Contact Center agents are more important than ever for running effective customer service operations. Despite the advent of sophisticated Interactive Voice Response (IVR) systems and website bots designed to automate assistance, the demand for audio interactions is actually increasing rather than decreasing.
This trend highlights the crucial role that human agents play in delivering personalized support and resolving complex issues that automated systems may struggle to address. As customers seek more meaningful connections and tailored solutions, the value of skilled contact center agents continues to rise, making them indispensable in fostering customer satisfaction and loyalty.
The Need for Stronger Quality Assessment in Contact Centers
As online interactions increase, high-value/difficult interactions rely on human interaction. What does that mean for agents? They need to be “Swiss army knife” agents with tools for every call and the skills to navigate complex issues. What does that mean for contact centers? Training and coaching are more important, not less. Assessing quality at the agent, department, and contact center level has to be precise, current, and thorough.
What is Automated Quality Assurance (AQA)?
Automated Quality Assurance (AQA) refers to the use of technology to evaluate and enhance the quality of interactions within contact centers. By leveraging advanced algorithms and machine learning, AQA systems can analyze conversations between agents and customers, scoring them against predefined metrics.
This process seeks to streamline the quality assessment operations while providing actionable insights for training and development, thus ensuring that agents continuously improve their interactions and service levels. Overall, AQA plays a crucial role in maintaining high standards of customer service in an increasingly automated environment.
Alternatively, “traditional” QA is conducted manually by a team of specialized users, typically former contact center agents/supervisors. It involves a manual inspection of an interaction via recording, assessing the agent’s performance against a set of predefined criteria. The term QA is borrowed from the manufacturing world, where products were inspected individually to verify the quality levels of the manufacturing line and spot problems quickly. The challenges with traditional QA stem from the fact that interactions are so varied, it’s difficult to create an effective sample with a limited QA team. As pressures mount in terms of customer churn and interaction complexity, traditional QA loses steam.
With Traditional QA, successes are remedial, at best: This agent isn’t cutting it (6 months after the typical probation period); this agent does task A well but not task B (one year after getting their headset), etc. So, how can AQA help to make this process more efficient and effective?
Automated Speech Recognition (ASR)
Until the advent of ASR and Artificial Intelligence (AI), the only approach was to supersize the QA team in the hopes of getting from 2% of calls to 10% (closer to 5%, more likely). Ultimately, however, this approach doesn’t work, as it can double the cost of contact center staffing.
Why is this the case? Because there are so many criteria that QA wants to check (see paragraph 1).
It’s not reasonable to tell QA, “Let’s look at these three things this month and three things next month that are different.” You won’t be able to make accurate assessments due the time it takes to baseline behaviors, develop scoring and acceptability thresholds, understand the techniques and limits of training and coaching, etc. This is an evolutionary process that requires lots and lots of data. It’s not reasonable to ask QA to limit their scope (and change their scope every quarter) in order to “right-size” the QA staffing requirement.
So what can be done? AQA uses ASR to assist the QA team by taking low-hanging fruit off their plate. What’s “low-hanging fruit?” Those tasks can be assessed using ASR.
How ASR Can Be Used in QA
Sometimes, what’s best measured by ASR is an indicator, not an actual scorable assessment. ASR (and AI in general) is best at pattern matching; did a particular event or occurrence happen? How many times within the inspection period (call, day, week) and against a target (agent or team)? AI is less effective when it comes to assessing a skill UNLESS (and until) an AI-model has been thoroughly trained.
While training AI models can be highly effective at large scales, it requires a significant investment, which can offset the benefit of using ASR to aid QA. However, the information provided by pattern matching, as an indicator, offers valuable data for the QA team in drawing their attention to (or away from) a particular skill set or other nuance.
Here are some examples of how ASR can be used in a QA context:
- Detection of required phrases for transaction compliance
- Detection of courteous or professional language
- Detection of slang or other undesirable language
How this plays out in a Quality Management program looks like this:
Step 1 | Select QA metrics that can be analyzed with transcript matching (eg., reduce or eliminate use of slang). |
Step 2 | Validate matching patterns and quantification strategy (eg, less than 5% of calls contain matches). |
Step 3 | Validate manual scoring with automated results. |
Step 4 | Determine monitoring frequency; eliminate metric from manual evaluations. |
Step 5 | Optimize training, coaching, and evaluations as needed. Continue monitoring using Conversational Analytics. |
In this manner, QA teams can reduce the time spent on manual evaluations, allowing them to focus on higher-value assessment criteria while continuing to assess performance against targets across the entire assessment spectrum. At the same time, QA teams can capture and present quantifiable evidence of conformance or performance against specific interaction types.
In the future, QA Tools from Encore and others can expect to provide reporting that combines manual and automated evaluation assessments.
DVSAnalytics: Delivering Optimal QA Tools to Contact Centers
Contact centers must adapt to the increasing demands for human interaction and the complexities of customer needs. AQA offers a pivotal solution to fulfill such needs. Leveraging technologies like ASR and AI, AQA empowers contact center teams to focus on the most critical aspects of their operations, ensuring a comprehensive approach to quality evaluation.
Ultimately, QA is here to stay, and AQA is here to help transform the way contact centers uphold their standards, making it possible to deliver consistently excellent customer experiences.
