Desk Score: The Quality Revolution in Customer Service with AI – Blip | Blip Help

Index

Overview

Welcome to the new era of human service in Blip Desk. We present Desk Score, a powerful feature powered by Blip Copilot.

Desk Score automates the quality assurance (QA) of your operation. Instead of manual sampling audits, our Artificial Intelligence analyzes 100% of finished conversations, offering a fair score and instant constructive feedback.

"With Blip Copilot Score, your service improves with every conversation."

How It Works

The process is transparent and automatic. As soon as a ticket is finished in Blip Desk, Blip Copilot kicks in:

Full Reading: The AI reads the entire history of the conversation between agent and customer.
Criteria Analysis: The service is cross-referenced with 6 fundamental quality pillars.
Scoring: A score of 1 to 5 stars is generated.
Feedback: The system highlights what was done with excellence and suggests practical improvement tips for points of attention.

The 6 Evaluation Pillars

Desk Score evaluates service based on technical and behavioral criteria. See what makes up the score:

Time Metrics (Configurable)

These parameters can be customized for each Bot in the Copilot settings, respecting the reality of your operation.

First Response Time: Evaluates the agent's agility in welcoming or the first handling after the ticket is distributed.
Average Handling Time (AHT): Measures total efficiency, from the beginning to the end of the conversation.

Quality and Behavior Metrics

Availability: Checks the flow of the conversation, penalizing long unjustified interruptions or the absence of the agent during active service.
Conduct and Empathy: Analyzes the "tone of voice". Was the agent cordial? Did they maintain a professional stance? Did they demonstrate empathy with the customer's problem?
Spelling and Grammar: Evaluates linguistic correctness, proper use of Portuguese (or the local language), and clarity in communication.
Assertiveness and Size: Checks if the answers were concise or detailed enough for the complexity of the doubt presented.

How Copilot performs the evaluation: The Score Logic

To ensure a fair and consistent evaluation, Blip Copilot uses a methodology based on the Presumption of Excellence.

The score calculation follows three fundamental principles:

Start with Maximum Score: Every service starts, by default, with a score of 5 (five). We start from the principle that the agent is prepared to perform an excellent job.
Deduction by Evidence: The AI analyzes the conversation looking for specific deviations in the established criteria. Only when Copilot identifies, based on evidence in the text or metadata (time), that a criterion was not met, does it subtract a fraction of the score corresponding to that item.
Safety Criterion: The system operates in favor of the agent. If the AI does not find enough elements in the conversation or clear parameters that justify the loss of points, no score is withdrawn. That is, if there is no detectable error, the score for that criterion is fully preserved.

Summary: The score is not built from scratch; it is preserved. The agent's goal is to maintain their 5 stars, avoiding the discounts that the AI applies when detecting opportunities for improvement.

How to Configure Desk Score

Desk Score is a tool contained exclusively within Blip Copilot. By subscribing to Blip Copilot, you will have full access to Blip Score.

To ensure that the evaluation is fair and consistent with the reality of your operation, the manager or bot administrator must configure the time parameters.

Blip Copilot allows you to customize what is considered a "good time" for your business.

Access the Portal: Enter the Blip portal and select the desired service Bot.
Copilot Menu: In the lateral bar of the "Helpdesks" menu, locate and click on Blip Copilot.
Evaluation Settings: Go to the Score tab.
Define SLAs (Service Level Agreements):
- First Response Time: Define the ideal maximum time (in minutes) for the agent to send the first message.
- Total Service Time: Define the expected average duration for resolving a ticket.
Save: Click save to apply the rules to the next services.

After that, simply release the permissions for SCORE on the agent permissions page.

Note: The Conduct, Spelling, Availability, and Response Size metrics are already calibrated by Blip's AI based on market best practices and do not require initial manual configuration.

For the Agent: Autonomy and Continuous Evolution

Desk Score is each agent's personal "coach".

In the new "My Metrics" menu in Desk, the agent has immediate access to their performance. There is no need to wait for the manager's monthly feedback; evolution happens now.

Detailed Scores: View of the score for each recent ticket.
Practical Tips: The AI doesn't just point out the error, it teaches.
Agent Feedback: If the agent thinks the score or details do not match reality and disagrees, they can signal it in the menu at the bottom of the screen, so we can always improve our algorithms.

Example: In this service, response size was not adequate; the AI identified it and gave a tip directly on how to proceed in the next tickets.

For the Manager: 360º View of the Operation

Forget manual Quality Assurance spreadsheets. The new Consolidated Quality Report delivers the health of your service in real-time.

To access the management report, access it within the “Helpdesks” menu in the new side menu, the “Service Quality” option:

Overall Score: Team average score.
Score evolution in the period: Visualize the behavior of service quality over time.
Tickets by criterion: Analyze the highlighted criteria identifying those that have opportunities for improvement in the operation.

Detailing of evaluated services: Here you view each service, the ticket ID, the scores, and you can access the evaluation the agent received in full.

Daily performance by agent: Team average score on the filtered days.

Expected Benefits

Standardization: Ensure all customers receive the same level of excellence.
Agility in Feedback: Fast course correction for new agents.
Reduction of Operational Cost: Less time spent by supervisors manually auditing conversations.
Motivation: Natural gamification of the service process through the score.

Frequently Asked Questions (FAQ)

Can I adjust what is considered a "good response time"?

Yes! Each operation has its particularity. In the Blip Copilot settings menu, you define the time limits (SLA) for both the first response and the total duration of the service, ensuring that the AI evaluates your team according to the reality of your business.
Does the AI score replace the satisfaction survey (CSAT)?

No. They are complementary metrics. While CSAT measures the customer's perception and feeling about the service, Desk Score measures the agent's technical and behavioral compliance regarding best practices. Having a high CSAT and a low Desk Score, for example, may indicate that the agent is friendly but needs to improve technical efficiency.
Can I customize or create new evaluation criteria?

At the moment, no. The current 6 criteria were carefully selected and validated during months of testing with real customers. They cover the fundamental pillars of excellent service. However, we are constantly evolving, and our users' feedback will guide the inclusion of new customization possibilities in the future.
Does the AI always get the evaluation right?

The technology is extremely precise, but it is not infallible. Desk Score uses state-of-the-art generative AI and has undergone thousands of tests in the beta phase to ensure high reliability. However, in cases of very atypical conversations or very specific slang, variations may occur. We work continuously so that these cases are rare exceptions.
What should I do if I notice the AI evaluated a service incorrectly?

Your collaboration is fundamental to the evolution of our intelligence! In the new Desk menu, we encourage agents and managers to use the "thumbs up/down" (positive/negative) icons on each evaluation. This direct feedback helps us calibrate and improve the analysis algorithms specifically for the context of your operation.

Conclusion

Desk Score is not just about numbers; it's about the continuous development of your team and ensuring that every customer receives the best version of your brand. It is Blip's commitment to helping you every day deliver value to your customers.

For more information, visit the discussion on the subject at our community or videos on our channel. 😃