Traffic notifications help monitor your website's traffic patterns by alerting you to significant spikes or dips. The system uses machine learning and anomaly detection to identify statistically significant changes in your traffic patterns.

For anomaly detection to work effectively, your website needs at least 28 days of traffic history. The system compares current traffic levels against this historical data, flagging changes that fall outside 99% of your normal traffic distribution.

Finding and Analyzing Anomalies:

1. Navigate to Analytics > Traffic
2. Select your desired date range from the dropdown menu
3. Look for pulsating exclamation marks on the traffic analysis chart
4. Click warning icons to view detailed information

When reviewing an anomaly, you'll see:

• Date of occurrence
• Traffic volume during the event
• Top affected pages (up to two) with traffic percentage
• Contributing traffic channels
• For spikes: Most visited pages during the increase
• For dips: Pages with the largest decrease in traffic

Using Anomaly Data:

• Identify successful traffic drivers
• Spot underperforming marketing channels
• Adjust your marketing strategy based on insights
• Test different approaches to improve traffic

The system continuously improves through machine learning, becoming more accurate as it collects more data. You can provide feedback on anomaly notifications by using the "Is this helpful?" feature at the bottom of each anomaly detail page.

To provide feedback:

1. Click the thumbs up/down button
2. Add optional comments in the text field
3. Click Send

This feedback helps improve the accuracy and usefulness of the anomaly detection system.

Related Articles

Previous Articles