Bruin Academy
Guide
Marketing Analyst
Go from zero to a working marketing analyst that pulls Google Ads, Klaviyo, and GA4 into DuckDB - then lets an AI agent answer real attribution and CAC questions. The AI agent is the coder - every step is a plain-English prompt.
What
Learn the modern marketing analyst stack in one sitting. Pull ad spend, email engagement, and web analytics into one local database, then ask an AI agent attribution and CAC questions your team is usually stuck guessing at.
- Pull Google Ads, Klaviyo, and GA4 into a single local DuckDB database
- Build staging and report layers so the AI agent has a unified cross-channel surface
- See blended CAC by channel across paid, owned (email), and organic
- Ask natural-language questions like "which ad campaigns drove high-LTV customers?" and get SQL-backed answers
How
- The AI agent does the typing - you prompt it in plain English and approve what it does
- Bruin CLI scaffolds the pipeline, and ingestr connectors handle the data pulls
- Bruin MCP gives Claude Code, Cursor, or Codex direct query access to your database
- Pipeline-scoped
AGENTS.mdplusbruin ai enhanceteach the agent your domain
Before you start
- An AI coding tool - Claude Code, Cursor, or Codex (the agent does the typing)
- Bruin CLI - the agent installs this for you in Step 1
- Access to a Google Ads, Klaviyo, and Google Analytics 4 account (or use the sandbox accounts in Step 3)
Guide overview
- 1Set Up Your WorkspaceInstall your AI coding tool, then have it install Bruin CLI and wire up MCP for you. Every step from here on is a plain-English prompt.
- 2Scaffold Your ProjectCreate an empty Bruin project with DuckDB as the local database - the workspace your ad, email, and web analytics data will land into.
- 3Ingest Google Ads, Klaviyo, and GA4Have the AI agent write ingestr assets for three marketing data sources and pull ad spend, email engagement, and web analytics into DuckDB.
- 4Build Staging + Report LayersAdd a clean middle layer on top of the raw ad, email, and web data, then build a final report table that unifies channel-level performance - the analysis-ready table your AI agent will query.
- 5Teach the AI Your DomainUse bruin ai enhance to document every column, then add a pipeline-scoped AGENTS.md with marketing context so the agent reasons about attribution, CAC, and LTV like a real analyst.
- 6Ask the Real Marketing QuestionPut the stack to work. Combine Google Ads, Klaviyo, and GA4 to answer the kind of cross-channel question your BI tool can't handle.
Resources
Get help & contribute