marketing-attribution-analyst — Claude Code subagent for attribution
A subagent that works out which channels actually earn credit for conversions and how to split marketing budget accordingly.
npx claude-code-templates@latest --agent business-marketing/marketing-attribution-analystWhat marketing-attribution-analyst does
This subagent handles the question of which channel or touchpoint gets credit for a conversion, and whether that credit holds up. It works from your own tracking, channel spend and CRM revenue data. Model is sonnet; tools are Read, Write, Glob, Grep, WebFetch, WebSearch.
It starts by asking about your analytics stack (GA4, GTM, CDP or warehouse), attribution window, available data sources, business model and spend level. Then it checks current platform documentation through search and reads any local tracking code or SQL you share. The method it recommends matches your confirmed data rather than defaulting to the fanciest model on the shelf.
For quarterly budget calls it leans on marketing mix modeling over aggregate spend and outcome data. Causal checks come from incrementality tests such as geo holdouts and matched-market tests. Multi-touch attribution stays for campaign-level tuning. MMM needs roughly 1-2 years of weekly observations and real spend variation across channels, so a large budget alone will not make it identifiable.
What it helps with
- find out whether paid social really drives the credited 40% of revenue
- pick a measurement method that fits spend level and data maturity
- validate a channel with a geo holdout before shifting budget
- tell a real channel-mix shift from a tracking break after migration
- justify a channel budget split to the board
Example prompts for the agent
- Our last-click model credits paid social with 40% of revenue. Check whether that lift is real and propose a test.
- Help me choose an attribution method for our spend level and tracking data.
- GA4 numbers changed a lot after our tagging migration. Is this a real channel shift or a measurement problem?
How to install marketing-attribution-analyst
1claude-code-templates
npx claude-code-templates@latest --agent business-marketing/marketing-attribution-analyst2Manually: put the files in place
.claude/agents/marketing-attribution-analyst.mdFor all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention
Manually: put the files in place
.codex/agents/marketing-attribution-analyst.tomlCodex defines subagents in TOML: name, description, developer_instructions. Move the Markdown text into developer_instructions by hand. For all projects use ~/.codex/agents/
Manually: put the files in place
.cursor/agents/marketing-attribution-analyst.mdCursor also reads .claude/agents/
Step by step, with commands for each platform: Claude Code subagents: how to create one, call it and keep context clean
Questions about marketing-attribution-analyst
How do I install marketing-attribution-analyst in Claude Code?
Run npx claude-code-templates@latest --agent business-marketing/marketing-attribution-analyst, then confirm the file landed at .claude/agents/marketing-attribution-analyst.md.
What does marketing-attribution-analyst do?
It builds multi-touch attribution, measures marketing mix impact and validates channels with incrementality testing. That helps you decide which channel to trust and how to allocate budget.
What data does the subagent need?
Your tracking stack, attribution window, spend by channel and campaign, and revenue or CRM data. MMM also needs 1-2 years of weekly observations with spend variation across channels.
