Personnel cost budgeting is still done in Excel in many companies, even where the logic of the calculations outgrew spreadsheets long ago. We explain how multi-level analytics with a non-standard calculation structure was moved into 1C:ZUP for an international manufacturing company.
Personnel cost budgeting at the client company includes several layers of management analytics. Beyond the standard items, annual bonuses must be planned depending on the manager's level, PIT calculated separately, records kept by financial responsibility centre and reports produced for different controlling purposes.
The data for the calculations is stored in different sources: before preparing the budget it had to be exported manually and consolidated into a single Excel file. Over time the file grew to dozens of sheets with cross-referenced formulas, and changing one parameter required checking the whole chain. Only one specialist understood the logic of those formulas in detail. After key staff changes and the loss of the file, it became practically impossible to reproduce the correct calculation methodology.
UCMS Group specialists, who have been running HR records and payroll for the company since 2014, proposed automating the budgeting process in 1C:ZUP.
The personnel cost budgeting module lets you plan the payroll fund, taxes, bonuses and other personnel costs directly inside the payroll system. Actuals are pulled automatically from the same data used for payroll and compared with the plan in the required dimensions.
After a demonstration of the capabilities, the client decided to implement it.
There were instructions for the old Excel file, but they no longer reflected how the model actually worked. Many links between items existed only inside macros and cross-referenced formulas, and the intermediate calculations were not documented.
Elena Terekhina, software developer at UCMS Group:
“We had the source data and the final table, but not the intermediate steps. We rebuilt the logic from the old figures: we took what should come out at the end and derived the formula backwards for each item separately. It was real detective work.”
At the same time it was important not just to reproduce the old model but to make it manageable: so that when a new cost item appears or parameters change, there is no need to go into the code and everything can be reconfigured through reference books.
The module's standard functionality covers most typical tasks. This project, however, required customisation for the client's complex management model.
Examples of customisations:
We also configured work with several budget versions. The company reviews its budget once a quarter, when new employees join, conditions change or cost items are refined. Each version is saved separately so that the original can be compared with the current one or you can see how the figures changed. Recalculating the budget for new conditions now takes minimal time, because all calculations are performed automatically according to the configured rules.
Implementation and training ran in parallel. The first budget calculation for 2025 was done together: a UCMS Group specialist and a client representative calculated at the same time, constantly comparing the results with the old Excel file. This made it possible to identify and fix discrepancies quickly and to show in practice how the new model works.
As the client mastered the system, it began calculating the budget independently. Questions that arose were handled in short consultations. The specialist explained the logic and the client then did the configuration itself. Gradually almost all operations moved to the client's finance team.
In parallel we prepared documentation: a user guide describing the standard scenarios and a separate section on non-standard settings. The guide is written so that even a new employee can quickly understand the process without diving into the project history.
Companies with non-standard management analytics often postpone budgeting automation, because it seems the system will not be able to reproduce logic that has evolved over years for specific tasks. This project shows that the complexity of the model is not an obstacle: the tool is configured around real processes, not the other way round.
