What is a data warehouse for marketing?

A marketing data warehouse centralizes data from all marketing tools — CRM, ads, email, website analytics, social media — into one queryable database. It enables cross-channel analysis impossible with individual tool reports. Why you need one: your marketing data lives in 10-20 tools. Each tool reports differently. You cannot answer "which channel drives the most revenue" without combining CRM closed-deal data with ad spend data with web analytics. A data warehouse makes this possible. Platform options: BigQuery (Google, serverless, pay per query — best for most), Snowflake (enterprise, best performance at scale), PostgreSQL (self-hosted, free, good for under 100GB), and Redshift (AWS ecosystem). Data pipeline: extract data from tools (Fivetran, Airbyte, or custom API scripts), transform it (dbt is the standard), load into warehouse. This ETL pipeline runs daily or hourly. Key tables to build: unified_contacts (merge CRM + email + web identity), channel_attribution (which channels drive conversions), campaign_performance (spend, impressions, clicks, conversions per campaign), and revenue_attribution (tie marketing activity to closed revenue). Cost: BigQuery is nearly free for small volumes (<1TB queried/month = $5). Fivetran starts at $120/month. dbt Cloud starts free. Total: $200-500/month for a complete pipeline.