Will my Power BI Premium license automatically convert to Fabric F-SKU?

No, migration is not automatic. Your P-SKU capacity will continue to work during the grace period (August 2025 - December 2026). You must purchase F-SKU capacity separately and manually reassign workspaces. Microsoft provides tools to make this process seamless, but it requires deliberate action. If you do nothing, your P-SKU will eventually be retired in December 2026, forcing migration. We recommend planning your migration in 2025-2026 to avoid last-minute rush.

Can I migrate some workspaces to F-SKU and keep others on P-SKU temporarily?

Yes, this is a common strategy. You can run P-SKU and F-SKU capacities in parallel during the migration period. Assign low-risk workspaces to F-SKU first, validate functionality, then gradually move high-risk workspaces. Many enterprises run parallel capacities for 3-6 months during transition. However, you cannot assign a single workspace to both P-SKU and F-SKU simultaneously - it must be one or the other. This gradual approach reduces risk but increases short-term cost due to dual capacity charges.

What happens to my data after I migrate to Fabric F-SKU?

Your data remains in the same region and workspace. Migration does not move or copy data - it simply changes which capacity powers the workspace. Semantic models, reports, dataflows, and paginated reports continue to function identically. The workspace ID remains the same, so external tools and APIs do not need reconfiguration. The only change is the underlying compute capacity (P-SKU → F-SKU). After migration, you gain access to new Fabric features like OneLake, Data Engineering, and Real-Time Intelligence, but existing Power BI workloads are not affected. Data sovereignty and compliance certifications (GDPR, HIPAA, SOC 2) remain intact.

Power BI Premium to Microsoft Fabric Migration: Complete Enterprise Guide for 2025

Microsoft is transitioning Power BI Premium customers to Microsoft Fabric F-SKU capacities. This migration unlocks unified analytics capabilities beyond traditional BI, but requires careful planning for licensing, capacity sizing, and workload migration. This comprehensive guide walks you through the enterprise migration process. Our Microsoft Fabric consulting services provide end-to-end migration support.

Understanding the Transition

What's Changing

Microsoft announced that Power BI Premium Per Capacity (P-SKUs) will transition to Microsoft Fabric F-SKU capacities. This is not a forced migration, but a strategic evolution of the platform.

Timeline: - August 2025: Grace period begins for transitioning - December 2026: P-SKU retirement (18-month window) - Ongoing: F-SKU enhancements and feature parity

What this means: - Existing P-SKU licenses continue to work during grace period - New Premium capacity purchases use F-SKUs - Migration is one-way (cannot downgrade F-SKU to P-SKU)

For capacity planning context, see our Fabric sizing guide.

Why Microsoft Is Making This Change

Unified Platform Strategy: Fabric brings together Power BI, Data Factory, Synapse Data Engineering, Data Science, Real-Time Intelligence, and Data Activator under one capacity model.

Simplified Licensing: One capacity type (F-SKU) instead of multiple (P, EM, A SKUs).

Future-Proofing: New features (OneLake, Copilot, Real-Time Intelligence) require Fabric capacity.

Cost Efficiency: Pay-as-you-go and autoscaling options not available with P-SKUs.

P-SKU vs F-SKU Comparison

Capacity Mapping

| Power BI Premium (P-SKU) | Microsoft Fabric (F-SKU) | vCores | Price/Month* | |--------------------------|--------------------------|--------|--------------| | P1 | F64 | 8 vCores | $8,384 | | P2 | F128 | 16 vCores | $16,768 | | P3 | F256 | 32 vCores | $33,536 | | P4 | F512 | 64 vCores | $67,072 | | P5 | F1024 | 128 vCores | $134,144 |

*Approximate pricing based on pay-as-you-go rates. Regional variations apply.

