Analyze the structural shape of the winning AI answer across several
related keywords/topics in one call: does each lead with a list, how
long is the opening, how many sources it cites. Use this for content
planning across a topic cluster, e.g. before writing several related
pieces meant to get cited, instead of calling analyze_citation_structure once
per topic.
Read-only: no side effects, safe to retry. Costs 1 quota unit per
keyword in the batch (free tier is 30 units/month shared across all
the metered tools, so up to 30 keywords total that period if nothing else
is used). A per-keyword provider error doesn't fail the whole batch -
that keyword's entry just carries an "error" field instead.
Returns: {"results" (list, one {"keyword", ...same shape as
analyze_citation_structure, or "error"} per keyword, in the order given),
"summary": {"topics_analyzed", "topics_requested", "list_led_count",
"avg_sources_cited", "avg_community_pct" (average source_mix.community_pct
across the analyzed topics)}}.
Args:
keywords: topics/queries to analyze, e.g. ["how to reduce churn",
"churn rate benchmarks", "reduce customer churn saas"]. Max 10.
country: market to read the answers in, e.g. "Italy". Defaults to
"United States".
language: language code, e.g. "it". Defaults to "en".
engine: "chat_gpt" (default), "gemini" or "perplexity", as in
analyze_citation_structure. One answer per topic.
ConnectorAPI key