What-if analysis in BI is the interactive capability that allows dashboard users to modify input parameters — revenue growth rates, cost inflation assumptions, FX rates, headcount changes, or price changes — and observe the real-time impact on downstream calculated metrics (revenue, margin, EBITDA, cashflow) within the BI report, without changing the data model or creating a new planning scenario. In Power BI, what-if analysis is implemented through What-If Parameters — a feature that creates a parameter table (a single-column table with a range of possible values defined by the user) and a corresponding slicer on the report page. When the user moves the slicer to a new value, DAX measures that reference the parameter table’s selected value recalculate in real time — a revenue sensitivity measure that multiplies the current revenue by 1 + [GrowthRate Parameter] updates instantly as the growth rate slicer is adjusted. The dashboard becomes a dynamic sensitivity tool rather than a static report.
What-If Parameter Types in Finance BI
| Parameter Type | Finance Use Case | DAX Pattern |
|---|---|---|
| Growth rate assumption | Revenue sensitivity to growth rate change — how does EBITDA change if revenue grows 5% vs 10% vs 15%? | Revenue Sensitivity = [Actual Revenue] * (1 + SELECTEDVALUE(GrowthParam[Growth%], 0)) |
| FX rate override | Retranslate GCC subsidiary results at alternative rates — USD/SAR at 3.75 vs 3.80 | FX Revenue = [USD Revenue] * SELECTEDVALUE(FXParam[FXRate], 3.75) |
| Cost inflation | Project cost base under different inflation scenarios — baseline vs high inflation vs recession | Inflated Costs = [Actual Costs] * (1 + SELECTEDVALUE(InflationParam[InflationRate%], 0)) |
| Headcount change | Model cost of adding N headcount to a department | Additional Cost = SELECTEDVALUE(HeadcountParam[AddedHeads], 0) * [Avg Salary per Head] |
What-If Analysis vs Oracle EPM Planning
What-if analysis in Power BI or OAC is a visualisation-layer sensitivity tool — it does not persist scenarios, does not feed back to the planning system, and does not replace Oracle EPBCS or FCCS for formal scenario planning. Its value is in immediate sensitivity exploration: a CFO in a board meeting who wants to see the P&L impact of a 10% revenue shortfall can adjust the growth rate slicer in the Power BI dashboard and see the impact in seconds, without opening Oracle EPBCS, creating a new scenario version, loading assumptions, running a business rule, and waiting for the calculation. For quick sensitivity communication in executive discussions, Power BI what-if parameters are a practical complement to the formal EPM planning process — the formal scenario lives in EPM, the sensitivity exploration lives in the BI dashboard.
GCC Finance What-If Scenarios
GCC finance teams use what-if analysis in BI for several region-specific sensitivity scenarios. First, EGP devaluation sensitivity: for GCC holding companies with Egyptian operations, a Power BI what-if parameter for the EGP/USD or EGP/SAR exchange rate allows the CFO to see the group P&L impact of further EGP devaluation without rebuilding the consolidation model — the FX sensitivity is built into the dashboard’s DAX measures. Second, oil price sensitivity for Saudi entities whose revenue is directly or indirectly linked to oil price: a Brent crude price parameter feeds an estimated revenue impact measure, giving the board a rapid view of earnings sensitivity to oil market movements. Third, Saudi Vision 2030 investment return modelling: what-if parameters for investment timeline and return rate enable scenario exploration of long-term investment portfolio outcomes.
What Goes Wrong in Practice
The most common what-if analysis design failure is a parameter that can be set to a value that makes the downstream measure mathematically meaningless — a negative growth rate that produces negative revenue, or a zero divisor in a percentage calculation. What-if parameters must be bounded (minimum and maximum values defined in the parameter table) and the DAX measures that consume them must handle edge cases (DIVIDE with an alternative return for zero denominators, IF conditions that guard against implausible input values). A what-if parameter with no bounds and no guard conditions in the consuming measures will eventually produce a nonsensical result during a board presentation.
How Loop Wise Solutions Designs What-If Features
We implement what-if parameters in finance dashboards with defined value ranges, step increments calibrated to realistic assumption ranges (a revenue growth rate parameter that steps in 0.5% increments from -20% to +30%, not an unconstrained free-text field), and clear visual labels that distinguish what-if sensitivity outputs from actuals — ensuring users cannot mistake a sensitivity projection for reported figures.
