Money brain · Forecast · 11 Nov 2026 · 4 min read
A 90-day cash forecast the board will trust
Boards don’t lose faith in forecasts because they’re wrong. They lose faith because nobody can say how wrong.
Every forecast is wrong. The ones a board trusts are wrong by an amount everyone knows in advance. Here is what that takes.
Show a range, not a line
A single number on day 60 looks precise and isn’t. A range, say between $780K and $1.1M, tells the board what you actually know. It also tells them when to worry: when the bottom of the range gets close to the buffer.
Rebuild it often
A forecast built on the first Monday of the month is stale by Wednesday. Rebuilding it weekly is good. Rebuilding it every morning from bank and ERP data is better, because then a late customer payment shows up the day it’s late, not at month end.
Name the assumptions
- Which customers pay late, and by how many days, from their own history.
- Which costs are fixed and which move with revenue.
- Which big payments are certain and which still need approval.
When the forecast changes, the board should be able to see which assumption moved.
Keep score
Compare what you forecast four weeks ago with what happened. Report the gap every month. A forecast whose error is shrinking earns trust quickly, and a smaller error means a smaller buffer. That is cash you can put to work.
How Kairos does it
Kairos rebuilds the 90-day forecast every morning from your bank feeds and ERP, shows it as a range, and explains what changed since yesterday. It runs on your servers. Your team reviews it in about ten minutes.
Want this worked out on your own numbers?