Alerts & Trends
Your Dataset fills on its own — Alerts and Trends are how you watch it without polling. Alerts push new matches to you; Trends chart how topics move over time.
Alerts
An Alert is a persistent query over your Dataset. Whenever a new record matches — a keyword hit, a tag, a search — the Alert fires and delivers the record to the destination you chose. Set them up under Alerts in the dashboard.
- Slack — matches arrive as messages in the channel you connect
- Webhook — matches POST to any URL: n8n, Make.com, Zapier, or your own service. See Webhooks for the payload, signatures, and retry behavior.
On the trial
Trial organizations get up to 5 Alerts. They fire normally during the 14 days and pause when the trial expires — nothing is deleted, and they resume when you subscribe.
Trends
A Trend computes a time series over your Dataset — how often a topic, tag, or query appears per day — so you can spot spikes in mentions early and see the shape of demand instead of reading every record. Trends are available on the paid plan.
The series is sparse: only intervals with at least one match are stored, so a 30-day trend whose query matched on a single day has one point. Treat missing buckets as zero.
Limits
Alerts and Trends share one quota: 5 on the trial (Alerts only, since Trends are paid) and 10 combined on the paid plan. See Plans & limits.
From your agent (MCP)
Your agent can manage both with the alert_* and trend_* tools — alert_create, alert_list, alert_update, alert_delete, and alert_get_deliveries to inspect what fired and where it went (same names for trends). Before pointing an Alert at a destination, the agent can call get_account — its integrations block shows whether Slack is connected (and to which channels) and lists your webhook endpoints with the ids destination.webhookIds expects. Or just ask: "Alert me in Slack when someone mentions our brand".