Gong MCP Server

Gong

Gong is a conversation intelligence platform that analyzes sales calls and emails to provide insights into customer interactions and sales effectiveness. Use it to improve sales strategies and team performance by understanding communication patterns and market feedback.

Data Model

• Call index — lightweight metadata and summaries for scanning activity over a date window or by id.
• Rich call exports — transcripts, AI topics, trackers, interaction stats, and action items (higher payload cost—use only when content is required).
• Scorecard templates — rubrics managers use to grade conversations.
• Completed scorecards — reviewer submissions tied to calls and reps.
• Workspace users — licensed Gong seats for attribution and filtering.

Workflow hints

• Browse recent calls: constrain by from_date_time / to_date_time.
• Hydrate one call: request by call_id when you already know the target.
• Deep analysis: use the extensive-call export with JSON filters (call ids, users, dates).
• Coaching loops: filter answered scorecards by reviewee or date after picking template ids from the scorecard catalog.

Gotchas

• Date windows: Listing the call index requires explicit bounds unless you already have a call id.
• POST-heavy exports: Both extensive calls and answered scorecards expect POST bodies with filter objects—not query strings.
• Scorecard dates: Filters use YYYY-MM-DD, not full ISO-8601 timestamps.
• Pagination: Paged collections expose cursor-based continuation via the cursor parameter.
• Workspace scope: Data is scoped to your Gong workspace; use workspace_id to target a specific one.