In the Age of AI, You Need a Data Lake in the First 100 Days
Microsoft Fabric, now generally available, has delivered on the promise that motivated this article. By consolidating Microsoft’s data stack into a single platform, Fabric enables natural language queries for business users to ask questions of their company’s data directly, and thanks to delta lake technology, companies are positioned to leverage machine learning and AI applications on top of their consolidated data. This platform provides both near-term benefits in the form of data visibility and automated BI, and long-term benefits in establishing the foundation for advanced analytics. The case for building a data lake in the first 100 days has only grown stronger since we first made it.
Why Data Lakes?
The first 100 days are critical to the success of an acquisition. The early period after deal closing is the most important for launching the value creation roadmap and ensuring alignment between PE sponsors and portfolio executives. A key component of that roadmap is identifying needed infrastructure investments and the right timing for their implementation. In terms of the data investments that should be prioritized in the first 100 days to yield both immediate and long-term ROI, we recommend a data lake-centered approach.
Listen to Brick and Kate discuss this in the podcast episode below.
Brick recommends taking all data from transactional systems across the platform company and its add-ons and consolidating it into a data lake. From there, analysts and report writers can gain access almost immediately, and a semantic layer on top of it allows technical executives to reach that data and meaningful outputs quickly, without waiting for full system integration across every acquired entity.
Data Lakes for Buy-and-Build Strategy
With the prevalence of the buy-and-build strategy in private equity and the midmarket, platform companies face the complexity of integrating data from disparate companies and source systems. A data lake enables integration in as little as a few weeks, as opposed to the several months a traditional data warehouse can take to build. This approach does not preclude later migration of some or all data into a traditional data warehouse for enhanced governance and master data management. It simply removes the requirement that you wait for that migration before anyone can see anything useful.
Data Lakes and AI Tools
Beyond bringing order to the sometimes chaotic data landscape of buy-and-build investments, a data lake sets companies up to leverage AI analytics capabilities that are now fully operational. Power BI Copilot and natural language query tools within Microsoft Fabric have arrived and are delivering on their promise. Companies that consolidated their data into a lake before these tools were widely available gained an immediate advantage: they could activate AI analytics from day one rather than scrambling to build the foundation after the fact.
That advantage compounds going forward. Your ability to effectively leverage AI depends on clean underlying data. As you plan your data strategy, read our perspective on investing in data fundamentals before pursuing AI and our interviews with CTO Andy Scott and CIO John Manzanares.
The Benefits of a Data Lake in the First 100 Days
For buy-and-build companies, quick consolidation of data across business units, source systems, and subsidiaries into a data lake greatly speeds access to valuable insights. A data lake provides immediate access to data for early reporting as well as a sandbox for analysts. It is a highly scalable and flexible data repository with the option for future migration to a traditional data warehouse when governance and quality management require it. And with AI technologies now mature and operational, a consolidated data lake environment provides a direct path to putting them to immediate use, without waiting to build the foundation later.
Blue Margin’s managed data service builds data lakes on Microsoft Fabric using an agile approach that delivers the first working reporting layer in weeks. If your organization is planning an acquisition or managing a growing portfolio of data sources, contact our team to discuss what building a data lake in the first 100 days would look like for your situation.
