Customer Service Predictive Analytics

Advance Information from Hindsight to Foresight

Highlights

  • Without customer service predictive analytics, the view and information for every person in your contact or call center is entirely backward looking.
  • Contact or call center data becomes much more actionable when it advances from historical to predictive.
  • The ability to convert data into actionable insights is one of only four sustainable competitive advantages.
Johnny Grow Revenue Growth Consulting

A recurring pattern among lower performing contact or call centers is that they don't know what activities or methods deliver the biggest financial returns, so they pursue the easiest instead of the most effective.

This results in a best-case scenario of low, incremental and often temporary ROI, or a much more likely scenario of preserving the status quo.

Many analysts and call center consultants refer to this as random acts of customer support and an aimless execution path.

The Johnny Grow Contact Center Growth Formula uses Predictive Revenue Analytics (PRA) to answer the question of how to improve ROI or most effectively grow service revenue. It's helpful to transition from a customer service cost to profit center or accelerate revenue and profit growth.

The below Growth Plan Pyramid is one of the Growth Formula dashboards. It shows the cascading effects and calculated pro forma results from operational execution to final results.

Customer Service Predictive Analytics

See Customer Service predictive analytics models that connect data, insights, action and outcomes and shift information reporting from hindsight to foresight.

Click to Tweet

In addition to showing exactly how the data rolls up to drive company revenues, the contact center analytics model above delivers insights that show how multiple drivers work together. For example, we found that when improving FCR (First Contact Resolution) or CES (Customer Effort Score) by themselves the impact was far less to CSAT (customer satisfaction) than when improved together. Similarly, the data demonstrated a synergistic effect when the AES (Account Engagement Score) and the 360-degree customer view were pursued in parallel.

A growth plan must be supported by contact center metrics that roll up from the lowest level of execution to achieve the company's top priorities (i.e., ROI or revenue growth). Unless there is holistic alignment from departmental execution to company results the company's top business priorities will be delayed, degraded or just not achieved.

There are two overarching steps to build this type of contact center predictive model for your business.

  1. First, you must know what customers want (by customer segment) and the ROI of delivering those services. One method for this is to implement a Voice of the Customer (VOC) program to prioritize what customers want and align customer and company objectives. The below diagram is a visual illustration of how to convert customer feedback into prioritized and measurable insights.
Voice of the Customer Data Transformation
  1. With acquired data you can build a customer service analytics model. You can then interrogate the data in the model to show correlations, causal relationships and impact. You can show how changing one variable in the model impacts the others and apply What-If scenarios to find the biggest upside opportunities. The model allows you to experiment, compare, contrast and find the most direct and efficient route to achieving whatever is most important to the contact center and the company.

This forward looking information can help shift the role of contact centers from a cost-based service to a value-based or differentiating service that drives customer satisfaction improvements and company financial growth. But common customer service analytics challenges include not knowing your cost information or how to optimize cost while at the same time improving customer satisfaction.

Another dashboard is used to answer the question, "How do we reduce cost to serve?"

Customer Service Predictive Analytics

The above model is something we created for a midsize manufacturing company. But every business is unique so any cost takeout or cost optimization plan must first ensure it is using the most influential levers and drivers.

There are other cost drivers available, such as call management measures (i.e., average handling time (AHT) or speed of answer (SoA)) and workforce management. But for this company the data showed these variables were less influential in impacting the agent, customer and company outcomes.

Again, determining what factors and drivers are most influential starts with a VOC program that identifies what customers most want, how delivering those services impacts customer revenues and deducts the investment and cost of delivery.

The predictive pyramid models are essentially graphical blueprints that show the most direct path from effort to company outcomes. The metrics show real-time measurable progress and the cascading benefits when achieved. They also show the impact of doing nothing.

The Point is This

There are many contact and call center predictive analytics use cases. But if the goal is to contribute to the company's priorities, shift from a cost to profit center, or achieve some type of business transformation, starting with a strategic view that aligns efforts with the company's most important goals will deliver a precise plan that leads to the most significant, predictable and sustained payback.