Next step to improve operations: A performance diagnosis by the right Conversational AI analytics

How LivePerson’s proprietary 4E metrics provide actionable insights for conversational and operational performance improvement

Michelle Klaasen

August 23, 20223 minutes

illustration of how our 4E conversational ai analytics affect business operations

From the outside, conversational analytics — like natural language processing and AI solutions — can seem shrouded in mystery. Especially when it comes to setting goals and measuring success accurately. But LivePerson’s 4E Framework makes effectively managing conversational performance possible. 

The 4E Framework focuses Conversational AI analytics around Efficiency, Effectiveness, Effort, and Emotion. Currently, 4Es consist of 22 metrics to measure the brand performance and consumer experience across all brands. And it actually helps companies improve performance and efficiency, which drive reduced labor costs to save money.

Using the 4E Framework to maximize business operations

LivePerson’s proprietary 4E metrics use a simple format to improve business operations, though it may not sound simple. With it, we build a performance diagnostic process system to maximize value. The process is generalized so it can be easily adopted across brands, identifying the improvement opportunities by pinpointing where issues come from, leading to a recommended action in line with business goals.

Let’s take a look at how this performance diagnosis works.

Diagnosis through data insights

1. Gaining insight

Key performance indicators from LivePerson's 4E operational and conversational analytics

4E metrics introduce general insights to measure performance. Shown in the graph above, we can track metrics trends in a timeframe or a specific time point for a specific brand to understand where their 4E metrics are standing compared to the industry benchmark. These insights raise further questions of “what,” “why,” and “how”:

  • What creates the low efficiency? Why are the metrics off target? How can we improve?
  • What is the reason for the poor customer experience that leads to low emotion metrics? How do we handle this with employees and/or processes? 
  • How do I compare against my industry peers? How can our team implement optimization tactics to eliminate weak spots and become best in class?

2. Performance diagnostic analysis

To answer the above questions, we conduct a diagnostic analysis. Data segmentation and the metrics correlation analysis are the core of the diagnosis. Based on the conversation operation and management, the data can be segmented in different ways: 

  • Vertical
  • Industry
  • Region
  • Agent group
  • Intent
  • Skill

The segmentation can be parallel or across from each other with multiple levels. We compare the 4E metrics between the segmented group and find the correlation between metrics in different segment levels. 

data segmentation to gain deeper insight

Through this process, we are able to answer the “what” and “why” questions. For example, we could identify the intent or line of business that is not performing as expected. Perhaps we come to understand a metric’s poor value is caused by the correlation with other metrics, helping resolve small issues, or big ones. 

3. Data-based recommendations

Finally, we draw a conclusion based on the diagnostic analysis above and provide recommendations to pinpoint the root cause. Brands are able to significantly reduce the cognitive load of analysis and improvement plans, helping them achieve business goals and potentially cut costs with improved productivity.

Performance diagnosis workflow example

Workflow of the performance analysis to uncover the actionable insights

Learn more about improving business operations with our 4E Framework, built on conversational analytics