A centralized storage system that holds raw data in its native format, structured or unstructured, at any scale, deferring transformation until the data is needed for a specific use.
A data lake stores raw data in whatever format it arrives: structured tables, semi-structured JSON, unstructured text, images, log files, video. Unlike a data warehouse that requires data to be cleaned and structured before loading, a data lake accepts everything and leaves the transformation for later.
The appeal is flexibility. Store everything now, figure out what you need later. When a new analytics use case emerges, the raw data is already available.
The data swamp problem
Flexibility without governance produces a data swamp: a massive repository where nobody knows what data exists, where it came from, whether it is accurate, or how to use it. Data lakes that lack metadata management, access controls, and cataloging become storage costs with no analytical value.
What most people get wrong
Teams build data lakes because they can, not because they have a plan for the data. A data lake is useful when you genuinely have diverse data types that do not fit a warehouse schema and you have the data engineering capability to process raw data when needed. For organizations that primarily run structured analytics, a data warehouse or lakehouse is a better fit.
The lakehouse convergence
The market has moved toward convergence. Data lakehouses combine the raw storage flexibility of a lake with the querying and governance capabilities of a warehouse. For most new implementations, the lakehouse architecture has become the starting point rather than choosing between a pure lake or a pure warehouse.