Myth: Marketing needs a data lake to be data-driven

Myth: Marketing needs a data lake to be data-driven

Someone always proposes a data lake for the marketing org. The correction: marketing's data is structured and modest in scale — a warehouse with dbt covers ~95% of use cases without the lake's complexity tax.

— A thread on warehouse-vs-lake for marketing-sized data.
— A dbt writeup on modeling marketing data without a lakehouse.
— A cost-and-headcount comparison of running a lake vs. a warehouse for a marketing team.
— A checklist of the rare cases (raw event firehose, ML feature store) where a lake earns its keep.

Skip if you're ingesting billions of raw events. Essential if someone is scoping a lake for your email data.

That's the stack for this week. Forward to a teammate.