What Is Data Analytics in CX? Leveraging Predictive Analytics for Fair Agent Utilization

Janna Pugh at Five9 explores how contact centre data analytics is evolving from historical reporting into predictive insights that help CX leaders improve operations, manage agent performance, and anticipate customer demand.

Even before AI took the tech world (and, really, the world at large) by storm, businesses ran on data.

Now that we’re several decades into The Information Age, and with AI hungrily consuming every piece of information we give it, the importance of data in business – and the data analytics that stems from it – has never been higher.

For CX leaders, understanding what data analytics is, its relationship to predictive analytics, and its use cases is crucial to attain and maintain healthy contact centres.

What is Data Analytics in CX?

Data analytics is the process of ingesting many individual pieces of information or KPIs and turning them into a story that helps to understand a particular situation, often for future decision-making.

For CX leaders, a good example is average handle time (AHT). Each individual call has a duration that you can capture — a single piece of information.

Over the course of a day, you should have hundreds (or even thousands) of these call duration data points, which you can then analyze and determine how long it takes, on average, to solve customers’ issues when they call in.

This is, of course, a single, simple example. Data analytics is most powerful (and valuable) when used at scale, taking in thousands of individual data points or contact center KPIs across multiple types of contact center metrics and identifying larger, contact center-wide trends.

Moving From Hindsight to Foresight With Predictive Analytics

Traditional data analytics tells the story of what has happened. Leaders can and should use it to inform their future decision-making, but it has most often been a “best guess” situation.

Today, however, we’re entering an age where predictive analytics is becoming the norm. The premise of predictive analytics is largely the same as traditional data analytics: take thousands of pieces of individual data points and analyze them to form trends and patterns.

With more advanced algorithms, machine learning (ML), and artificial intelligence (AI), however, these data points can forecast what will happen rather than what has already taken place.

Key Use Case: Agent Burnout Detection

To better explore a practical example of predictive analytics, let’s consider one of CX leaders’ key contact center health metrics: agent utilization (also known as the agent occupancy rate).

While many of the metrics that CX leaders track want to sit at the extremes, the highest possible CSAT, the lowest possible outage rate, agent utilization is one metric that should always sit below the maximum, around 75 – 85%.

After all, 100% agent utilization leaves no time for activities like training or note-taking, and quickly leads to burnout, employee dissatisfaction and even attrition.

But variable call volumes and call difficulty, among other factors, mean staffing levels for prime agent occupancy rates aren’t straightforward.

This is where predictive analytics can be used to better forecast trends and suggest staffing levels that correlate to your contact centre’s actual needs.

Advanced Predictive Analytics Examples For Achieving Fair Agent Utilization

Of course, staffing levels are just the beginning for the potential of using predictive data analytics in your contact centre.

When you have an intelligent CX platform collecting many contact centre health metrics, its business intelligence (BI) capabilities can help you streamline operations across your entire contact center.

Take routing, for example. With AI-powered data analytics, you could know which agents handle specific types of calls more effectively and efficiently, intelligently pairing them with customer calls — where you already know the impetus for their call because of your AI virtual agent ingest.

Rather than use more traditional distribution methods (e.g., round robin), intelligent routing ensures your agents are well-utilized and avoids overwhelming them.

Another example is with measuring cognitive load and sentiment. When an agent (or a customer) is becoming more frustrated on a call, real-time sentiment analysis can help predict — and prevent — confrontation through any number of solutions: alerting a manager, offering real-time assistance such as suggested responses, etc.

Effectively, this could help agents keep a high utilization rate without needing to take an extra breather from a difficult call.

Predictive analytics can also help with the age-old problem of “burstiness” — sudden increases in call volume that quickly stress the limits of your agents.

Whether it’s an outage or simply a certain time and day of the week that seems to inexplicably test your contact centre’s capabilities, predictive analytics can use past data (or even external data such as uptime and status trackers) to predict when an influx may occur so you can have or call in the appropriate resources.

Putting Data and Predictive Analytics in Practice

Used well, data can tell you any number of insights. But getting those insights relies on having good, clean, accurate data and then using it to tell the story so you get buy-in on future actions from across the business.

From helping agents on a personal level understand their performance data and how to use it to improve to board-level conversations about the health and effectiveness of your contact center, data analytics is what enables your ideal future state.

This blog post has been re-published by kind permission of Five9 – View the Original Article

For more information about Five9 - visit the Five9 Website

Call Centre Helper is not responsible for the content of these guest blog posts. The opinions expressed in this article are those of the author, and do not necessarily reflect those of Call Centre Helper.

Author: Five9
Reviewed by: Robyn Coppell