Comparison
Strategy Mosaic vs Databricks metric views: universal semantic layer or native metrics?
Strategy Mosaic and Databricks Unity Catalog metric views compared: scope across platforms, how metrics are defined, governance, which BI and AI tools can use them, pricing model and when each one fits.
Hemant Walchale
· 3 min read
Both products promise the same thing: define a metric such as revenue or churn once, and get the same answer in every dashboard and AI assistant. They differ in where that definition lives. Databricks metric views live in Unity Catalog, next to the data. Strategy Mosaic is an independent layer above several data platforms.
The short answer
- Choose Databricks metric views if most of your governed data is in Databricks. Definitions sit next to the data, inherit Unity Catalog permissions and lineage, and work with Databricks dashboards and Genie as well as external BI tools.
- Choose Strategy Mosaic if your metrics must combine Databricks with Snowflake, BigQuery or other sources, or you want definitions that stay put if you change warehouse.
- Check the native options first in any case. Snowflake has semantic views and Power BI has semantic models; see semantic layers for AI.
How they compare
| Databricks metric views | Strategy Mosaic | |
|---|---|---|
| Scope | Data in Databricks Unity Catalog | Multiple data platforms and sources |
| Vendor | Databricks | Strategy (formerly MicroStrategy) |
| Defined in | SQL DDL or Catalog Explorer, with a YAML specification | Mosaic’s own modelling environment |
| Governance | Unity Catalog permissions, lineage and audit | Its own governance, integrating with Unity Catalog and other catalogs |
| Consumers | SQL, notebooks, AI/BI dashboards, Genie, alerts, Power BI, Tableau, Sigma | BI, AI and productivity tools through SQL (JDBC), DAX, REST and Python |
| Advanced metrics | Window measures, parameters, joins across star and snowflake schemas, materialization | Varies; check the current documentation |
| Cost | Part of Databricks; you pay for the compute that runs queries | Separate licence (named-user pricing, according to Strategy) |
| Lock-in | Definitions tied to Databricks | Definitions tied to Mosaic |
What usually decides it
- How many platforms hold governed data. One platform favours native metric views. Several, with metrics that join across them, favour an independent layer.
- Who consumes metrics. If people mostly use Databricks dashboards, Genie and one BI tool, native is simpler. Many tools across the enterprise strengthen the case for Mosaic.
- Existing investment. Organisations already running MicroStrategy have models and skills that carry over to Mosaic.
- Cost and operating effort. A separate semantic layer is another licence and another system to run. Make sure it removes more work than it adds.
The short version
Metric views for a Databricks-centred estate, with no extra product to buy; Strategy Mosaic when one set of metrics must span several platforms or outlive your current warehouse. Either way, agree metric definitions with the business before choosing the tool.
Related: semantic layers for AI, what is Unity Catalog and Databricks vs Snowflake. Our AI-ready data service builds governed metrics that AI assistants can use.
Sources
- Unity Catalog metric views, Databricks documentation
- Strategy announces general availability of Strategy Mosaic, Strategy
- Strategy launches the world’s first universal intelligence layer, Strategy blog (vendor claims)
Not sure where your own platform stands? Take the free 3-minute data platform scorecard, or see how we work on Databricks. Unfamiliar terms are explained in the data engineering glossary.