← Article review

Power BI and Fabric: incremental TCO model

Illustrative scenarios using verified US-site user rates and Central US USD PAYG/hot-storage rates checked 29 October 2026. Not a contract quote. Taxes, existing platform costs, discounts and currency conversion are excluded unless you enter them in additional services. Authors and viewers are distinct people. Capacity candidates are not performance recommendations.

Enter costs before calculating.

Full-price PPU already includes Pro features: do not enter their sum. Existing Pro/E5 alone does not satisfy a PPU-content requirement. For F64+, the model assumes eligible Free viewers with correct access and no remaining PPU dependency; authors remain separately licensed. For smaller F SKUs, viewers remain paid. Special embedding, guest and mixed-licence cases require a separate assessment.

Pro/PPU patterns use existing preparation/warehouse infrastructure; add its incremental cost in services where appropriate. Fabric mode adds hot storage only: other storage tiers, BCDR, network/API charges, overage, Spark autoscale and other workloads must be entered explicitly. Native Power BI storage is not automatically chargeable OneLake hot storage. Reservations do not shrink when active hours fall in this model. Pausing capacity does not cancel commitments, storage or potential consumption settlement.

Formula: newly required user seats × applicable rate + PAYG active capacity hours × whole-SKU hourly rate (or quoted reservation monthly equivalent) + applicable GB × storage rate + services + operations + implementation/amortisation months. No automatic recommendation or measured performance is produced.