Engineering-led mid-market: SQL-first reverse-ETL over marketer-self-service tooling
When a mid-market organization ($25M–$500M revenue, 50K–1M profiles) has data engineering or IT — not marketing — holding RACI-Accountable for the CDP decision, and already operates a CDW with dbt-modeled data, this recommendation names the correct evaluation lane: SQL-first composable reverse-ETL, not the marketer-self-service or point-solution paths the organization may otherwise default to.
Why this recommendation exists. archetype.engineering-led-mid-market-composable-evaluator documents the presenting symptoms directly: segments built in Snowflake and emailed as CSVs to marketing, churn models built in dbt with no activation path, and 15 hand-maintained custom sync jobs consuming sprint capacity. Each of these is a reverse-ETL gap, not a CDP-selection gap in the marketer-self-service sense.
What this recommendation is not. It does not recommend a no-code, marketing-operated CDP — that tooling is optimized for the distinguishing signals of the marketing-led counterpart archetype (UI accessibility, marketer self-service), which this archetype explicitly does not have. It does not pick between Hightouch and Fivetran Activations; both are named as primary candidates for further evaluation on connector depth, schema-evolution handling, and pipeline governance.
Scope limitation. This recommendation assumes an existing CDW with dbt pipelines already owned by data engineering. An organization making a net-new CDW purchase alongside its CDP decision is evaluating a different, larger problem than this archetype covers.