Today, we’ll explore a recent enhancement to the Content Analyzer: audio file analysis. This new capability allows businesses to uncover powerful insights from voice interactions by transcribing and analyzing audio files.

In this blog post, we’ll showcase how the Content Analyzer works with a real-world example—a contact center call recording. Imagine a scenario where a customer calls to cancel a service due to pricing concerns. The agent, trained to retain customers, engages with the caller, offering various discounts and incentives to meet their expectations.

The Analysis

Once the call recording was transcribed, the Content Analyzer provided detailed insights across several key aspects of communication and service quality.

1. Communication Clarity
   The analysis highlighted how effectively the agent conveyed information about pricing adjustments and service features. Phrases like “Let me clarify how this discount works for you” scored high for clarity.
2. Empathy and Customer Service
   The tool detected empathetic language used by the agent, such as “I completely understand your concern, and I want to make this right for you.” These moments were flagged as critical for customer satisfaction and retention.
3. Problem-Solving Skills
   The agent’s ability to propose viable solutions, including tiered discounts and alternative service plans, was scored and noted as a key contributor to resolving the issue.
4. Professionalism
   The analyzer evaluated the agent’s tone, word choice, and adherence to company policy, confirming a professional demeanor throughout the call.
5. Technical Knowledge
   The agent’s ability to clearly explain how pricing adjustments would reflect in future billing cycles demonstrated strong technical expertise.

Here’s an example of how the call analysis was summarized:

Once you have an high level understanding of the call analysis, you can dive into the specific aspects of the analysis. For instance, here’s what the Professionalism aspect looks like:

In this example, we’re leveraging existing capabilities to get answer to an ad-hoc question about the call.

Note this analysis contains two particular aspects to the automation of the analysis:

1. The determination of the aspects used for the analysis were generated dynamically using what was described in Dynamic Structuring of Content Analysis
2. Based on this analysis plan, the call audio was then transcribed into segments and each segment was then analyzed

Since the analysis dynamically determined was of type content review, the call was analyzed using a list of aspects. If the objective of the analysis was to make certain assertions about the call, it could look different. Same principle would apply if you wanted to do sentiment analysis or data extraction on the call transcription.

Here’s what sentiment analysis on the call recording transcript segments looks like:

Here’s an example when doing assertion against the call recording:

Here’s an example for an extraction analysis task on the same content:

Bottom line, this method to analyze content, whether it’s video, audio or text allows for multiple different analysis and extraction tasks of the same content in order to achieve different business objectives.

The Outcome

The final analysis revealed that the agent’s approach—balancing empathy with problem-solving—resulted in the customer accepting a new pricing arrangement rather than canceling the service. This outcome illustrates how effective communication strategies can transform potentially negative interactions into positive customer experiences.

Why This Matters

For contact centers, every call represents an opportunity to improve customer satisfaction and drive business outcomes. With the new audio analysis capability, the Content Analyzer equips teams with actionable insights to optimize agent performance, enhance customer retention, and deliver exceptional service.