QA – Scorecards Scoring Guide

📊 Instructions for Using the Scorecard in Oversai

The Scorecard is an evaluation tool that allows you to assess agent performance based on a defined score goal.

  • If the evaluated person meets or exceeds this goal → they are marked as “Passed”

  • If they don’t reach it → they are marked as “Failed”


🧩 Scorecard Structure

1. Sections and Questions

  • The Scorecard is organized into sections.

  • Each section is broken down into evaluation questions.

2. Section Weights

  • Each section has a percentage weight.

  • The total of all section weights must add up to 100%.

  • The weight of each section is defined manually.

3. Questions and Scoring

  • Each question has a maximum possible score (the value of the highest-scoring option).

  • The maximum score of a section is the sum of the maximum scores of its questions.

4. Question Option Types

  • default → Adds only the indicated points.

  • fail_section → If selected, the entire section is scored with 0 points, regardless of the other answers.

  • fail_all → If selected, the overall Scorecard score is automatically set to 0.

  • N/A (Not Applicable) → Excludes the question from the calculation (removes its points from the total available, adjusting the final score proportionally).


🧮 Final Score Calculation

The final score is obtained by summing the results of each section, weighted by its percentage.

Step-by-step example:

  1. A section has a maximum of 10 points.

  2. The agent earns 5 points.

    Section calculation:

    (5/10) *100 = 50%

  3. If that section is weighted at 40%, then:

    50% * 40 = 20%

  4. Repeat the calculation for each section.

  5. The final Scorecard score is the sum of all section results.


🏆 Example Outcome

  • Agent’s final score: 90%

  • Defined score goal: 80%

  • Since the agent exceeds the goal → Result: “Passed”


✨ Benefits of Using Scorecards in Oversai

  • Objective and consistent evaluations.

  • Flexibility to adapt the structure and weights according to your policies.

  • Ability to include critical criteria (fail_section, fail_all).

  • Clear visibility of agent performance with simple, transparent metrics.



Still need help?

Contact us

Quality Assurance - QA