How do you analyze customer support data in Metabase?
Help desks like Zendesk and Intercom hold the record of every customer conversation your team handles. To analyze them in Metabase, you sync the tool into a database, map its objects to a shared support model, and build dashboards on top. Metabase has no native connector for these tools, so the sync comes first.
TL;DR — Most help desks share the same shape: tickets/conversations, messages, agents, customers, tags, SLA/status history, and CSAT. Model that shape once and most dashboards and metrics port across tools — only the source-specific fields differ.
Which tools does this cover?
This pattern applies to help desks and shared inboxes, including:
- Zendesk — tickets, comments, SLA policies, satisfaction ratings
- Intercom — conversations, parts, Fin AI resolutions, contacts
- Front — shared-inbox conversations, comments, assignments
- Freshdesk — tickets, conversations, agents, groups, CSAT
- Gorgias — ecommerce tickets tied to orders and customers
- Kustomer — omnichannel conversations and timelines
- Pylon — B2B issues across Slack, Teams, and email
- Plain — threads, customers, and tenants via a GraphQL API
- LiveAgent — multi-channel tickets and call logs
- Jitbit — IT help desk tickets and categories
- Crisp — live-chat conversations and sessions
- Drag — Gmail shared-inbox cards and boards
What is the shared support data model?
Almost every tool maps onto these entities. Model them as clean tables, not raw connector JSON:
| Concept | Zendesk term | Intercom term | Used for |
|---|---|---|---|
| Ticket / conversation | Ticket | Conversation | The unit of customer work |
| Message | Comment | Conversation part | Response times |
| Agent | Agent | Admin/teammate | Workload |
| Customer | Requester | Contact | Volume by account |
| Tag | Tag | Tag | Drivers, topics |
| Status / SLA event | Audit / SLA policy | State change | Resolution time, breaches |
| Satisfaction | Satisfaction rating | Conversation rating | CSAT |
The single most important field is a reliable status/SLA history. With it you can compute true resolution time, time-in-status, reopen rate, and SLA attainment. Without it, those metrics must be caveated.
How do you connect a help desk to Metabase?
- MCP route (AI-assisted) — pair the tool's MCP server with the Metabase MCP server for fast, exploratory questions. Treat it as exploratory, not governed reporting.
- Pipeline route (warehouse-backed) — sync the tool into a database with a managed connector, dlt, or the API, then build durable dashboards.
See the per-tool setup on the Zendesk andIntercom pages, or the support analytics overview for the full pattern.
What can you analyze across support tools?
- Responsiveness — first-response time, resolution time
- Volume & backlog — created vs. solved, open work aging
- SLA — attainment and breaches against your targets
- Quality — CSAT, reopen rate
- Drivers — load by channel, tag/topic, and account
Which dashboards should you build?
- Support overview — volume, FRT, resolution time, backlog, CSAT
- SLA & response time — attainment, breaches, aging
- CSAT & quality — satisfaction trend, reopen rate, drivers by tag
- Agent & team performance — workload balance and handle time
Common mistakes
Treating tools as fundamentally different.→ 80% of the model is shared; standardize it.
Skipping the status/SLA history.→ Point-in-time snapshots can't reconstruct resolution time or reopens.
Counting bot replies as first response.→ Define first response as the first human or first meaningful reply.
Related
Analytics
Integrations
FAQ
Can I use the same dashboards for Zendesk and Intercom?
Mostly yes, if you model both onto the shared support schema. Metric definitions and chart structure port across; only source-specific fields (e.g. Zendesk SLA policies, Intercom conversation parts) differ.
Do I need status history?
For volume, backlog, and first-response time, no. For resolution time, reopen rate, and SLA attainment, yes — sync the status/SLA event history.
Is there a native Metabase connector for support tools?
No. Sync to a database first, then connect Metabase to that database.