
Marketing Attribution Advisor
Helps agents work out which marketing drives conversions and reconcile conflicting dashboard numbers
- Repository
- coreyhaines31/marketingskills
- Stars
- 57,951
- Licence
- MIT
- Files
- 6 files
What it does
Once installed, the agent approaches questions about which marketing channels actually drive sales differently. It explains that attribution is an estimate rather than a single truth, compares models such as first-touch, last-touch, linear and data-driven, and advises showing two models side by side for long sales cycles. It can also help reconcile mismatched numbers between ad platforms and analytics tools, and point to other skills for event tracking, pixels, CRM pipelines or AI search. A second, build-oriented track covers self-hosted first-party attribution when the user controls the site.
Great for
- Marketers whose Google, Meta and analytics dashboards disagree and need a defensible way to explain results
- Founders trying to work out their real customer acquisition cost or which channel to fund
- Data or growth teams choosing between attribution models or planning incrementality and media mix tests
- Developers who want to build their own first-party attribution tracking on a site they control
Skip it if
- Teams expecting one objective 'true' revenue number per channel: the skill itself says attribution is directional and every model embeds an opinion
- Anyone relying on the output for high-stakes financial decisions without checking the underlying conversion data, since models can mislead when data volume is thin
Try asking
- “My Google Ads says 120 conversions but GA4 says 80. Which channel should I trust and how do I reconcile them?”
- “Should I use first-touch or last-touch attribution for a B2B product with a six-month sales cycle?”
- “Help me set up first-party attribution so I can stitch my SavvyCal bookings back to their original source.”
- “Explain to my CEO why branded search looks like our best channel and what we can do about it.”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 6 original files, the MIT licence and an attribution note.
attribution.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip attribution.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip attribution.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Give the agent your business type, sales cycle and primary conversion so it can follow the optional product-marketing context file it checks for.
- Ask for first-touch and last-touch side by side rather than a single model when your sales cycle is long.
- Share the conversion counts from each tool and your deduplicated conversion source so the agent can explain the gap instead of guessing.
What's inside
- evals/evals.json
- references/attribution-models.md
- references/by-business-type.md
- references/first-party-tracking.md
- references/measurement-paradigms.md
- SKILL.md
Redistributed unmodified from coreyhaines31/marketingskills under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.