Monitoring & Anomaly Detection

Fload's Monitoring Agent continuously watches your metrics 24/7 and alerts you when something unusual happens — revenue spikes, download drops, conversion rate changes, or any other significant deviation from normal patterns.

What it does

The Monitoring Agent runs automated anomaly detection across all your metrics:

• Continuous monitoring — checks every metric every day for unusual patterns
• Automatic alerts — notifies you via email, Slack, or Discord when anomalies are detected
• Surge and decline detection — identifies both positive spikes and negative drops
• Multi-metric coverage — monitors revenue, downloads, subscriptions, engagement, ad performance, and more
• Contextual analysis — understands seasonal patterns, day-of-week effects, and historical trends

How it works

Detection algorithm

Fload's anomaly detection uses statistical analysis to determine what's "normal" for each metric based on historical data. It requires at least 30 days of data to establish a baseline.

For each metric, the algorithm:

1. Analyzes historical patterns (trends, seasonality, variance)
2. Calculates expected range for today's value
3. Flags values that fall significantly outside the expected range
4. Assigns a severity score based on how far the value deviated

Anomaly types

Severity levels

• High — deviation is very significant and requires immediate attention
• Medium — noticeable change but may be explainable
• Low — minor deviation that's likely noise

Monitoring dashboard

The Monitoring section provides a unified view of all detected anomalies:

Tabs

• All — every anomaly (surges and declines)
• Surges — positive anomalies only
• Declines — negative anomalies only
• Dismissed — anomalies you've reviewed and dismissed

Anomaly details

Click any anomaly to see:

• Chart visualization — metric over time with the anomaly highlighted
• Expected vs. actual — what the value should have been vs. what it was
• Context — dimensional breakdowns (which countries, products, or campaigns drove the change)
• Suggested actions — what to investigate or do next

Notifications

Configure how you want to be alerted when anomalies are detected:

Channels

• Email — daily digest or real-time alerts
• Slack — post to a designated channel
• Discord — post to a designated channel

Filters

• Severity threshold — only notify for high-severity anomalies
• Anomaly type — only surges, only declines, or both
• Specific metrics — choose which metrics to monitor (e.g., only revenue and downloads)

Go to Monitoring → Settings to configure notification preferences.

Use cases

Catch issues early

• Sudden drop in downloads → app store visibility issue or technical problem
• Revenue decline → payment processing issue or pricing problem
• Trial conversion drop → onboarding flow broken or paywall not showing

Capitalize on wins

• Organic traffic surge → feature press coverage or viral moment
• Conversion rate spike → winning experiment or seasonal trend
• Ad ROAS improvement → campaign optimization working

Validate changes

• After a major app update, monitor for unexpected behavior
• After a pricing change, track revenue and conversion impact
• After a marketing campaign, see if metrics move as expected

Dismissing anomalies

Once you've reviewed an anomaly and understand why it happened, you can dismiss it. Dismissed anomalies are archived and won't clutter your active monitoring dashboard.

Dismissals help the system learn what's normal for your app — for example, if you run a big promotion every Black Friday, the system will eventually stop flagging the annual revenue spike as unusual.

Availability

Anomaly detection is available on all tiers:

• Free — limited to basic metrics
• Startup — full coverage across all connected metrics
• Pro — full coverage + advanced notification options

Monitoring requires at least 30 days of historical data to activate.