AI Agents for SAP
Business One FP&A
Large company FP&A explains profit. At this size the question that keeps somebody awake is whether there is enough money on the fifteenth.
- At this size the cash forecast matters more than the profit forecast, and five of its six inputs are already in Business One exactly.
- Every cash forecast built on stated payment terms is wrong in the same direction. Customers pay how they pay and the receipt history already records it per customer.
- In a worked 13 week view, the forecast on terms showed a comfortable week 6 and the forecast on behaviour showed it 41,000 short, four weeks in advance.
- Committed purchase orders not yet invoiced are the line most small business cash forecasts omit entirely.
- A customer drifting from day 30 to day 50 over six months is the earliest credit signal you will get, and an average collection period across the book hides it completely.
AI agents for SAP Business One FP&A build a 13 week cash view from open payables and receivables priced on how customers actually pay rather than on stated terms, include committed orders not yet invoiced, and explain a profit movement with the documents attached. ChatFin rebuilds it weekly at no cost and posts nothing.
The reason to start with cash rather than profit is that at this size it is the question being asked, and it is more mechanical than most teams assume. Five of the six inputs are already in the system exactly. The sixth, new sales, is a judgement that belongs to a person. The correction that matters is replacing payment terms with observed behaviour, which usually moves the forecast more than any assumption argued about in a review.
At this size the forecast that matters is cash
Large company FP&A is about explaining profit. On Business One the question that actually keeps somebody awake is whether there is enough money on the fifteenth. Those need different work, and building the second is far more mechanical than most teams assume.
| Input | Where it lives | How reliable |
|---|---|---|
| Open supplier invoices by due date | Business One, exactly | Very. This is a fact |
| Open customer invoices by due date | Business One, exactly | Facts, but the dates lie |
| Observed customer payment behaviour | Receipt history | The important correction |
| Recurring costs | Payroll, rent, subscriptions | Known and dateable |
| Committed orders not yet invoiced | Purchase orders | Known, often ignored |
| New sales | Nowhere in finance | A judgement. Not the agent's |
Five of the six are already in the system. A 13 week cash view built from them and corrected for how customers actually pay is more useful than a profit forecast, and it can be rebuilt every week at no cost.
ChatFin prices receipts on observed payment behaviour rather than stated terms. The gap between the two is usually the most useful line in the whole forecast.
Stated terms are the wrong input
Every cash forecast built on payment terms is wrong in the same direction. Customers pay how they pay, and the ledger already records it.
ChatFin reports each customer's observed payment habit alongside their stated terms, so the gap is visible rather than buried inside the forecast.
A worked 13 week cash view
Built on terms, week 6 shows a comfortable position. Built on observed behaviour it does not.
| Line | On stated terms | On observed behaviour | Why it moved |
|---|---|---|---|
| Receipts, weeks 1 to 4 | 412,000 | 331,000 | Three large customers habitually pay late |
| Receipts, weeks 5 to 8 | 388,000 | 402,000 | The week 4 slippage arrives here |
| Payments, weeks 1 to 8 | 640,000 | 640,000 | Unchanged. You pay when you say you will |
| Committed orders not yet invoiced | not included | 118,000 | Purchase orders already placed |
| Lowest weekly position | +96,000 | negative 41,000 | Week 6 |
The forecast on terms says there is no problem. The forecast on behaviour says week 6 is short, and it says it four weeks early, which is enough time to do something ordinary about it rather than something expensive.
Explaining a movement without a finance team
The profit question still gets asked, usually by an owner rather than a board, and the answer is expected quickly.
Split the movement into components that add back
Volume, price, mix, timing and a residual. If it does not reconcile to the reported movement it is not an explanation.
Separate accounting causes from trading causes
A released accrual or a reclass is not a trading event, and presenting it as one is how an owner draws the wrong conclusion.
Name the documents
At this size the owner frequently knows the specific customer or supplier, so naming them makes the answer immediately checkable.
Stop where the data stops
Where the cause is not in the ledger, report what moved and say the rest is not knowable from the system.
ChatFin drafts the explanation with the documents attached. At this size the person reading it usually recognises the names, which makes checking it a thirty second job rather than an afternoon.
Where this underperforms on Business One
Three, honestly.
Deeper in SAP Business One FP&A
Three use cases sit under this hub, in build and shown rather than linked.
Worked on this page
Natural language querying of finance data · Financial forecasting · Financial modeling
Covered as sections here
Cash forecasting · Variance analysis · Budget vs actual · Management commentary · Customer payment behaviour
Questions Business One teams ask first
Why is a cash forecast built on payment terms wrong?
Because customers pay how they pay, and the ledger already records it. Most have a stable habit, day 45 against 30 day terms for years, which is predictable rather than risky and belongs in the forecast as day 45. The habit differs per customer, so an average collection period across the book hides it and produces a forecast that is never right for any particular week. ChatFin prices receipts on observed behaviour.
What do most small business cash forecasts leave out?
Committed purchase orders not yet invoiced. The order is placed, the money is going to leave, and nothing in a receivables and payables view shows it. In a worked 13 week example that line was 118,000. Adding it, and correcting receipts for observed payment behaviour, turned a comfortable week 6 into a shortfall of 41,000 visible four weeks in advance.
What if the business has no budget?
Many Business One businesses do not, and budget versus actual simply does not exist as a report. Prior period comparison, trend and the cash view all still work and none of them needs a budget. ChatFin will say plainly that budget versus actual is unavailable rather than inventing a baseline, and the 13 week cash view is the better place to start anyway.
Week 6 is short, and you know four weeks early.
ChatFin builds the 13 week cash view from open payables, receivables and committed orders, and prices receipts on how each customer actually pays rather than on their terms.
It explains profit movements with the documents attached, separates accounting causes from trading ones, and stops where the ledger stops.
Bring your open ledger. The cash view takes an hour.
Book a demoAll images are served from the Pexels content network. No third party product screenshots, vendor logos or licensed media are used on this page.
Bring your open ledger.
We will build the 13 week cash view through ChatFin on observed payment behaviour and show you the difference against the same view on stated terms.