Why your quality assurance scorecard is driving your team crazy (and how to fix it)

Executive Summary

The average attrition in Customer Service teams across all industries is still between 20% and 45% in 2025.

Half of the reasons frontline agents give for leaving could be avoided. Better quality assurance, more feedback, and helpful tools can make a difference.

Existing QA scoring systems aren't truly 'objective.' They often harm company culture by 'nitpicking' instead of supporting customer service agents.

We can improve customer service systems for our teams by using real measures, PII masking, redaction, or hiring an external consultant.

At VerticalAI, we can help you develop a truly objective QA system with 100% call coverage for your customer service team.

---

In our talks with customer service teams in Perth, we found that most firms track 'conversation success' with blunt metrics.

For example, customer service agents may be required to ask a specific question at the end of the conversation.

"Is there anything else we can help you with?"

Agents are scored on whether they recite a closing script and the customer doesn't ask for anything else.

This is considered a successful conversation, even if the customer's issue remains unresolved, or they waited on the phone for an hour prior to being served.

Some larger teams might have their manager, company founder, or QA manager listen in to a sample of their calls. They then rate these chats using a quality assurance scorecard.

We've found that this process is managed primarily through spreadsheets, and that only around 1-3% of all calls receive attention.

These measures are often the best that the team can do.

It is unfeasible for teams to hire someone to listen to more calls most of the time. Listening to every call for quality assurance would mean investing in additional full-time employees, essentially doubling the size of the team.

There is also a real issue with existing objective 'score card systems'. If handled poorly, they can harm company culture, cause burnout, and lead to a lack of recognition or appreciation. These are all direct contributors to attrition rates (6%, 17% and 11% respectively).

This score card, while using numbers for scoring, is also highly subjective. It is up to the QA manager, founder, or whoever else is making the scoring, to figure out if they scored a 6 or 7 for empathy. Furthermore, based on a 2% sampling rate, if this agent has made 100 calls that week, they would only have received feedback on 2 calls on average.

Score cards like these are also ways that a difficult personality or toxic manager can assert power. Because of the subjective metrics used, it's easy for a difference in opinion to lead to an agent being scored lower than their peers.

When done correctly, these score cards also miss important metrics that a QA manager is not positioned to make a judgement on. For example:

• "Is this call an accurate representation of the average customer's experience?"
• "Is there an up-sell opportunity in the call that has been missed?"
• "Should this query have been escalated directly to a technical support team?"

The low sampling rate can lead a QA manager to miss the forest for the trees. Often, they focus too much on an arbitrary metric like "Our calls should have 80% quality based on the score card." What they miss is that 40% of the calls could be automated or better handled through a website or app.

Customer service agents, skilled at handling tough situations, often answer simple questions. They deal with inquiries about opening times, confirm past appointment booking dates, or spend hours each week looking up the status of an invoice or location of an asset.

Furthermore, our research shows that up to 50% of agent attrition are factors that are within our control. We can make jobs more enjoyable and rewarding for our teams, which can happen with improved feedback and QA systems. When agents feel supported and recognized, they perform better.

From our research and experiences, 10% of attrition is due to negative company culture, or a bad relationship with their leader, which can be improved with better QA systems.

Burnout is also a major cause of attrition. This factor gets worse when we understand little about the calls customer service teams handle, and when we create KPIs that don't match the calls we are handling.

Finally, 19% of attrition happens because of a lack of recognition and appreciation. This, along with dissatisfaction, can improve with better, more timely, and objective feedback.

So, how can we improve this process?

Start by tracking more objective metrics like time to resolution, or how long customers are on hold.

Another way is to review calls collaboratively, which avoids them being used as nitpicking sessions. We can do this by anonymising calls through PII masking for both the customer service agent and the customer.

Finally, you can hire an external consultant, or enlist the help of a third party that can give more objective feedback, which avoids personal opinions and biases.

---

Many companies are sitting on a wealth of information hidden within their call recordings. From our research, around 97% of all call recordings are never reviewed. This is a massive missed opportunity to improve the quality of your customer service.

Here at Vertical AI, we've helped our clients get 100% coverage and visibility on the quality of their calls. Because we help them analyse all their calls, we are also able to identify opportunities previously missed.

Here are some opportunities we can help you find:

• Reasons to improve the website, mobile app, or social media strategy to reduce inbound calls.
• Reasons to update email templates and SMS systems to make sure customers don't have to call for minor queries about their orders or appointments.
• We can quantify the number of missed opportunities for up-sells or cross-sells, and start training the team to spot them.
• By applying live sentiment analysis on calls, we can proactively avoid burnout, attrition, and customer dissatisfaction.
• We can use AI to automate simple, low-value call flows, so your team can focus on more complex issues.

Need help? At VerticalAI, we analyze 100% of calls to deliver objective QA that reduces attrition, helps you save costs, and improves your customer service.

Email us at: hello@verticalai.com.au

Visit our website at: https://verticalai.com.au

---

References:

2025 Contact Centre Best Practice Report, 6th Edition Bk Top 5 Customer Service Trends and Priorities that Matter Most in 2025