JUST AS IMPORTANT: DON ' T ONLY TRACK WHAT WENT WRONG. TRACK WHAT WENT RIGHT.
This is where turning QA activity into insight— rather than just scores— makes all the difference. It brings the QA process full circle; my April article on QA focused on seeing it through the agents’ eyes while in the December issue I will look at integrating QA with training. I will also discuss training here as well.
START WITH A SIMPLE, STRUCTURED TRACKING SYSTEM
About 10 years ago, when I worked as a team lead, I was responsible for completing weekly QA evaluations for my team of 12 to 16 agents. Like many new leaders, I approached QA the same way everyone else did at the time: complete the required quota and move on to the next task.
Each agent received three evaluations per week: a short call( one to two minutes), a medium call( three to four minutes), and a long call( anything over four minutes). I took detailed notes, explained deductions to agents, and even gave some agents second chances when warranted.
On the surface, it felt thorough, at least to me at the time. But over time, I realized something was missing.
I couldn ' t easily answer basic questions, such as:
• Which agents were repeating the same mistakes?
• How often was I overlooking an issue versus coaching it or placing the agent on corrective action for repeat offenders?
• Were the specific problems isolated? Or were they showing up across the team?
Once the week ended, the insights disappeared with it. You know that feeling when the whistle blows, you slide down the dinosaur ' s tail, as in the old " Flintstones " cartoon, ready to clock out.
That ' s when I built a simple tracking system: nothing fancy, just an Excel spreadsheet.( Google Sheets works just as well, especially for smaller centers or teams without the budget for a dedicated QA platform.)
Here ' s what I tracked:
• Category of markdown( e. g., Call Authentication, Empathy, Long Hold Without Check-Back).
• Specific reason for the issues( for example:“ Agent failed to verify member ' s date of birth”). Be specific; two months later you won ' t remember what the deduction was really about.
• Call details( channel, date and time, call type [ such as an inbound member’ s call in Spanish ]).
• Agents’ names and call length.
• CSAT scores or other relevant performance indicators, when available.
• Agent scores, color coding those that passed versus those that failed.
That simple structure changed everything. Suddenly, QA stopped being a series of disconnected evaluations and became a source of insight.
WHAT THE DATA REVEALS
If you have a sufficient budget for tracking QAs— but not enterprise-level spend— you can take this a step further. Use SharePoint as a data repository and Power BI for analytics, creating dashboards that reveal trends over time.
If tools like Copilot are available, they can accelerate analysis by surfacing patterns such as: " Agent Terry has failed five QA evaluations in the past 10 months due to improper call disconnects."
That ' s not just data. That ' s a signal to investigate call histories and provide targeted coaching before the behaviors escalate or send you down a rabbit hole looking for other instances of inappropriately disconnected calls.
WHY TRACKING WHAT GOES RIGHT MATTERS
Just as important: don ' t only track what went wrong. Track what went right.
Capturing 100 % QA scores helps define what " good " actually sounds like. High-performing calls can serve as learning examples, be included in training, or even be leveraged to identify potential mentors for new hires.
Recognition, when grounded in data, becomes meaningful rather than subjective. What gets recognized gets repeated.
When I joined as a call center supervisor and trainer in my new role, I was asked to help with QA and I brought back my old tracking system. I began noticing that agents handling payment-related calls performed well on straightforward questions:
• " How much do I owe?"
• " When is my payment due?"
But when callers asked more complex questions— such as why their monthly payments fluctuated— performance dropped sharply. Agents redirected callers unnecessarily, guessed at explanations, or provided vague responses that created confusion rather than clarity.
As you can imagine, callers were calling back, which led to a drop in first call resolution( FCR), higher AHT, and, you guessed it, a reduction in CSAT; it ' s a domino effect most of us reading this don ' t like. QA data made the knowledge gap that had developed impossible to ignore.
16 CONTACT CENTER PIPELINE