Case Study: Gong

Overview

Field	Value
Company	Gong.io
Industry	Revenue Intelligence / Sales AI
Headquarters	San Francisco, CA
Deployment Date	October 2025
MCP Version	Anthropic MCP spec
Enterprise Readiness Score	⭐⭐⭐⭐

Use Case

Gong introduced MCP support to solve bidirectional enterprise AI fragmentation. Gong's own AI features now pull in data from external tools (inbound), while external enterprise agents — Microsoft Copilot, HubSpot AI, Salesforce — can query Gong's revenue intelligence data (outbound).

This turns Gong into a shared intelligence layer across the full revenue stack rather than a siloed application.

Connected Systems

System	Type	Direction	Auth Method	Notes
Salesforce	CRM	Both	OAuth	Bidirectional; write-back gated by Gong logic
HubSpot	CRM / Marketing	Both	OAuth	Part of announced MCP integration
Microsoft Copilot	AI Agent	Inbound query	MCP Protocol	External agents query Gong via MCP Server
Internal Gong data	Revenue Intelligence	Read	MCP Gateway	Deal data, call intelligence, pipeline

Architecture Pattern

Bidirectional Bridge

Gong exposes an MCP Gateway (for inbound agent queries) and an MCP Server (for outbound context enrichment). This makes MCP a bidirectional protocol rather than a one-way data source.

Governance Controls

Access management: MCP Gateway controls which external agents can query Gong
Risk classification: Revenue data access gated by deal-level permissions
Approval gates: Write-back to CRM requires Gong-side logic validation
Data policies: Not publicly detailed beyond access controls

Outcomes

Announced as a platform capability; adoption metrics not yet publicly disclosed
Enables cross-platform AI workflows without manual data export or API stitching

Source Links

Evidence Quality Notes

Systems and architecture are confirmed by press release. Governance details and outcome metrics are not publicly disclosed. Score of 4/5 reflects named systems + architecture with partial governance evidence.

Gong.io — Bidirectional Revenue Intelligence