Skip to content
Happitu Happitu
Docs /Leaderboard
Open app

Leaderboard

Track performance rankings across agents, scorecards, and teams with comprehensive leaderboard analytics.

Leaderboard showing agent rankings and metrics

The Leaderboard provides performance rankings based on quality metrics. Use it to:

  • Identify top performers
  • Spot coaching opportunities
  • Compare team performance
  • Track improvement over time
  • Recognize achievement
  1. Click Leaderboard in the main navigation.
  2. The default view shows agent rankings.
  3. Use tabs to switch between ranking modes.

Rank individual agents by performance metrics.

Columns displayed:

  • Rank position
  • Agent name
  • Group assignment
  • Quality metrics (see below)
  • Trend indicator (improving/declining)

To view agent rankings:

  1. Select Agents tab.
  2. Apply filters as needed (date range, group, workspace).
  3. Sort by clicking column headers.

Rank performance on specific scorecards.

This view shows:

  • Scorecard name
  • Average score across all evaluations
  • Pass rate
  • Number of evaluations
  • Top and bottom performing agents on this scorecard

To view scorecard rankings:

  1. Select Scorecards tab.
  2. Choose which scorecard to analyze.
  3. See performance breakdown.

Compare performance across teams.

This view shows:

  • Group name
  • Team size
  • Aggregate metrics
  • Best performing agents in each group

To view group rankings:

  1. Select Groups tab.
  2. Compare team performance side-by-side.
  3. Drill down into specific groups.

The mean quality score across all evaluations.

  • Range: 0-100%
  • Calculation: Sum of all scores ÷ number of evaluations
  • Usage: Primary performance indicator

Higher AQS indicates better overall quality.

Percentage of evaluations that meet the passing threshold.

  • Range: 0-100%
  • Calculation: Passing evaluations ÷ total evaluations × 100
  • Usage: Measures consistency

High pass rates with high AQS indicate strong, consistent performance.

Percentage of interactions that received evaluations.

  • Range: 0-100%
  • Calculation: Evaluated interactions ÷ total interactions × 100
  • Usage: Measures evaluation completeness

Target coverage varies by organization, typically 10-20%.

Average duration of interactions.

  • Format: Minutes and seconds
  • Usage: Efficiency metric
  • Context: Compare against quality scores

Very low AHT with low quality may indicate rushing. Very high AHT may indicate inefficiency.

Customer sentiment analysis:

  • Average sentiment: Overall customer satisfaction
  • Positive rate: Percentage of positive interactions
  • Negative rate: Percentage of negative interactions
  • Sentiment trend: Improving or declining over time

Sentiment complements quality scores to show customer perception.

Select the time period for rankings:

  • Today
  • Yesterday
  • Last 7 days
  • Last 30 days
  • This month
  • Last month
  • Custom range

Rankings update based on interactions within the selected period.

Narrow rankings to specific teams:

  1. Click the Group filter.
  2. Select one or more groups.
  3. Rankings refresh to show only selected groups.

Filter by workspace:

  1. Click the Workspace filter.
  2. Select a workspace.
  3. See rankings for that scope.

Rank by specific scorecard:

  1. Click the Scorecard filter.
  2. Select a scorecard.
  3. Rankings use that scorecard’s metrics.

By default, the leaderboard sorts by:

  • Agents: Average Quality Score (descending)
  • Scorecards: Average score (descending)
  • Groups: Average Quality Score (descending)

Click any column header to sort by that metric:

  • First click: Sort descending (highest first)
  • Second click: Sort ascending (lowest first)
  • Third click: Return to default

When agents have identical scores:

  • Higher pass rate ranks higher
  • More evaluations ranks higher
  • Alphabetically by name

Each agent shows a trend indicator:

  • ↑ Green: Improving (score increased 5%+)
  • → Gray: Stable (within 5%)
  • ↓ Red: Declining (score decreased 5%+)

Trend compares current period to previous period of same length.

Click an agent’s name to see:

  • Historical performance chart
  • Score progression over time
  • Evaluation history
  • Detailed metrics breakdown

Download leaderboard data:

  1. Configure your view (filters, rankings, date range).
  2. Click Export button.
  3. The data downloads as a CSV file compatible with Excel and other spreadsheet tools.

Exports contain:

  • All visible columns
  • Current filter configuration
  • Rankings data
  • Metadata (export date, parameters)

Note: Scheduling is not available for the Leaderboard. To schedule automated reports, create a saved view in Explore and configure a schedule there.

  • Compare fairly: Use filters to compare similar roles/time periods
  • Look at trends: One-time rankings matter less than trends
  • Consider context: High scores with low coverage may not represent true performance
  • Balance metrics: Don’t focus only on AQS—consider pass rate and sentiment
  • Celebrate success: Recognize top performers publicly
  • Identify agents in bottom quartile for coaching
  • Compare top and bottom performers to identify best practices
  • Track improvement after coaching interventions
  • Use scorecard rankings to identify specific skill gaps
  • Compare groups to identify training needs
  • Set group goals based on leaderboard benchmarks
  • Use group rankings in team meetings
  • Share best practices across groups
How often does the leaderboard update? Rankings update in real-time as new evaluations are published. Historical data is recalculated nightly.
Can agents see the leaderboard? Depending on your organization's settings, agents may see their own ranking or team rankings. Check role permissions in Settings.
Why isn't an agent showing on the leaderboard? Agents need at least one evaluation in the selected time period to appear. Check date range and filters.
Can I customize which metrics appear? Yes. Click **Customize** to show/hide columns and reorder them to match your priorities.
How are new agents ranked? New agents appear after their first evaluation. Their initial trend indicator shows as stable until sufficient history exists.