Power BI Datamarts Licensing: Self-Service Data Management Guide 2026

Power BI Datamarts launched in 2022 as Microsoft's answer to the self-service data management gap between Power BI datasets and enterprise data warehouses. They provide business analysts with a managed Azure SQL Database backend, auto-generated semantic model, and no-code/low-code query experience — without requiring engineering involvement. The licensing bar is higher than standard Power BI: Datamarts require Premium per User or Premium capacity, not just Pro. At 50 analysts deploying departmental datamarts, the licensing gap between Pro ($10/user) and PPU ($20/user) represents $6,000/month in additional spend that many organisations fail to budget for.

Datamart Licensing Requirements

Datamarts have a strict licensing requirement: Premium per User (PPU) or Premium capacity (P/F SKU). Power BI Pro does not provide Datamart creation rights, regardless of workspace type.

Activity	Power BI Pro	Power BI PPU	Premium Capacity (P/F SKU)
Create Datamart	❌	✅ (in PPU workspace)	✅ (in Premium workspace)
Edit Datamart (data transforms)	❌	✅	✅
Query Datamart via SQL endpoint	❌ (own queries)	✅	✅
Build reports from Datamart	❌ (in Pro workspace)	✅	✅
View reports from Datamart	✅ (Pro workspace)	✅	✅ (Free licence in Premium workspace)
Share Datamart (read access)	Limited	✅	✅

Datamarts vs Fabric Warehouse: When to Use Each

The Datamart vs Fabric Warehouse decision is as much organisational as technical. Both provide SQL query capability over structured data, but they target different users and scale requirements.

Dimension	Power BI Datamart	Fabric Warehouse
Primary user	Business analyst (self-service)	Data engineer / advanced analyst
Storage limit	100 GB per datamart	No hard limit (OneLake)
T-SQL capability	SELECT queries only; no DDL/DML	Full T-SQL DDL/DML
Data ingestion	Dataflows Gen2 (Power Query)	Pipelines, Spark, COPY INTO, T-SQL
Auto-generated semantic model	✅ Auto-created	Manual or Fabric-assisted
IT governance	Low overhead (self-service)	IT/engineering managed
Licensing requirement	PPU or Premium capacity	Fabric F SKU (F64+)
Storage billing	Included in PPU/Premium	OneLake $0.023/GB/month
Fabric Lakehouse integration	Limited (SQL endpoint access)	Native (same OneLake)
Suitable data size	<100 GB	Unlimited

The practical decision: if your business analysts need a personal or departmental SQL-accessible store for data transformations and the data volume is under 100 GB, Datamarts serve the purpose within PPU licensing. If the data engineering team owns the data pipeline and the store needs to scale beyond 100 GB or integrate with Spark/real-time workloads, Fabric Warehouse is the correct choice — but requires an F64+ capacity investment.

Cost Modelling: PPU vs Premium Capacity for Datamarts

Scenario 1: Small Team, Datamarts Only

10 business analysts who need Datamarts and no other Premium features. PPU at $20/user/month × 10 = $200/month. Compared to F64 capacity ($5,980/month 3-year reserved) — PPU wins overwhelmingly for small analyst teams. The F64 break-even for Datamart-only usage is approximately 300 PPU users ($6,000/month vs $5,980/month).

Scenario 2: Mixed BI Platform with Datamarts

150 Power BI users, 25 of whom need Datamart authoring. Rest need Pro for standard reporting. Option A: Upgrade all 25 Datamart users from Pro to PPU = 25 × $10/month incremental = $250/month additional. Option B: Move all 150 users to Premium capacity (F64) = $5,980/month vs 150 × $10 Pro = $1,500/month. PPU selective upgrade is far more cost-efficient unless Premium capacity delivers value for other workloads (Fabric, XMLA, deployment pipelines).

Scenario 3: Enterprise Self-Service Analytics Programme

Organisation deploys Datamarts as a standardised self-service analytics tool for 200 business analysts across 15 departments. At PPU: 200 × $20 = $4,000/month. At F64 Premium capacity: $5,980/month but covers all Power BI viewers (eliminating 500 viewer-only Pro licences at $5,000/month). Total with Premium: $5,980/month vs $4,000 PPU + $5,000 viewer Pro = $9,000/month. Premium capacity saves $3,020/month when the viewer population is large.

