A planning model can be structured in more than one way, and the structure you choose decides a lot about what the model can do later. This session walks through three of them in , built directly on top of your semantic model, inside Microsoft Fabric where your data already lives. We build all three on the same dataset so the differences are easy to compare.
- Tabular structure. The plan is a grid. Dimensions come in as rows, and you build out across columns: input columns, simulation columns, visual measures, commentary. Top-down and bottom-up entry, spreads, allocations, scenarios. No DAX. Suits plans that are line items over time.
- Row model. Every row becomes a driver. Point the model at one hierarchy column, say chart of accounts, cost center or product SKU, and each row inside it turns into a planning driver. Built for lightweight semantic models, flat tables or star schemas, where the calculation logic has to be built in the planning layer. Analysts add drivers by adding rows, not by filing a request with IT.
- Measure model. Every DAX measure becomes a driver. The plan inherits governed measures like Revenue Actuals, and you build visual measures such as Revenue Forecast on top without touching backend DAX. Built for teams with a mature semantic layer, so actuals and forecasts stay anchored to the same definitions.
Between the row model and the measure model, the deciding factor is how developed your semantic model already is: whether the logic needs building in the plan, or already exists and just needs extending.
You'll be able to match a planning requirement to the structure that fits it.
Register today!
