User Analytics

Overview

The User Analytics page provides comprehensive visibility into individual user activity on the platform, including token usage, cost metrics, performance data, and request volumes. This feature enables administrators to track resource consumption, analyze usage patterns, identify optimization opportunities, and allocate costs across different users or departments.

Accessing User Analytics

1. Navigate to the left sidebar navigation panel
2. Click on "Monitoring" in the menu options
3. Select "Users Analytics" from the submenu (indicated by the arrow icon)
4. The User Analytics interface will load, displaying a tabular view of all user activity data

User Analytics Interface Components

Time Range Selection

Located near the top of the interface, this control allows filtering of user data by different time periods:

• 24H: Displays user activity from the past 24 hours
• 7D: Displays user activity from the past 7 days
• 1M: Displays user activity from the past month
• 3M: Displays user activity from the past 3 months
• All: Displays all historical user activity data

Export Functionality

• EXPORT ALL DATA: Button located next to time filters that allows exporting the current view of user analytics data for external analysis, reporting, or billing purposes

View Controls

Located at the top-right corner of the table:

• Search: Magnifying glass icon to search through user data
• Table View: Toggle between different table view modes
• Fullscreen: Expand the user analytics table to fullscreen mode

User Analytics Table

The main component displaying detailed user activity information in tabular format with the following columns:

Column	Description	Example Value
Selection Checkbox	Allows selection of individual or multiple user records	☐
User	User identifier or email address	user@example.com
Total Input Tokens	Number of tokens consumed in user queries/prompts	7500
Total Output Tokens	Number of tokens generated in responses	120
Total Cost	Aggregated cost of all user activity	0.0789
Average Latency	Average response time for user requests in seconds	3.0975
Requests	Total number of requests made by the user	5

Pagination Controls

Located at the bottom of the table, these controls allow navigation through large datasets:

• Rows per page: Dropdown selector to choose how many rows to display (typically 10)
• Page navigation: Buttons to move forward and backward through pages
• Page indicator: Shows current page and total pages (e.g., "Page 1 of 1")

Data Metrics Explained

Token Usage Metrics

• Input Tokens: Represents the volume of text sent to the AI models in user prompts/queries

  • Higher numbers indicate more verbose or frequent inputs
  • Useful for identifying users with inefficient prompt strategies
  • Example range: 1,000-10,000 tokens per time period

• Output Tokens: Represents the volume of text generated by AI models in responses

  • Indicates the verbosity of AI responses to user queries
  • Useful for measuring model output efficiency
  • Typically lower than input tokens for most use cases
  • Example range: 100-1,000 tokens per time period

Cost Metrics

• Total Cost: Calculated based on token usage (both input and output) and the specific models used
  • Directly correlates with total token usage
  • Varies based on model pricing tiers
  • Example range: $0.01-$1.00 per time period

Performance Metrics

• Average Latency: Measures the average response time for all user requests
  • Lower values indicate better performance
  • May vary based on query complexity and model used
  • Example range: 1-5 seconds

Usage Volume

• Requests: Count of distinct interactions initiated by the user
  • Provides insight into overall platform adoption
  • Can be compared against token usage to identify efficiency patterns
  • Example range: 1-100 requests per time period

Using User Analytics Effectively

Cost Management

• Budget Allocation: Track costs by user to establish departmental budgets
• ROI Analysis: Calculate return on investment based on user productivity vs. cost
• Cost Optimization: Identify users with high token-to-request ratios for prompt engineering guidance
• Usage Forecasting: Project future costs based on current usage trends

Performance Monitoring

• Latency Tracking: Identify users experiencing slower than average response times
• Resource Optimization: Balance workloads to maintain consistent performance
• Experience Improvement: Target optimizations for heavy users with performance issues

User Behavior Analysis

• Power Users: Identify users with high engagement (frequent requests)
• Efficiency Patterns: Compare input vs. output token ratios across users
• Training Opportunities: Identify users who might benefit from prompt optimization training
• Adoption Tracking: Monitor user onboarding and continued engagement

Comparative Analysis

• Team Benchmarking: Compare usage patterns between teams or departments
• Model Efficiency: Analyze which users achieve better input-to-output token ratios
• Cost Efficiency: Identify users achieving the most value (requests) for the lowest cost

Best Practices

Regular Monitoring

• Review user analytics weekly to identify usage trends
• Monitor cost metrics before billing cycles to avoid surprises
• Track new user adoption and early usage patterns
• Compare period-over-period metrics to identify changing patterns

Filtering Strategies

• Use time range selectors to focus on relevant periods:
  • 24H: For daily operational monitoring
  • 7D: For weekly trend analysis
  • 1M/3M: For month-over-month comparisons
  • All: For historical performance analysis

Data Export and Analysis

• Export data regularly for integration with business intelligence tools
• Create custom reports combining user analytics with other business metrics
• Maintain historical exports for compliance and trend analysis
• Use exported data for custom visualizations and presentations

User Education

• Share analytics with users to promote cost-awareness
• Provide targeted training for users with inefficient token usage
• Create internal benchmarks and best practices based on high-performers
• Develop prompt engineering guidelines based on efficiency patterns

Troubleshooting Common Issues

Issue	Possible Cause	Solution
Missing user data	New user or no activity in time range	Expand time range or check user account status
Unusually high costs	Inefficient prompts or excessive usage	Review user's prompt patterns and implement usage limits
Inconsistent latency	Complex queries or system load	Analyze specific requests and optimize prompt design
Disproportionate token ratio	Sub-optimal prompt design	Provide prompt engineering guidance to the user
Low request count despite high tokens	Large, infrequent operations	Evaluate use case and suggest breaking into smaller requests

Integration with Other Monitoring Features

The User Analytics interface is part of a broader monitoring ecosystem that includes:

• Dashboard: Provides overview metrics of platform performance
• Requests: Tracks specific API and user requests in detail
• Audit Trail: Records system changes and administrative actions
• Platform Monitoring: Monitors technical aspects of the platform infrastructure

These components work together to provide comprehensive visibility into all aspects of platform usage, performance, and cost management.