Datamart Storage and Data Refresh Limits

Understanding the Datamart technical limits prevents unexpected capacity planning failures:

• Storage per Datamart: 100 GB hard limit. No option to increase beyond 100 GB — data volumes above this require migration to Fabric Warehouse or Azure SQL Database.
• Scheduled refresh: Up to 8 refreshes per day (PPU); up to 48 refreshes per day (Premium capacity). For near-real-time use cases, the PPU refresh limit may be a constraint.
• Tables per Datamart: No published hard limit, but performance degrades above 200 tables in practice.
• Concurrent queries: Managed automatically, but heavy concurrent query load may require Premium capacity over PPU for acceptable performance.

Datamarts and Microsoft Fabric: Migration Path

Microsoft's Fabric roadmap treats Datamarts as a stepping stone toward Fabric Warehouse rather than as a permanent strategic destination. The practical implication for licensing strategy: do not build critical data infrastructure on Datamarts if your data volumes or query complexity are approaching the 100 GB or T-SQL limits. Instead, treat Datamarts as an accelerator for departmental analytics while planning Fabric Warehouse for production-grade workloads.

Migration from Datamart to Fabric Warehouse is possible: the Datamart's Azure SQL Database backend can be accessed via SQL endpoint and data extracted to Fabric Lakehouse/Warehouse. Microsoft has not provided a one-click migration tool as of Q1 2026.

EA Optimisation: Getting the Right Licence for Datamarts

At EA renewal, Datamart licensing is typically a line-item negotiation within the broader Power BI/Fabric discussion. The key levers:

PPU vs Premium capacity positioning: If your Datamart user count is below 300, PPU is commercially superior to purchasing Premium capacity solely for Datamart access. Use this analysis to resist Microsoft account team suggestions to "upgrade to Premium" when PPU meets your requirement at half the cost.

M365 E5 bundle check: Microsoft 365 E5 includes Power BI Pro but not PPU or Premium capacity. If your target Datamart users are on E5, they still need PPU or Premium access for Datamarts. Do not assume E5 covers Datamart authoring — it does not.

Fabric as the alternative: If Microsoft pushes for Premium capacity to enable Datamarts, counter with Fabric F SKU — at F64, you get Datamarts via Premium capacity equivalence AND Fabric Warehouse (which can replace Datamarts at scale) AND all other Fabric workloads. Use the Fabric discussion to negotiate better F SKU pricing rather than accepting P SKU positioning.

Frequently Asked Questions

What licence is required to create a Power BI Datamart?

Creating Power BI Datamarts requires Power BI Premium per User (PPU) at $20/user/month OR Power BI Premium capacity (P SKU or F SKU — F64+). Power BI Pro alone does not provide Datamart creation rights. This is one of the most common unexpected licensing gaps when organisations deploy self-service analytics programmes.

What is the storage limit for Power BI Datamarts?

Power BI Datamarts have a hard 100 GB per-datamart storage limit. This applies regardless of workspace type (PPU or Premium capacity). For data volumes exceeding 100 GB, Microsoft Fabric Warehouse (no per-warehouse storage cap, billed via OneLake at $0.023/GB/month) is the recommended path.

How do Power BI Datamarts compare to Fabric Warehouse?

Datamarts target self-service business analysts with limited SQL needs — up to 100 GB, SELECT queries only, auto-generated semantic model, Power Query ingestion. Fabric Warehouse supports unlimited data, full T-SQL DDL/DML, engineering-grade workloads, and deep OneLake integration. The choice maps to audience: Datamarts for business analysts, Warehouse for data engineers.

Are Power BI Datamarts being deprecated in favour of Fabric?

Microsoft has positioned Fabric Warehouse as the strategic path for managed SQL analytics. Datamarts receive limited new feature investment but are not deprecated as of 2026. New projects should evaluate Fabric Lakehouses and Warehouses as the primary path, with Datamarts serving as a transitional option for existing lightweight deployments.

Can Power BI Datamart data be accessed from Microsoft Fabric?

Yes. Datamarts expose a SQL connection endpoint that Fabric notebooks and pipelines can connect to using standard SQL connections. This allows Datamart data to participate in broader Fabric workflows without requiring immediate migration to Fabric Warehouse.