Performance dashboard

The performance dashboard helps manage labeling operations in Labelbox projects. It reports the throughput, efficiency, and quality of the labeling process. These analytics are reported at the overall project level and at an individual level. You can use filters to focus on specific information and dimensions. The performance of your data labeling operation is broken down into four components:

• Individual
• TEQ (Throughput, Efficiency, and Quality)
• Participation
• Instructions quiz

Each component has unique views to help you understand the overall performance of your labeling operation. Use the mouse to select and display details for any graph value.

​

Filters

Filters at the top of the performance dashboard allow you to analyze relevant subsets of data. When active, filters apply to all metric views, including Throughput, Efficiency, and Quality. Currently, the following filters are provided:

​

Throughput view

The Throughput view provides insight into the amount of labeling work being produced and helps you answer questions like:

• How many assets were labeled in the last 30 days?
• [What is the review time and rework time for labeled assets?](#review-time)
• What is the average amount of labeling work being produced?

The metrics shown above are available for all members of the project and for individual members. The throughput view displays the following metrics:

​

Tracking states for timing

A data row goes through many different states, which are tracked in the timer log.

​

Labeling time

Labeling time increments when:

• The user skips or submits the asset in the labeling queue

​

Review time

Review time increments when:

• The user views the asset in the data row browser view
• The user views the asset in the review queue

​

Rework time

Rework time increments when:

• The user submits the asset in the rework queue
• The user navigates to the review queue, selects edit, and approves/rejects the asset
• The user edits and saves the asset in the data row browser view

​

Inactivity

Only active screen time is recorded. Inactivity pauses the timer. When your system is idle for more than five (5) minutes, the timer pauses. This affects labeling time, review time, and rework time. The timer resumes when system activity resumes.

​

Efficiency view

The efficiency view helps you visualize the time spent per unit of work, per labeled asset, or per review. These metrics help answer questions such as:

• What is the average amount of time spent labeling an asset?
• How can I reduce time spent per labeled asset?

These metrics are available for individual project members and at the project level (for all project members). The efficiency view includes these charts: Label states in time calculations:

• “Skipped” time is excluded from labeling time.
• “Deleted” label time may be included or excluded using filters.
• “Abandoned” (unsaved) work is counted toward labeling time.

​

Quality view

The quality view helps you understand the accuracy and consistency of the labeling work being produced. These metrics answer questions like:

• What is the average quality of a labeled asset?
• How can I ensure label quality is more consistent across the team?

These metrics are available for individual project members and for the project itself, which summarizes performance for all project members. The metrics shown are available for each individual and for the project level, which summarizes the performance of members of the project. The quality view includes the following charts: Below are the descriptions for each chart in the quality view. Graphs display the components appropriate for your project. Neither benchmarks nor consensus display when they’re not enabled for your project.

​

Individual member performance

You can also view individual metrics for each team member that has worked on the project. The performance metrics are separated by intent ( labeling and reviewing) and are shown as distinct views in the table. Team members are listed individually in separate rows; members only appear if they actively performed tasks during the selected period; that is, they only show up if they labeled data rows or reviewed labels. Labeling metrics include: These metrics that appear on the Reviewing tab:

​

Instruction Quiz analytics

The Instruction Quiz tab provides comprehensive analytics for quiz performance when you have enabled quizzes in your ontology instructions. These analytics help you understand how well labelers comprehend your instructions and identify areas where additional clarification may be needed.

​

Access quiz analytics

To view quiz analytics for your project:

1. Navigate to your project
2. Go to the Performance tab
3. Select the Instruction Quiz tab at the top of the page
4. Use the date range picker to filter analytics by time period

​

Overview metrics

The dashboard displays key performance indicators at the top:

​

Visual analytics

The dashboard includes two key distribution charts: Pass Attempt Distribution

• Shows how many attempts labelers need to pass the quiz
• Helps identify if the quiz difficulty is appropriate
• Example: If most labelers pass on the first attempt, the quiz may be too easy; if most need many attempts, it may be too difficult

Score Distribution

• Displays the range of scores across all attempts
• Shows how scores are distributed on the 1-5 scale
• Helps identify if most labelers are performing well or struggling

​

Question performance table

This table shows detailed metrics for each quiz question: Use this table to identify problematic questions:

• Questions with low average scores may need clearer instruction content
• Questions requiring many attempts suggest the topic needs better explanation
• Negative improvement trends indicate labelers aren’t learning from retakes

​

User performance table

This table shows individual labeler performance: Use this table to:

• Identify labelers who may need additional training or support
• Track improvement over time for individual users
• Understand which labelers are struggling with the material

​

Using analytics to improve your quiz

Based on the analytics data, you can take several actions: If overall pass rates are low:

• Review your instructions for clarity and completeness
• Consider breaking down complex concepts into simpler explanations
• Add more examples to illustrate key points

If specific questions have low scores:

• Revise the related section in your instructions
• Ensure the question accurately tests the intended knowledge
• Adjust the expected answer to be more flexible

If labelers need many attempts to pass:

• Simplify your quiz questions or make instructions more explicit
• Add practice examples in your instructions
• Consider reducing the passing threshold if appropriate

If improvement trends are negative:

• Review the feedback provided by the AI scoring
• Ensure questions test understanding, not memorization
• Consider whether the quiz is testing the right concepts

​

Import timer impact

Importing labels affects labeling time. When you import:

• Ground truth labels: no label time is recorded on the Data rows tab or the performance dashboard for that label. Label time is displayed as a zero (0).
  • When team members modify the label, time is recorded as review time
• For model-assisted learning (MAL) pre-labels, no label time is recorded on the Data rows tab or the performance dashboard for that label. Label time is displayed a zero (0).
• When team members open the data row in the editor and click Edit, the time spent before selecting Skip or Submit is recorded as labeling time