Feature Comparison

| Feature | P-SKU | F-SKU | |---------|-------|-------| | Power BI Reports & Dashboards | ✅ | ✅ | | Paginated Reports | ✅ | ✅ | | Dataflows Gen2 | ✅ | ✅ | | Deployment Pipelines | ✅ | ✅ | | OneLake | ❌ | ✅ | | Data Engineering (Spark) | ❌ | ✅ | | Data Science Notebooks | ❌ | ✅ | | Real-Time Intelligence | ❌ | ✅ | | Data Factory | ❌ | ✅ | | Copilot | Limited | ✅ | | Autoscaling | ❌ | ✅ | | Pay-as-you-go | ❌ | ✅ |

Key takeaway: F-SKU = everything P-SKU offered + modern analytics platform capabilities.

Migration Decision Framework

Should You Migrate Now or Wait?

Migrate Immediately If: ✅ You need OneLake for unified data lake ✅ You want to use Data Engineering or Data Science workloads ✅ You need Real-Time Intelligence (Eventstream, KQL) ✅ You want autoscaling to handle usage spikes ✅ You prefer pay-as-you-go over annual commitment

Wait Until Later If: ⏳ You only use basic Power BI reports (no Fabric features needed) ⏳ You have multi-year P-SKU contract not expiring soon ⏳ Your organization is risk-averse (let others migrate first) ⏳ You lack Fabric skills in your team (need training first)

Must Migrate By: ⚠️ December 2026 - P-SKU retirement deadline

Pre-Migration Assessment

Step 1: Inventory Current P-SKU Usage

Capacity utilization: - How many P-SKU capacities do you have? - What is average CU consumption? - Are you hitting capacity limits regularly?

Workspaces and workloads: - How many Premium workspaces? - Which workloads are enabled? (Dataflows, paginated reports, etc.) - Any custom connectors or XMLA endpoints in use?

User base: - How many Pro users consuming Premium content? - Any Power BI Embedded scenarios? - Guest user access patterns?

Step 2: Calculate F-SKU Requirements

Use Microsoft's capacity calculator: 1. Input current P-SKU (e.g., P1) 2. Review recommended F-SKU (e.g., F64) 3. Account for growth (add 20-30% headroom)

Example: - Current: P1 (8 vCores) - Usage: 70% average utilization - Growth projection: 30% user increase in 12 months - Recommendation: F64 with autoscale to F128 for peaks

Step 3: Cost Analysis

P1 (Annual Contract): $4,995/month × 12 months = $59,940/year

F64 (Pay-as-you-go): $8,384/month × 12 months = $100,608/year

F64 (1-Year Reserved): $8,384 × 80% discount = $6,707/month × 12 = $80,484/year

F64 (3-Year Reserved): $8,384 × 62% discount = $5,198/month × 12 = $62,376/year

Verdict: 3-year reserved F-SKU is cost-competitive with P1 annual pricing.

For detailed cost optimization strategies, see our Fabric capacity planning guide.

Migration Strategies

Strategy 1: Direct Upgrade (Recommended for Most)

Best for: Organizations with single P-SKU and straightforward BI workloads

Steps: 1. Purchase F-SKU capacity (same region as P-SKU) 2. Assign all Premium workspaces to F-SKU capacity 3. Test reports, dataflows, and refresh schedules 4. Verify no functionality breaks 5. Decommission P-SKU capacity

Downtime: Near-zero (workspaces remain online during reassignment)

Strategy 2: Gradual Migration

Best for: Large organizations with multiple P-SKUs and complex dependencies

Steps: 1. Purchase initial F-SKU capacity 2. Migrate low-risk workspaces first (dev/test) 3. Monitor for 2-4 weeks 4. Migrate medium-risk workspaces (non-critical production) 5. Migrate high-risk workspaces (mission-critical reports) 6. Decommission P-SKU after all workspaces migrated

Timeline: 2-3 months for large enterprises

Strategy 3: Parallel Run

Best for: Organizations with strict uptime requirements (24/7 operations)

Steps: 1. Purchase F-SKU capacity 2. Clone workspaces to F-SKU capacity 3. Run parallel BI environments (P-SKU + F-SKU) 4. Validate F-SKU performance matches P-SKU 5. Cut over users to F-SKU workspaces 6. Decommission P-SKU

Cost: Temporary double capacity cost during parallel run

Step-by-Step Migration Process

Phase 1: Planning (Week 1-2)

