Why the spreadsheet budget stops working
The problem is not that Excel is bad for planning — for planning it is excellent. The problem is what happens next.
The budget is built in November, agreed, and saved as a file. In February someone needs a comparison against actuals: they export the ledger, reconcile it against the budget by hand, find a variance and spend half a day establishing whether it is a real movement or a difference in method. By April the procedure has been simplified. By June it has stopped.
The cause is always the same: plan and actual live in different places and are reconciled manually. Once the budget sits in the same system as the postings, comparison stops being work and becomes a report.
A practical readiness test. If answering "how are we tracking against budget" takes more than ten minutes, the budget is being kept in the wrong place.
Three kinds of budget
In Business Central these are three separate mechanisms, and they get conflated regularly.
| Type | What it plans | Who needs it |
|---|---|---|
| General ledger budget | Amounts by account, broken down by dimension | The main one. Income, expenditure, result |
| Sales and purchase budget | Quantities and values by item and customer | Distribution, for purchase planning |
| Job budget | Cost and revenue within a specific project | Services, construction, implementation work |
Start with the first. A general ledger budget answers "what do we expect to spend and earn" and gives budget-versus-actual in the familiar reporting structure. The others are additions that not everyone needs.
Dimensions decide everything
This is the central idea, and it is the one that gets skipped.
The budget has to be held in the same breakdown as the actuals. If costs post with a department dimension, the budget must carry a department dimension. Otherwise comparison is only possible at company level — which is precisely the number that explains nothing.
The practical consequence: before starting on budgeting you need a working dimension structure. Companies that begin with the budget and only later reach dimensions usually rebuild both.
A minimum working set
- Department or cost centre — essential. Without it the budget has no owner.
- Business line — if you have several and they have different economics.
- Project — for project-based businesses.
- Region — for companies with branches.
More than four dimensions in a budget is almost always an attempt to plan something nobody intends to control.
Versions: plan, revision, actual
One of the genuinely useful things a system provides and a spreadsheet does not: several parallel budget versions.
| Version | Created | Purpose |
|---|---|---|
| Original plan | Before the year starts | Fixed and never edited. The basis for judging planning quality |
| Revision | After the first quarter | A realistic reference point given what has actually happened |
| Best and worst case | As required | For decisions about investment and hiring |
The original plan is never overwritten. This matters: if the budget is revised quarterly and old versions are discarded, you will always have "nearly hit plan" at year end and will learn nothing about the quality of your own forecasting.
The annual cycle in practice
- October–November. Export nine months of actuals broken down by dimension — the basis for planning.
- November. Departments fill in their own sections. The practical approach is to give each one their slice in Excel and load it back: Business Central supports this in both directions, and resistance from the team is noticeably lower.
- December. Consolidate, agree, freeze the original version.
- Monthly. Budget versus actual alongside period close. Not as a separate ritual but as part of the routine.
- April. Revision as a new version; the original stays.
The key point is in step two. Forcing department heads to work directly inside the ERP purely to complete a budget is a reliable way to derail the process.
Four mistakes
The budget is more detailed than the ledger. Planning across twenty lines when actuals post to five makes comparison impossible, and the system gets the blame.
Only costs are planned. Without a revenue plan there is no result and therefore no conversation about profitability. Half a budget is not a budget.
The original version gets overwritten. A year later you have no idea how well you forecast, because there is nothing to compare against.
Budget versus actual is reviewed quarterly. Three months is too long to correct anything. A variance spotted in February can still be worked; spotted in April it is simply a fact. What to review monthly is covered in financial reporting.
Frequently asked questions
Can budgets be uploaded from Excel?
Yes, and it is the recommended approach. Each department gets its own slice in a familiar format, fills it in and returns it, and the system loads it against the correct dimensions. Forcing managers to complete budgets directly in the ERP is a reliable way to derail the process.
How many budget versions are worth keeping?
Usually two: the original plan and a quarterly revision. Best and worst case scenarios are worth building for specific decisions — investment, hiring — rather than permanently. The one firm rule is never to overwrite the original.
Are dimensions necessary for budgeting?
Essential, and in the same breakdown as actuals. A budget without dimensions only permits comparison at company level, which is exactly the number that explains nothing. If the dimension structure is not set up yet, start there.
How often should budget versus actual be reviewed?
Monthly, alongside period close. A quarter is too long a cycle: a variance found in the third month can no longer be worked. A monthly rhythm leaves time to react.
Can we plan quantities rather than just values?
Yes, through the sales and purchase budget by item and customer. This is useful to distributors for purchase planning, but it is an addition — start with the general ledger budget, which answers the main question about income and expenditure.