Microsoft is the only company in the cloud era that didn't have to build a data audience — it inherited one. Every Fortune 500 enterprise was already a Microsoft customer in 2010 when Azure launched. Every analyst was already using Excel. Every IT department was already running Active Directory and SQL Server. Microsoft's data strategy from the very beginning has been: take that captive audience and progressively raise the share of their data spend that flows to Microsoft, by bundling new data products into existing Microsoft enterprise agreements.

In plain English: Microsoft doesn't usually win on product. It wins on contract. Power BI is the most-deployed BI tool in the world, not because it has better visualizations or a better semantic layer than Tableau or Looker (it doesn't), but because it shows up at near-zero marginal cost in any Office 365 E5 license. Microsoft Fabric, launched in 2023, is the same playbook applied to the entire data stack: warehouse + lake + ETL + BI + governance, packaged as a single SKU sold to people who already buy Microsoft.

Origin Story: From Excel to Azure to Fabric

Microsoft has been in the data business longer than almost anyone — SQL Server has shipped continuously since 1989 (originally licensed from Sybase, then rewritten as a fully Microsoft product in the SQL Server 7.0 release in 1998). Excel, dating to 1985, is by some measures still the most widely used analytical tool in the world.

The cloud chapter started with Azure. Microsoft announced "Project Red Dog" — the codename for what became Windows Azure — at PDC 2008, and the platform launched commercially in February 2010. For the first several years, Azure was a slow-moving second-place follower to AWS, focused mostly on porting Windows-stack workloads to the cloud. The decisive turn came when Satya Nadella became CEO in February 2014. Nadella reframed Azure as "the cloud for everyone," dropped the Windows-only attitude, embraced Linux and open source, and started using Microsoft's enterprise muscle aggressively. By the late 2010s Azure had passed Google Cloud and become a credible #2 to AWS.

The data-stack story starts with Power BI. Power BI began life inside Excel as an add-in — "Power Query," "Power Pivot," "Power View" — in the 2010-2013 era, and was spun out as a standalone product in July 2015 under the name Power BI. The strategy was clear from day one: bundle it into Office 365, sell it for $9.99/user as Power BI Pro, undercut Tableau on price by an order of magnitude. By 2019 Gartner had moved Power BI into the unambiguous lead position in the Magic Quadrant for Analytics and BI Platforms, and it has stayed there.

Underneath Power BI, Microsoft assembled the rest of the data stack piece by piece:

• Azure SQL Data Warehouse launched in 2016, was rebranded as Azure Synapse Analytics in 2019, and added a data lake / Spark / pipelines story to become Microsoft's "unified analytics" platform.
• Azure Data Lake Storage (ADLS Gen2) layered hierarchical namespaces on top of Azure Blob Storage to give it lake semantics.
• Azure Data Factory became the managed ETL / pipelines tool.
• Azure Purview (now Microsoft Purview) was the catalog and governance layer.
• Microsoft Fabric, launched in May 2023 and made GA in November 2023, then absorbed all of the above into a single SaaS product with a single capacity-based pricing model. Fabric is the most ambitious thing Microsoft has done in data in fifteen years.

Their Data Products

• Microsoft Power BI — The dominant BI tool in the world by deployments, by a wide margin. Strong with Excel-fluent business analysts, weaker with the SQL-fluent analytics engineering crowd. The fastest-shipping product in the BI category by far — monthly releases, hundreds of features per year, all driven by the bundle's economics: Power BI doesn't have to be good, it just has to be there.
• Azure Blob Storage — Microsoft's object storage. The S3 / GCS equivalent, with the additional twist that Azure Data Lake Storage Gen2 is just Blob Storage with hierarchical namespace turned on. ADLS Gen2 is the substrate underneath nearly every Azure data product, including Fabric's OneLake.