1. Stakeholder Alignment - Executive sponsor approval - IT, BI, and finance alignment - Communication plan for end users

2. Technical Readiness - Fabric capacity provisioned in Azure - Admin access configured - Monitoring tools set up (Fabric Capacity Metrics app)

3. Risk Mitigation - Backup all workspaces and datasets - Document current refresh schedules - Identify critical reports and dependencies

Phase 2: Pilot Migration (Week 3-4)

1. Select Pilot Workspace Choose a non-critical workspace with: - Representative report types - Dataflows and datasets - Moderate user base (10-50 users)

2. Migrate Pilot

PowerShell script to reassign workspace: Set-PowerBIWorkspace -Scope Organization -Id <workspace-id> -CapacityId <F-SKU-capacity-id>

3. Test and Validate - [ ] Reports load correctly - [ ] Scheduled refreshes execute - [ ] Data gateways function - [ ] Dataflows process successfully - [ ] User access permissions maintained

4. Monitor Performance - CU consumption via Fabric Capacity Metrics - Report load times - Refresh duration - User feedback

Phase 3: Full Migration (Week 5-8)

1. Batch Workspace Migration Migrate in waves (10-20 workspaces per wave): - Wave 1: Dev/test workspaces - Wave 2: Non-critical production workspaces - Wave 3: Critical production workspaces

2. Post-Migration Validation For each wave: - [ ] Reports accessible to users - [ ] Scheduled refreshes successful - [ ] Performance meets SLAs - [ ] No capacity throttling

3. User Communication - Email users 48 hours before migration - Notify during migration window - Provide support contact for issues

Phase 4: Optimization (Week 9-12)

1. Right-Size Capacity - Review CU consumption patterns - Scale up if utilization exceeds 80% - Scale down if utilization below 50% - Enable autoscale for variable workloads

2. Leverage New Fabric Features - Enable OneLake shortcuts for data integration - Migrate dataflows to Dataflows Gen2 - Explore Data Engineering for ETL workloads - Pilot Real-Time Intelligence use cases

3. Cost Optimization - Switch to reserved capacity for 20-38% discount - Pause dev/test capacities outside business hours - Implement incremental refresh to reduce CU consumption

Common Migration Issues and Fixes

Issue 1: Capacity Throttling After Migration

Symptom: Reports load slowly, refreshes queued

Cause: Underestimated capacity requirements (P1 → F64 not sufficient)

Fix: - Enable autoscale temporarily - Monitor peak usage hours - Upgrade to next F-SKU tier if consistently above 80%

Issue 2: Dataflows Fail After Migration

Symptom: Dataflows that worked in P-SKU fail in F-SKU

Cause: Dataflow uses deprecated connectors or legacy settings

Fix: - Recreate dataflow in Dataflows Gen2 - Update connectors to latest versions - Verify data gateway connectivity

Issue 3: XMLA Endpoint Connectivity Lost

Symptom: External tools (Tabular Editor, ALM Toolkit) cannot connect

Cause: XMLA endpoint requires re-authentication after capacity change

Fix: - Reconnect external tools to new F-SKU capacity endpoint - Update connection strings in automation scripts - Verify firewall rules allow Fabric XMLA endpoints

Issue 4: Cost Overrun in First Month

Symptom: F-SKU costs 2x higher than expected

Cause: Pay-as-you-go pricing + no autoscale limits

Fix: - Review Capacity Metrics to identify high CU consumers - Optimize dataset refresh schedules (stagger refreshes) - Implement incremental refresh for large datasets - Switch to reserved capacity for predictable pricing

Post-Migration Enablement

Train Your Team on Fabric

For BI Developers: - OneLake data integration patterns - Dataflows Gen2 vs Dataflows Gen1 - Fabric shortcuts for data reuse - Notebooks for advanced transformations

For Data Engineers: - Spark notebooks in Fabric - Delta Lake optimization - Pipeline orchestration - Real-Time Intelligence (Eventstream, KQL)

For Analysts: - Copilot for natural language queries - Real-time dashboard creation - Direct Lake mode for fast queries

Establish Fabric Governance

Capacity Management: - Assign workspace admins for capacity allocation - Set up alerts for 80%+ utilization - Review CU consumption monthly

OneLake Governance: - Define OneLake folder structure - Implement sensitivity labels - Configure retention policies

Security & Compliance: - Extend RLS from Power BI to Fabric datasets - Configure Purview integration for data catalog - Implement DLP policies for sensitive data

For governance best practices, see our Fabric governance guide.

