"In one engagement, a Nordic retail group was running Power BI Premium P2 at $10,000 per month — a capacity tier sized for peak load that was idle 70% of the time. Our independent analysis modelled a migration to Fabric F64 with autoscale burst capacity, reducing their monthly BI infrastructure cost from $10,000 to approximately $3,800 while maintaining peak-load performance. The annualised saving of $73,000 was achieved without reducing any reporting capability. The engagement fee was less than 9% of the first-year saving."
The Changing Landscape of Power BI Premium
Power BI Premium — Microsoft's capacity-based BI licensing tier — has been a staple of large enterprise analytics deployments since its introduction in 2017. The core value proposition was simple: purchase a dedicated capacity at a fixed monthly rate and allow unlimited users to consume Power BI content without each requiring an individual Pro licence. For organisations with hundreds or thousands of report consumers, the economics were compelling.
The landscape has changed materially. Microsoft's strategic direction is to retire traditional Power BI Premium P-SKUs in favour of Microsoft Fabric F-SKUs. After February 2025, P-SKUs are no longer available for new purchases — existing customers can renew but cannot expand on P-SKUs. For all new deployments, F-SKUs are the only capacity path, accessed through Azure billing rather than traditional EA volume licensing. This transition has significant commercial and operational implications for every organisation running Power BI at scale.
Understanding the implications requires clarity on four licensing models: Power BI Pro (per user, standard), Power BI Premium Per User (PPU, per user with premium features), Fabric F-SKUs (capacity-based, replacing P-SKUs), and legacy Power BI Premium P-SKUs (capacity-based, renewal only). Each has different economics, feature coverage, and suitability profiles depending on the organisation's size, usage pattern, and analytics maturity.
Power BI Pro: The Baseline
Power BI Pro is the entry-level per-user licence at $10 per user per month (standalone) or included within M365 E5 and certain other M365 bundles. Pro covers report creation, publishing, sharing, and consumption — but only between Pro-licensed users. If any report consumer does not hold a Pro licence, content cannot be shared with them without a Premium or Fabric capacity backing the workspace.
Pro is the correct licensing choice for small to medium organisations where all report consumers are active creators or power users, and where the total user count is below the capacity break-even threshold. As soon as the organisation introduces a significant population of report consumers who do not create content, Pro's per-user cost model becomes inefficient relative to capacity-based alternatives.
Within the M365 SKU context, M365 E5 includes Power BI Pro as part of the bundle, making it available to E5-licensed users at no incremental per-user cost. Organisations on M365 E1, E3, or the newer E7 SKU (Microsoft's current top-tier above E5, bundles advanced AI and security capabilities) need to assess whether the E5-to-E7 migration or a standalone Pro licence is the more economic path to enabling Power BI access for their user population.
Power BI Premium Per User (PPU)
Premium Per User (PPU) was introduced as a bridge between the per-user simplicity of Pro and the capacity-based model of traditional Premium. At $20 per user per month, PPU provides access to all Power BI Premium features on an individual basis — including larger datasets (up to 100 GB per model), 48 scheduled refreshes per day, deployment pipelines, paginated reports, XMLA endpoints, and AI-powered features — without requiring dedicated capacity.
The critical constraint is that PPU is all-or-nothing at the user level. Every user who accesses PPU-enabled content must hold a PPU licence. Unlike capacity-based Premium or Fabric, where free consumers can access content shared from a Premium workspace, PPU requires licensing for every consumer — creators and viewers alike. This makes PPU most suitable for analytics teams and departments where all consumers are active users who justify the $20 per user per month spend based on feature access.
The PPU Break-Even Calculation
The practical question for any Power BI deployment decision is: at what user count does PPU become more expensive than a Fabric F-SKU capacity, and at what point does Pro suffice? The break-even analysis is straightforward. An F64 Fabric capacity (the lowest tier with full Premium feature parity, including XMLA read-write, paginated reports, and unlimited viewer access) costs $4,995 per month at list price. At PPU pricing of $20 per user per month, this is equivalent to 250 users. Below 250 total users who need Premium features, PPU is cost-competitive or cheaper. Above 250 users requiring Premium feature access, an F64 capacity is more economical — and allows unlimited additional viewers at no incremental per-user cost.
This break-even is a simplified calculation. EA-negotiated discounts on Fabric F-SKU capacity — typically 20 to 40 percent below list for organisations with strong Azure consumption commitments — shift the break-even materially. At 30 percent discount, F64 costs $3,497 per month, equivalent to 175 PPU licences. Organisations with 175 to 250 total Premium-feature users sit in the evaluation zone where individual pricing and EA commercial terms determine the optimal model.
Unsure whether PPU or Fabric capacity makes sense for your BI environment?
Our Microsoft EA advisory specialists team provides independent break-even analysis and EA negotiation support.Power BI Premium P-SKUs: Retirement in Progress
For organisations currently running Power BI on P-SKU capacity (P1, P2, P3, or P4), the transition to Fabric F-SKUs is a near-term planning requirement. P-SKUs are no longer available for new purchases. Existing customers can renew existing P-SKU commitments but cannot expand capacity or migrate to new P-SKU agreements. Microsoft's published direction is full migration to F-SKUs at the next renewal opportunity.
The P-to-F SKU migration is more than a naming change. P-SKUs are procured through traditional EA volume licensing and appear on the EA schedule. F-SKUs are Azure resources, billed through Azure subscriptions and subject to Azure Hybrid Benefit eligibility, Azure Reserved Instance pricing, and Azure Cost Management tooling. The commercial process, billing model, and governance framework are materially different.
