Manually entered assumption values, or calculated assumption values, appear in a centralized interface where they can be adjusted once and cycle through the entire model. This enhances data integrity by providing a single pane of glass view for core drivers and simplifies maintenance.
Scenario: A global company uses CPM software for budgeting and financial forecasting. They have different versions of budgets for divisions like marketing, sales, and production, and each department makes individual assumptions which guide these budgets.
Solution: By centralizing and sharing assumptions, individual departments have access to each other's estimations. For instance, the sales department's assumption of a 20% growth in customer acquisition can be shared, factored into the marketing budget for customer acquisition campaigns, and into production for product development and inventory management. This results in accurate budgets and forecasts that are aligned across departments.
This requirement describes a centralized assumption interface which can be incredibly helpful. It also alludes to a centralized location that pushes out to different versions. In other words, the assumption table can be independent from a version if it needs to be. Some systems call that a “global assumption” meaning that, no matter what, changing it will effect everything in the system.
You may want instead to have a version-specific or local assumption that, when modified, only affects the active version. In either case we recommend you discuss this in detail with the vendor.
However the vendor does it, make sure there is some sort of centralization. Rogue assumptions floating around the model are a hassle to manage.