Anomaly Alerts - Appeeky REST API & MCP Server

Appeeky Analytics checks each app once a day. When installs, revenue, trials or install-to-paid conversion move far from a normal day, you get an anomaly alert.

Drop in installs for Mtioon
Installs dropped 70% vs the usual day.
Installs on 2026-10-24: 12 vs 40 usual (-70%)

How It Works

Baseline: the median of the previous 14 days, in the app’s time zone.
Metrics: installs, revenue, trials and install-to-paid conversion.
Direction: a spike (well above the baseline) or a drop (well below it).
Once per day: at most one alert per app, metric and day.

Where Alerts Go

Where	What you see
Telegram	A message to every chat paired with the Appeeky Alerts bot. See Sale Alerts to pair one.
Attribution → Overview	The latest anomalies above the KPIs
Co-founder	Each alert is recorded as an anomaly signal for the app, so the co-founder can explain it in chat and in briefs. Drops in money metrics are treated as high priority.

API

Method	Path	Returns
`GET`	`/v1/analytics/apps/:appId/anomalies?range=30d`	`{ alerts }`

{
  "data": {
    "alerts": [
      {
        "appId": "app_9f2c",
        "storeAppId": "123456789",
        "metric": "installs",
        "direction": "drop",
        "day": "2026-10-24",
        "current": 12,
        "baseline": 40,
        "changePct": -70,
        "message": "Installs dropped 70% vs the usual day."
      }
    ]
  }
}

The analytics_anomalies tool on the Appeeky MCP server returns the same list.