Archetype — Engineering-Led Mid-Market Composable Evaluator

Organizational profile

Mid-market organization ($25M–$500M revenue, 50K–1M customer profiles) where a data engineering team or IT function is RACI-Accountable for the CDP evaluation and subsequent activation operations. Marketing requests audience segments through a data-request or ticketing process; it does not operate the CDP tooling directly.

Distinguishing signals from marketing-led counterpart. In contrast to archetype.marketing-led-mid-market-composable:

• A data engineering team already operates a CDW (Snowflake, BigQuery, or Databricks) with existing dbt pipelines and SQL-defined data models.
• Marketing is RACI-Responsible (executes campaign sends) but not Accountable — the data engineering lead or data/analytics manager holds Accountable.
• Evaluation criteria weight toward connector catalog depth, schema evolution handling, pipeline observability, query cost implications, and reverse-ETL operational overhead, rather than UI accessibility or no-code audience building.

Distinguishing signals from enterprise engineering-led archetypes. Unlike archetype.aep-heavy-enterprise-evaluating-composable and archetype.salesforce-ecosystem-enterprise-evaluating-cdp:

• No incumbent enterprise CDP (no AEP, no Salesforce Data 360 investment to protect or augment).
• Making a net-new CDP selection, not evaluating augmentation of an existing platform.
• Budget-constrained ($100K–$400K/year), below enterprise AEP/Salesforce licensing thresholds.

Trigger pattern

Usually precipitated by one or more of:

• The data engineering team has a CDW with customer event data and marketing is asking for activation to external channels without the data team rebuilding pipelines per destination.
• A point-solution CEP (Klaviyo, Braze) operates on incomplete or stale customer data because it lacks direct CDW connectivity; marketing wants segments derived from live CDW models.
• A new ML propensity model was built in the CDW and marketing wants to activate on its scores — but no pipeline exists to push scores to email or ad platforms.

Common presenting symptoms

• "Our data team builds segments in Snowflake and emails them as CSVs to marketing for upload to Klaviyo — there has to be a better way."
• "We built a churn prediction model in dbt but marketing can't use it because there's no connection to our email platform."
• "We're maintaining 15 custom sync jobs ourselves instead of using a connector catalog. It's eating half the sprint."

Recommended direction

SQL-first composable reverse-ETL: audience definition in SQL or dbt, activation via reverse-ETL to the destination catalog. Hightouch (400+ destinations, SQL/dbt audience definition) and Fivetran Activations (formerly Census — acquired May 2025; data-contract-first syncs, KG node: vendor.census) are the primary candidates for this archetype's evaluation stage. See org-dim.operational-profile.engineering-led-cdp for selection guidance.