Why this matters
CS teams do not need another dashboard. They need a brief: what changed, why the account is at risk, and what to do before renewal.
What you get
- Accounts renewing soon
- Usage trend and seat utilization
- Error/support friction
- Recommended CS motion for each account
Walk through it
You
Which renewals are at risk this quarter?
Agent
I’ll find renewal cohorts and join them with usage and friction signals.
POST /v1/projects/:project_id/analytics/query {
project_id: "default",
hogql: "SELECT properties.account_id AS account, argMax(properties.renewal_date, timestamp) AS renewal_date, count() AS recent_events FROM events WHERE timestamp > now() - INTERVAL 60 DAY GROUP BY account HAVING renewal_date != '' ORDER BY renewal_date ASC LIMIT 50"
}The output
The agent returns one brief per account: risk level, evidence, and action. It should be written for a CSM, not for an engineer.
Setting it up
Emit renewal_upcoming or attach renewal date to account snapshot events. Keep account IDs consistent across usage, errors, and billing.
Variations
- “Only include enterprise renewals.”
- “Draft the CSM email for the top three at-risk accounts.”
- “Create a weekly renewal-risk digest.”