Keith Dawson's Analyst Perspectives

The pandemic laid bare the tenuous relationship between many contact centers and their agents. The abrupt shift to remote work was just one component of the changes still under way. The past four years have also seen enterprises reckoning with questions about how to hire, what to pay their employees, what skills are needed and how an avalanche of new technology will change their operations.  

Agents are critically important in two ways. First, they represent the most direct interaction with customers any business will have and so are a point of great risk and opportunity at all times. Second, they are the most expensive ongoing component of center operations and so must factor into any discussion of technology deployment that touches their day-to-day work. In the next few years, both software providers and contact center buyers will need to reevaluate how technology impacts agents, including their productivity, engagement, skills and performance.  

Managers should be thinking about policy and technology changes now while we’re still in the early stages of an industrywide transformation. We’re experiencing a shift in how agents are sourced and trained, how they work with automated systems and how they approach interactions that are more complex and unique than in years past. Anticipating the consequences of these shifts allows managers to make key decisions about what behaviors and best practices they value and will put resources behind. Meanwhile, providers will be challenged to develop tools that help agents and managers make sense of the bewildering array of information sources and action-options with which they work.  

With these challenges in mind, here are five ways the agent environment will change along with the technology either in development or currently available to help centers cope. It is important to note that sometimes the technology is itself the change that needs to be responded to, and the agent's environment is downstream of that. 

1. Workforces are now permanently distributed with work-from-home remaining in place for the majority of agents most of the time. The managerial practice of close collaboration with agents is much harder to provide, especially for supervisors who began their careers in on-site centers. Supervisors need tools that allow them to "virtually" look over the shoulder of agents with visibility into real-time metrics and the ability to listen in and coach just as smoothly as if the agent was in the next seat. 
   
   Companies have had to adapt some practices to make that collaboration less clumsy. Many have adopted more robust internal communications systems, often from UCaaS providers, that combine contact center communications with standard business communications platforms. Companies have also benefitted from a wider geographic talent pool. Allowing for remote agents means an enterprise is not restricted to the catchment area of a particular location and instead can recruit and hire almost anywhere. They are helped by relatively underutilized tools that automate the pre-hire assessment process, including automated interviewing.
    
2. Tools embedded with AI are becoming pervasive throughout the agent work environment. There will be predictable and unpredictable consequences of this trend. We know, for example, that AI has already increased the containment rate within self-service systems which can reduce certain types of volume that reach agents. Interactions that can't be handled by automation are more complex, more judgement-dependent and more likely to require sensitivity and soft skills than interactions did in the past. Some see this leading to fewer agented interactions that will take longer and require more informational resources to resolve, necessitating increased or refocused training. Others look to AI tools for agent guidance and predict that rather than a smaller, more expensive agent pool, agents with lesser or average skills can handle more complex cases precisely because AI is backstopping them. 
   
   There's no way to know which point of view will dominate over time. One should not assume that more automation technology will allow for cost savings through headcount reductions. It might, in some cases. But businesses are going to have to rethink the cost equations that tie service level, agent expenses and customer satisfaction together because the agent/AI combination represents a severe shift from standard practices.
     
3. The next generation of agents will need to be just as skilled with textual communications as they have always needed to be with voice. That means making sure that they are trained to understand how tone and nuance are conveyed in channels like chat and email. They will need the skills to quickly absorb textual information in real time, and to pivot from one mode to another in ways that appear seamless to the customer. 
   
   Here, again, AI will complicate this process in a good way. Agent guidance and assistance tools can quickly summarize, interpret and add value to long text strings by highlighting customer sentiment along the way. AI can also produce text that the agent can use or adapt into communications. Businesses will need to fine-tune training for empathy and social skills to meet the needs of communicating in the emotionally neutral text channels. It will be more important than ever to screen for the ability to express those skills in text before hire and throughout the agents' tenure. This, too, can become an AI function.  

4. The demands of training and skills development will need to take into account characteristics that are less universal. The emphasis will shift to teaching agents how to manage increased complexity, not just what set of processes to follow. Training will have to encourage flexible thinking and incorporate more in-the-moment coaching based on real-time assessments of interactions. Supervisors will look to coaching and spot training interventions that take place close to the actual event rather than waiting until monthly evaluations are completed, often long after an interaction. This trend is deeply entwined with the idea noted above that supervisors will need help figuring out how and when to intervene when agents are out of sight line. 
   
   
5. Finally, the data and analytics environment surrounding centers will change daily work routines and loads. The use of real-time customer sentiment, for example, is something new for most agents, providing better insights into the effects of his or her behavior. More data is available about customer histories, especially around multithreaded, multichannel interactions that are often partly asynchronous, making it easier for agents to see the context and preserve continuity of the customer's experience. This is asking the agent to juggle more balls in the air than they have been responsible for in the past, though, and again, that leads us down the path to needing new criteria for hiring and performance measurement. 

Some of the new criteria should be related to the longer-term outcomes of agent activity. Agents often hit their performance KPIs of speed and volume without having a clear sense of whether interactions produce positive outcomes for the company as a whole. More and better data suggests that managers may be able to create new KPIs that reflect agent impact on customer value, loyalty and advocacy. This, in turn, may produce downstream effects in reduced agent attrition, or even highlight unrealized skills among unrecognized agents. Once agents are recognized for skills related to generating revenue or increasing sales, new ways to motivate performance will surely emerge, along with a desire to spread knowledge about those skills among the wider agent pool. By 2027, we assert that three-quarters of organizations will automatically evaluate data to capture all sources of value an employee contributes to customer value and loyalty. 

Software providers have been rapidly adding technology to their offerings that will directly impact agent management. For the most part, the technology has been developed in advance of the actual agent-related use cases. Examples include generative AI that leads to next-best-action guidance and conversational chatbots that reduce volume but ratchet up interaction complexity. Providers are starting to better map their innovations to those use cases, increasingly highlighting the specific cost savings derived from new features and encouraging their buyers to explore a broader palette of KPIs for agent performance.  

For their part, buyers need to think about how they will maintain cost controls and optimized operations during an extended transformational period. It is particularly important for contact center leadership to consider how the technology they add in the short term will impact their workforces in the long term. This is especially true for ubiquitous platform technology like AI and knowledge management. Providers and buyers need to jointly connect these technology shifts to their very real downstream effects and develop business cases for technology enhancements that directly address ongoing business concerns about managing labor costs.  

Regards,

Keith Dawson