Licensing Considerations

User Licensing (Unchanged)

F-SKU capacity still requires user licenses: - Power BI Pro: $10/user/month (create and share reports) - Power BI Premium Per User (PPU): $20/user/month (advanced features) - Fabric Trial: Free 60-day trial (limited capacity)

Example: - F64 capacity: $8,384/month - 500 Pro users: $5,000/month - Total: $13,384/month

Fabric-Specific Licensing

New with F-SKU: - Data Engineering: Included in F-SKU - Data Science: Included in F-SKU - Real-Time Intelligence: Included in F-SKU

No additional licenses needed beyond capacity + user licenses.

Migration Checklist

Pre-Migration

[ ] Document all Premium workspaces and datasets
[ ] Calculate required F-SKU capacity
[ ] Purchase F-SKU capacity (same region as P-SKU)
[ ] Provision Fabric capacity in Azure portal
[ ] Assign admin access to Fabric capacity
[ ] Install Fabric Capacity Metrics app
[ ] Communicate migration plan to stakeholders
[ ] Backup all workspaces and datasets

During Migration

[ ] Reassign pilot workspace to F-SKU
[ ] Test and validate pilot workspace (1 week)
[ ] Batch migrate remaining workspaces
[ ] Validate each wave before proceeding
[ ] Monitor Fabric Capacity Metrics for throttling
[ ] Provide user support during transition

Post-Migration

[ ] Validate all reports and dataflows functional
[ ] Verify scheduled refreshes executing successfully
[ ] Review capacity utilization (target 60-80%)
[ ] Enable autoscale if needed
[ ] Optimize refresh schedules to reduce CU consumption
[ ] Switch to reserved capacity for cost savings
[ ] Train users on new Fabric capabilities
[ ] Decommission P-SKU capacity

Business Case for Migration

ROI Calculation

Costs: - F64 capacity (3-year reserved): $62,376/year - Migration project (consulting, internal labor): $50,000 one-time

Benefits: - Data engineering capability: $100,000/year (eliminates third-party ETL tools) - Real-Time Intelligence: $75,000/year (faster decision-making, operational efficiency) - Autoscaling cost avoidance: $30,000/year (vs. over-provisioning P-SKU) - Developer productivity (Copilot, unified platform): $50,000/year

Total 3-Year TCO: - Cost: $62,376 × 3 + $50,000 = $237,128 - Benefit: $255,000 × 3 = $765,000 - Net ROI: $527,872 (222% ROI)

Conclusion

Migrating from Power BI Premium to Microsoft Fabric F-SKU is a strategic investment in your analytics future. Key takeaways:

Timeline: - Grace period: August 2025 - December 2026 - Recommended: Migrate by mid-2026 to avoid rush

Costs: - F64 3-year reserved ($62,376/year) = cost-competitive with P1 ($59,940/year) - Additional value: OneLake, Data Engineering, Real-Time Intelligence, Copilot

Migration: - Direct upgrade: 2-4 weeks for simple environments - Gradual migration: 2-3 months for complex enterprises - Parallel run: 3-6 months for 24/7 critical systems

Enablement: - Train teams on Fabric capabilities beyond BI - Establish governance for OneLake and capacity management - Leverage autoscaling and reserved capacity for cost control

Ready to migrate to Microsoft Fabric? Contact our migration specialists for assessment and implementation services.

**Sources**: - Microsoft: Grace Period for Power BI Premium to Fabric Transition - Microsoft: Power BI Premium Licensing Update - Microsoft Learn: What's New in Power BI January 2026