A Claude Code skill that adds consumer research expertise on top of the GWI (Global Web Index) MCP server. Makes anyone using Claude with GWI immediately more effective at pulling, interpreting, and presenting consumer insights.
This skill teaches Claude how to:
- Query GWI effectively — prompt construction patterns, audience definitions, follow-up strategies
- Interpret results correctly — index scores, percentages, sample sizes, statistical caveats
- Format findings professionally — persona cards, comparison tables, research reports, executive summaries
- Avoid common pitfalls — one question per call, trend splitting, US English, Core-only limitations
You need the GWI MCP server (gwi_spark_ai) configured in your Claude Code environment. This requires a GWI account with Spark API access.
# Add the marketplace
/plugin marketplace add LeonHartmann/gwi-research-skill
# Install the skill
/plugin install gwi-research-skillgit clone https://github.com/LeonHartmann/gwi-research-skill.git
cp -r gwi-research-skill/skills/gwi-research ~/.claude/skills/skills/gwi-research/
├── SKILL.md # Core skill — workflows, rules, quick reference
└── references/
├── query-patterns.md # Query construction, 54 markets, audience filters
├── data-interpretation.md # Index scores, 5 metrics, rebasing, caveats
└── output-formats.md # 7 templates (quick lookup → trend narrative)
The skill activates when you ask about: audience research, consumer insights, market research, demographic analysis, media planning, brand health, audience profiling, competitive audience analysis, trend identification, GWI queries, consumer trends, digital behavior, or purchase intent.
"Tell me about esports viewers in Germany"
"Compare Gen X and Millennials on social media usage in the UK"
"What motivates eco-conscious consumers to purchase sustainable products?"
"Build me an audience profile of EV owners in the US"
MIT