share one timeline. Here is what the two look like side by side, and why the single AR column quietly hides the week the CFO needs a heads-up on.
How single-sided spreadsheet forecasting works
The default shape of spreadsheet cash forecasting is a one-column AR ledger. One column, one date, one dollar amount per row — every scheduled receipt the AR system knows about, sorted ascending. The forecast is whatever the controller writes next to it: a manual running net that lives as a comment, a sum cell at the bottom, or a pivot table rebuilt before every board meeting.
The math does not require a join, and that is precisely the point. AR receipts enter on the day they are expected to clear, payroll and vendor runs enter on whatever spreadsheet the controller happens to have open, and the net cash position is the human subtraction of one column from another. On a quiet week the two sides line up and the number survives. On the weeks where they do not, the forecast was always a guess — printed confidently and reconciled badly.
What the two-sided weekly methodology on /forecast does differently
The /forecast methodology is a two-sided weekly rollup of AP and AR against the same timeline. Every scheduled inflow — autopay receipts, renewals, mid-tier renewals, new-MRR batches — sits next to every scheduled outflow — biweekly payroll, vendor AP runs, payment processor fees, contractor payouts — and the net weekly position is computed against the running ledger rather than subtotaled from one column into another.
The structure reads as: two-sided weekly inflow vs outflow, lined up week by week, with the net column sign-aware so a negative week reads as red on the page before the wires clear. The CFO stops being the human middleware between two dashboards; the controller stops pasting a pivot table into the board deck; and the net cash position is an output of the system, not a manual subtraction that drifts.
Before / after: a few weeks of cash movement
The fastest way to see the gap is to lay them next to each other. Same weeks, same real cash movement, two different views. The AR column is the one a spreadsheet has been printing all quarter; the AP payouts column is the one the controller has been subtracting by hand; the Net weekly position is what the two-sided view computes for free.
| Week | AR receipts (one column) | AP payouts | Net weekly position |
|---|
| Wk 1 (Jul 6 – 12) | $95,000 | $148,000 | |
| Wk 2 (Jul 13 – 19) | $215,000 | $158,000 | |
| Wk 3 (Jul 20 – 26) | $130,000 | $165,000 | |
| Wk 4 (Jul 27 – Aug 2) | $240,000 | $148,000 | |
Read the Wk 3 row through a single-sided AR column and the week looks healthy — $130,000 worth of receipts, no question about it. Read the same row through a two-sided weekly view and Wk 3 closes at −$35K: the scheduled mid-tier renewals land, the same-week vendor run and contractor payouts leave, and the net is the number the CFO actually needs. The single-sided view does not see both sides of the week; the two-sided view does, and that is the entire difference between a forecast that quietly hides a shortfall and one that surfaces it.
Why the failure is structural, not a tooling bug
The Wk 3 shortfall in the table above is not a bug in the spreadsheet. It is the spreadsheet telling the truth about what it knows. A one-column AR ledger is built to answer one half of the question: when money is expected to come in. It has no place to write what is expected to go out on the same week, so it cannot compute the join. Even a perfectly maintained spreadsheet forecast cannot see both sides of the week without pulling the AP schedule into the same tab — and the moment the AP schedule becomes a separate tab, the controller is the join, and the controller’s Monday morning is already spoken for.
The two-sided weekly methodology exists precisely to remove the human join. AP and AR share a model, share a timeline, share a sign-aware net column, and the forecast becomes an output of the ledger rather than a function of whose tab is open at 9am. The shift is structural — the failure was never a missing formula, it was a missing column. Two-sided forecasting adds the second column back in, and the week that matters stops disappearing into a pivot table.