Many recent studies show that today’s consumers prefer self-service to live agent service – but, let’s face it, there will always be those situations where live agent assistance is required to resolve a particularly sticky or complex issue. Without a doubt, self-service via a web portal, mobile app or Interactive Voice Response (IVR) allows consumers to accomplish many common, routine tasks on their own at their convenience. But we also know from personal experience that the ‘intelligence’ of today’s self-service tools is limited, especially when compared to that of a live agent.
Historically, organizations have used very different internal systems to manage and monitor these two forms of customer service. To manage live agents, which account for about 70% of total contact center cost, almost all enterprises use some form of workforce optimization (WFO) software. These complex solutions have evolved over decades of thinking and experimenting in the real world of managing idiosyncratic human beings/agents. Contrast that with the tools and standards for managing self-service channels. To date, self-service channels have not been overly sophisticated and are often managed by IT rather than by the line of business.
Consider the often maligned IVR. Internally, organizations use simple KPIs such as call containment % and abandonment rate to measure IVR performance. Customer expectations are therefore understandably low, with consumers actually wanting to talk to a human being in order to get results. Enter Alexa, Cortana, Siri and behind them is a parade of chatbots, virtual agents and other digital assistants and employees. What happens when an enterprise expands its thinking beyond traditional IVRs, offering up Alexa as the first line of customer service? Customers recognize immediately that they are engaging with a new class of automated customer service thanks to technologies such as Natural Language Understanding (NLU). Expectations are instantly higher, certainly head and shoulders higher than with the traditional IVR or Automated Speech Recognition (ASR). With these ‘modern IVRs’, the lines between the historically separate worlds of live agent service and self-service are blurred.
Digital employees are growing at an astronomical rate. The Straits Times recently reported that chatbot growth is about 32% CAGR. That’s much higher than the projected growth rate of the agent population. That leads to the inescapable conclusion that fewer agents will be needed in the future as simple customer service tasks are increasingly handled by these ‘smarter’ digital employees. At first blush, you might think that workforce optimization systems, which are designed specifically to manage live agents, would become less important as more work is assumed by digital employees. Nothing could be further from the truth.
WFO has been honed over decades to make the human workforce as efficient and effective as possible. All of the standards for ideal agent quality, all of the measures for ideal agent productivity, all of the means for measuring the voice of the customer are in place. WFO embodies the standards of performance to which fully evolved digital employees should be held and will form the foundation for optimally integrating digital employees with a live workforce. Let’s look at some common WFO functions and consider how they will work in a world of service delivered from both live agents and digital employees:
Workforce Management – Workforce management is the prototypical example of a tool designed to manage a large human workforce, but in a world of both live agents and virtual agents, its traditional capabilities can be easily extended to optimally balance the work load between these two types of “labor”. Going forward, WFM will need to schedule employees based not only on the skills of employees but on the skills of digital employees. When volume spikes are forecast in the era of digital employees, WFM must determine whether work should be directed to digital employees with fairly limited skills (perhaps sacrificing quality) rather than staffing up with live employees and risking the cost of them being idle.
Quality Management – An integral part of all contact centers of any size, QA is a process of continuous improvement. Over a period of time, contact centers establish standards that define high quality customer interactions, constantly measure against these standards and recommend coaching to encourage agents to adhere to these standards. With digital employees ostensibly offering the intellect of a human being, contact centers will need to employ similar monitoring and critique of interaction quality, providing the digital employee’s source of intelligence the necessary quality evaluation data to change behavior. Existing live agent quality standards will naturally become the gold standard for digital employee interactions.
Performance Management – Performance Management is a powerful tool that allows contact center managers to quickly understand the performance against KPIs for individual agents, teams and contact centers as a whole. With the introduction of digital employees, there’s a new form of agent in town. It has the ability to simultaneously serve many customers, yet it has the skills and performance of a single agent, since its intelligence stems from a single rules engine or AI engine. Managers will want to closely monitor and quickly understand how this new type of agent is performing compared with human agents, so that they can use the digital employee where it performs well and recommend changes where there are KPI issues.
Speech/Text Analytics – Speech and text analytics solutions have become very important of late, since they can rapidly assess the quality of the customer journey across many channels of interaction, both for live agents and self-service. As we introduce this new type of agent, we’ll continue to want a complete picture of service being provided in all channels and for all levels of intelligence – human employee, digital employee and traditional self-service. Analysts will need to look for a whole new class of commonly occurring challenges of digital employees such as whether information is available in the digital employee knowledgebase rather than traditional speech analytics characteristics such as empathy and customer rapport. In fact, the concept of “automated quality”, increasingly available through speech/text analytics, takes on a whole new meaning with real-time feedback of interaction quality and success informing the digital employee AI engine, which can immediately self-correct based on this feedback with no human intervention at all.
As we move rapidly into the era of the digital employee, traditional WFO tools will be extended to provide a natural framework in which to optimize these automated agents. And we will continue to see refinements in traditional WFO to optimize a more skilled live agent workforce that increasingly services more complex issues, as simpler customer service become relegated to the digital employee. We are on the cusp of dramatic change in the contact center industry. Get ready for a fun and interesting ride.