How to Create a Good, Modern Agent Desktop

We all know that the agent talent pool is changing. Millennials are projected to form 44 percent of the U.S. workforce by 2025, and Gen Z (those between 18 and 25 years of age) are hot on their heels. These are the generations of digital natives. They expect to work with consumer-grade technologies that allow them to be more productive and that focus them on work that matters. Agents also want to have a direct impact on customers, to make a difference. They value time and experiences over long job tenures.

But we don';t make it easy for agents to do their jobs. One global communications company told me that its agents use up to a dozen disconnected applications in the course of a workday. It takes up to nine months to fully train agents, but they become so frustrated with their toolsets that they turn over after about six months.

In a study of how agents spend their time, [24]7.ai found that agents spend up to 35 percent of their day searching for information from a knowledge base or CRM, 15 percent of their day on repetitive, manual tasks, and 10 percent reaching out to subject-matter experts and leads for help. Customer service leaders complain that they can only tailor service interactions to broad customer segments. They can't optimize interactions, process flows, decisions, or next-best actions that help them attain better outcomes that foster relationships, trust, and loyalty.

The modern agent desktop is front and center in providing differentiated experiences. But, to be of value, it must do the following:

  • Empower agents to understand customers and their contexts. Agents must quickly understand their customers, their customers' value, and health. Customer insights must include not only immediate context from immediate journeys, but must also be enriched with broader attributes gathered from unstructured interaction data, such as occupation, conversational style. buying indicators. life events. relationship graphs, such as households, organizational hierarchies, and influencer relationships. This data must be used to guide agent actions and next steps.
  • Offload repetitive actions from agents to allow focus on work that matters. As self-service operations mature, agent work becomes harder and less reproducible. Chatbots help agents streamline information gathering from customers and clarify intent, surface data, and provide insights to agents. They also can offload post-call work, like classifying cases or auto-filling wrap-up forms. Digital process automation technologies hold agents' hands through predefined processes, automate reproducible steps, and can even autoscore agent performance evaluations, leaving more time for supervisors to coach. All these technologies let agents better focus on customer interactions, analyzing and solving customer issues.
  • Allow collaboration and just-in-time learning. Since agents troubleshoot harder issues, they must increasingly collaborate with customers, other agents, product experts, and even bots. Agent desktops must contain collaborative zones, and even video, to bring expert resources together to swarm around an issue and collectively work to understand its root cause, impact, and resolution. Agent knowledge must also be kept up to date. Agents don't have time to take day-long or even hour-long training classes. Learning needs to evolve to be continuous to keep up with new product introductions or emerging issues. How will you recommend the right training? Speech and text analytics can continuously monitor and score customer outcomes. Quality scores drive just-in-time coaching on specific subjects that could be only minutes long. And training is surfaced right in the agent desktop.

Great customer service rests on deep understanding of the customer, automation to free agents from repetitive tasks and to focus them on work that matters; and agent collaboration and training. Yet, no single-vendor customer service solution offers all of these capabilities today. This means that organizations must assemble these workspaces themselves.

Start with a modern customer service agent solution that provides omnichannel inquiry capture and workflow and integrated knowledge management. Layer on efficiency tools, such as RPA or digital process automation. Follow with solutions to make service more effective, such as cognitive search, agent-facing chatbots, and agent collaboration. Also invest in solutions that make agent actions more prescriptive, such as behavioral guidance, analytics-driven next-best actions, and speech and text analytics to monitor the quality of service and coach and train agents to be their best.


Kate Leggett is a vice president and principal analyst at Forrester Research.