The Claude Data Analyst Skill activates senior analyst mode: SQL, dashboards, segmentation, exploratory analysis, insight narrative. Turns raw data into decisions for non-technical stakeholders.
What it does
- SQL queries for BigQuery/Postgres/Snowflake.
- EDA: outlier, missing, distribution detection.
- Dashboard design (Metabase, Looker, Tableau).
- Executive narrative with prioritized insights.
- Cohort, churn, LTV, attribution analysis.
When to activate
- Ad-hoc analysis for immediate decision.
- Executive dashboard design.
- Investigation of drops in core metrics.
- Quarterly reporting with conclusions.
Core instructions
SKILL.md mandates: validate schema before query; never extrapolate without enough n; present 3 levels (what happened, why, what to do); do not use p-values without causal context.
Use cases
- Detect weekly retention drop with root-cause.
- 12-month cohort analysis for SaaS.
- Acquisition dashboard by channel.
- A/B post-mortem with confidence intervals.
Limitations
- Needs data access (MCP or copy/paste).
- Does not replace data engineer for pipelines.
- Causality requires designed experiments, not observation.
Frequently asked questions
Do I need SQL to use it?
No to activate it, yes to understand the output. If you do not know SQL, ask it to explain the query in plain English.
Combines with Product Manager Skill?
Yes. PM defines hypotheses, Data Analyst validates with data.
Does it work with Code Interpreter?
Yes, especially in Claude.ai with CSV/Excel analysis.