P-SKU to F-SKU equivalencies: P1 (8 v-cores) maps to F64 at $4,995 per month. P2 (16 v-cores) maps to F128 at $9,990 per month. P3 (32 v-cores) maps to F256 at $19,980 per month. These are list prices; negotiated rates typically carry a 20 to 40 percent discount depending on Azure commitment level and EA commercial terms.
Microsoft Fabric F-SKUs: The New Capacity Standard
Microsoft Fabric is the unified analytics platform that integrates Power BI, Azure Synapse, Data Factory, and other data workloads into a single capacity-based licensing framework. Fabric F-SKUs are the capacity unit, sold in Capacity Units (CUs) that scale from F2 to F2048. The entry-level F2 at $262 per month is priced far below the old P-SKU floor, making Fabric accessible for smaller deployments that previously could not justify $4,995 per month for a P1.
The critical feature threshold within Fabric F-SKUs is at F64. SKUs below F64 (F2 through F32) do not include unlimited viewer access — report consumers still require Pro licences. F64 and above include all Power BI Premium features, eliminating the need for individual Pro or PPU licences for report consumers. This means organisations evaluating Fabric for broad enterprise BI deployment should model F64 or higher as the effective entry point for the full Premium capability set.
Fabric also introduces the pause and resume capability, where Azure-billed F-SKU capacity can be paused during periods of no usage and resumed on demand. For organisations with predictable analytics workloads — batch refresh schedules and daytime-only report consumption — pause and resume can reduce effective monthly capacity cost by 30 to 50 percent compared with always-on P-SKU capacity.
Fabric and the Broader M365 Context
Fabric capacity is independent of the M365 licensing stack. An organisation on M365 E3 or E7 still needs to separately provision Fabric capacity for premium analytics capabilities. However, M365 Copilot features available within M365 E7 include integration with Copilot in Power BI, which leverages Fabric capacity for AI-generated report insights and natural language query. Organisations evaluating E7 as the M365 top SKU should assess whether the Copilot analytics integration justifies incremental investment in F64+ Fabric capacity alongside the E7 licence cost.
Dataverse Storage: The Hidden Cost
For organisations building Power Apps and Power BI solutions on Dataverse — Microsoft's low-code application platform database — storage costs represent a frequently underestimated component of the total analytics licensing spend. Dataverse charges $40 per GB per month for database capacity and $2 per GB per month for file capacity. At enterprise scale, where Dataverse tables accumulate years of operational data and attachments, storage costs can rival or exceed licence costs.
The default Dataverse storage allocation is 10 GB of database capacity and 20 GB of file capacity for the tenant, plus 250 MB of database and 2 GB of file capacity per user with a qualifying licence. Organisations with large datasets in Dataverse should model storage costs explicitly rather than assuming default allocations will suffice.
Key Decision Framework
When selecting the right Power BI licensing model, five questions drive the decision. How many total report consumers need access? Below 100 consumers, Pro or PPU typically suffices. At 200 to 500 consumers, F64 Fabric capacity becomes competitive. Above 500 consumers, Fabric capacity is almost always more economical. Do consumers need Premium features? If only basic report consumption is needed, Pro at $10 per user per month remains efficient. Premium features (large datasets, paginated reports, deployment pipelines, AI) require PPU or F64+. What is the current Azure commitment level? Higher Azure consumption unlocks better Fabric F-SKU discounts, potentially shifting break-even thresholds significantly. Is the organisation currently on P-SKUs? If yes, plan F-SKU migration at next renewal — do not wait for P-SKU forced retirement. What are the workload patterns? Variable or batch-only workloads benefit from Fabric pause and resume, which can materially reduce effective capacity cost.
Six Recommendations for 2026
1. Stop purchasing P-SKUs: The P-SKU retirement trajectory is clear. All new capacity investments should go to Fabric F-SKUs, regardless of existing P-SKU footprint.
2. Model the F64 break-even before committing to PPU at scale: If PPU licences will exceed 175 to 250 users, run the F64 vs PPU cost model with your actual EA-negotiated Azure pricing before committing. At negotiated rates, the break-even is often lower than list price calculations suggest.
3. Enable pause and resume for variable workloads: Any Fabric capacity backing workloads with predictable off-hours should implement automated pause and resume. The potential 30 to 50 percent effective cost reduction pays for the implementation effort in the first month.
4. Include Fabric capacity in EA Azure commit negotiations: Fabric F-SKUs billed through Azure are eligible for enterprise Azure pricing. Including Fabric capacity commitments in EA Azure consume negotiations typically yields 20 to 30 percent discount versus pay-as-you-go rates.
5. Audit Dataverse storage before enabling Fabric at scale: Large Dataverse deployments carry storage cost risk that must be modelled before committing to Fabric-based analytics architectures. The $40 per GB per month database rate accumulates quickly at enterprise data volumes.
6. Engage independent Microsoft EA advisory specialists before migrating from P to F: The P-to-F migration involves billing model changes, Azure subscription provisioning, capacity management process redesign, and EA commercial renegotiation. Independent advisory with direct P-to-F migration experience reduces the risk of commercial missteps during the transition.
Stay Current on Power BI and Fabric Licensing
Microsoft's Fabric licensing evolves rapidly. Subscribe for independent analysis on P-SKU retirement timelines, F-SKU pricing changes, and Fabric feature expansion.