Glossary
Data Lake
A storage system designed to hold raw, unstructured, or semi-structured data at scale. In dealership architectures, a lake typically holds raw vendor extracts before transformation into the unified warehouse layer.
Most modern dealer data platforms use a lakehouse, which collapses the lake and warehouse into one governed environment, instead of a separate lake and warehouse.
What it looks like in a dealership
- A nightly DMS extract lands as raw files in the lake before any cleaning or modeling happens.
- Website clickstream and call-tracking events land unstructured, then get modeled into customer journeys later.
Common questions
What is the difference between a data lake and a lakehouse?+
A lake is cheap raw storage without the transactional guarantees and SQL layer of a warehouse. A lakehouse adds those guarantees on top of the lake, so you get raw storage and reliable analytics in one place.
Is a data lake enough on its own?+
Rarely. A lake without governance becomes a swamp: raw files nobody can query. Dealers need the warehouse semantics on top, which is why the lakehouse pattern won.