Quality assurance (QA) is a key part of contact center performance management. Traditionally, managers and QA analysts use scorecards to evaluate conversations and provide agents with feedback and opportunities for improvement. The QA process aims to create a feedback loop that drives continuous improvement, tracking, evaluating, and helping agents over time. But the QA process as we know it is far from perfect. It’s reactive, manual, labor intensive and subjective.
Thanks to modern advances in AI, QA is getting a much needed uplift. Modern QA solutions monitor 100% of conversations, and in some cases, can even automate part of the scoring and evaluation process. But is AI enough to overcome the limitations of QA? Or do we need a new, more proactive approach to QA to truly transform contact center performance?
Why QA’s Impact Has Been Limited
The typical contact center performance management process contains 4 steps – planning, coaching, monitoring, and QA. Contact center managers determine a coaching plan for each agent, deliver the coaching to the agent, monitor the agent’s performance and then evaluate progress before repeating the cycle.
Historically, simply just monitoring 100% of conversations proved to be difficult. With hundreds of contact center agents taking thousands of conversations a day, manually sampling more than 2-3% of conversations is physically impossible.
Moreso, picking the right conversations to evaluate and getting an accurate read of each agent’s performance is no easy task. As a result, QA managers and agents alike often view the QA process as biased and subjective. And when you take into consideration that QA evaluations often dictate agent pay, bonuses and promotions, this becomes a big issue!
Thankfully, as contact center technology has evolved, the QA process has become more automated. With call recordings, managers could listen to past conversations. Then came modern transcription software, which allowed teams to search and review past conversations.
But which conversations should managers review? How do they know if they’re picking the right conversations? Focusing on the right problems? Providing fair and accurate feedback to agents? And would agents trust their opinions?
With so much on the line, some managers spend more than 50% of their week finding the right calls to evaluate! And limited budgets and staffing left a majority of conversations untouched by managers. Fortunately, modern AI and speech analytics solutions offered some much-needed help.
AI and Speech Analytics: Better QA (but Still Far From Perfect)
To truly scale the QA process, managers need a fast way to cut through thousands of conversations, select which conversations to score, and score the conversation. This is where modern AI-enabled QA solutions come into play. There are a few types of AI-enabled QA solutions.
Selection Assistance: Simple QA solutions focus on assisting managers with conversation selection. These tools use keyword-based rules and off-the-shelf AI models to recommend which conversations managers should review and evaluate. These tools do not score or evaluate conversations.
Evaluation and Scoring Assistance: Next are more advanced AI-enabled QA solutions. These solutions automatically evaluate and score conversations. The complexity of a solution’s evaluation logic depends on the sophistication of its AI.
Simpler solutions limit evaluations to surface-level attributes like keywords mentions, sentiment, tone, and silence. More advanced solutions use custom semantic-based AI to infer the meaning behind what’s being said. These systems can evaluate conversations for higher order behaviors that cannot be captured by simple keywords, like “Did the agent assume the sale?”
But the question still remains: can AI-enabled QA solutions overcome the limitations of QA and transform contact center performance?
Upleveling Contact Centers with Proactive QA
To truly transform contact center performance, QA needs to go from a reactive post-conversation process to a proactive during-the-conversation process.
It’s clear that AI-enabled QA is necessary for the modern contact center. Being able to use AI to quickly identify and evaluate conversations is currently the only way to scale QA. But then what? Agents are still required to remember feedback and use it during conversations. This is where QA breaks down. To truly elevate contact center performance, QA needs to become proactive.
“One time, after an agent finished a call, I gave him some feedback on how to correctly handle a customer question. The agent acknowledged the feedback and took his next call. On this call, he got the exact same question. The agent panicked, looked at me, and made the exact same mistake again!”
– Contact Center Manager
Proactive QA makes feedback actionable for agents. This is where real-time agent assist helps bridge the gap. With real-time agent assist, agents are presented with dynamic easy-to-follow coaching during every conversation. This coaching focuses on key development areas identified during the QA process and helps agents improve their skills on every conversation.
Taking a proactive approach to QA is a new concept that requires a new set of tools. This is where Cresta’s Real-Time Intelligence Platform comes in. Our platform takes a comprehensive approach to contact center performance. We combine agent assistance, coaching & QA workflows, and customer insights on a single platform, so managers can accelerate QA and put insights into action in every conversation.
To learn more, visit cresta.com/cresta-coach/ and check out our conversation with Michael Hopkins, Blue Nile’s SVP of Sales and Service.