Other Microsoft data products that don't yet have a wiki page:

• Azure Synapse Analytics — The cloud data warehouse and unified analytics platform. Originally a cloud version of the SQL Server MPP appliance (PDW), now repositioned as a "pool of compute on top of ADLS." Still sold as a standalone product but increasingly being absorbed into Fabric.
• Microsoft Fabric — The 2023 unified data platform. Combines a lakehouse (built on Delta Lake), a warehouse, Spark notebooks, real-time analytics (built on the Kusto / Azure Data Explorer engine), Data Factory pipelines, Power BI, and a governance layer, all sold as capacity units. The unifying storage layer is OneLake, which is essentially "ADLS with a single global namespace and a Delta-first opinion." Fabric is the most ambitious piece of data product packaging any vendor has done in years.
• Microsoft Purview — The catalog, lineage, and data governance layer. Plays the role of Lake Formation or Unity Catalog in Microsoft's stack. Also the home of Microsoft's data security and DLP tooling.
• Azure Data Factory — The managed ETL service. Visual pipeline builder with hundreds of connectors. Increasingly being subsumed into Fabric Data Factory.
• Azure Databricks — A first-party Microsoft re-sale of Databricks, launched 2017. Technically a Databricks product, but it ships under the Azure brand and bills through Azure invoices, which makes it Microsoft's main "real" lakehouse story. The unusually deep partnership between Microsoft and Databricks is one of the strangest and most consequential vendor relationships in the modern data stack.

The Strategy: The Bundle Always Wins

There is one strategic principle that explains essentially everything Microsoft does in data: the bundle always wins, eventually, against the standalone product. Microsoft has watched this play out, in its own favor, dozens of times over forty years. Excel beat Lotus 1-2-3 because it shipped with Office. PowerPoint beat Harvard Graphics for the same reason. Internet Explorer crushed Netscape (right or wrong) because it shipped with Windows. Teams crushed Slack — a far better product, by most accounts — because it shipped with Microsoft 365. Power BI is currently doing the same thing to Tableau.

Microsoft Fabric is the next escalation. The bet is that Fabric will do to the data stack what Power BI did to BI: undercut every standalone vendor in every category by a generous margin, package it all into a single SKU that an existing Microsoft enterprise customer can add with one signature, and let the bundle economics do the rest. The targets are clear: Snowflake (warehouse), Databricks (lakehouse), Confluent (streaming, via Real-Time Intelligence), Fivetran (ingest), and Collibra/Alation (governance), all in one product.

Whether this works is the big open question of the mid-2020s. Power BI is a successful precedent, but the data stack is more technical and more sticky than BI — you can switch BI tools in a quarter, you cannot easily migrate a petabyte warehouse. Microsoft is making the bet that new data workloads will land in Fabric by default at any Microsoft shop, and that the warehouse incumbents will be slowly squeezed from the bottom.

Honest Market Take

Microsoft's moat is the enterprise relationship, not product excellence. That's the unfashionable truth. Power BI's UI is creakier than Tableau's. Synapse is technically behind Snowflake and BigQuery. Purview is less polished than Collibra. Microsoft Fabric, in its first 18 months, shipped with serious early-stage rough edges that the warehouse vendors enjoyed pointing at. None of this matters very much, because none of it has to be the best — it has to be already in the customer's contract.

Where Microsoft is genuinely excellent: integration with the rest of the Microsoft ecosystem. Power BI's connection to Excel, to Teams, and to the M365 identity layer is deeper than any third-party tool will ever achieve. Fabric's connection to Power BI is similarly deep. If your company is already running on Microsoft 365, the friction of adopting Fabric is much lower than the friction of adopting Snowflake, and that friction differential is almost the entire game.

The risk for Microsoft is the same risk Microsoft has always faced: a market where a fast-moving best-of-breed competitor builds an ecosystem of fans and developers that the bundle cannot quite displace. Snowflake and Databricks have done this for warehousing and lakehousing the way Slack did for chat. Microsoft's job over the next five years is to make sure the data category goes the way of Teams (Microsoft wins eventually) rather than the way of AWS (Microsoft is permanently #2). The honest answer is that nobody knows yet.