Equipping the Enterprise for the AI-Powered Human Agent Era


From 1990 to 2010, some 4,400 children in the United States 17 years of age and younger made trips to the emergency room annually for the treatment of injuries resulting from amusement park rides. What does this have to do with contact center agents, you ask?

I think many would agree that working in the contact center of late is akin to being on a roller coaster—with unexpected ups and downs delivering the same adrenaline rush, but with less enjoyment. I also suspect many contact center agents succumbed to a form of whiplash such as these aforementioned kids—albeit not whiplash of the physical kind. In this case, the injuries sustained are being caused by situational whiplash.

Situational whiplash is how I describe the frantic need to navigate and execute in a fast-moving, twisting-and-turning landscape of new customer needs and organizational requirements—all while having to manage huge spikes in call volume and emotionally charged customer interactions while working in and among the distractions of home, and without the guidance of a peer or supervisor walking the floor.

As agents attempt to operate in the midst of these significant challenges, where things really go off the rails is when agents don’t have access to the information they need to do their jobs.

The customer service fallout from the COVID-19 pandemic has emphasized the need to accelerate the delivery of trend and issue identification along with the necessary insights right at the moment it matters most—on the agent front lines to positively change and impact outcomes during and immediately after the call.

This COVID imperative is an additive effect on top of other pre-COVID pressures affecting contact center agents, most notably rising customer expectations for responsiveness. According to research from customer service consultant Jeff Toister, more than 90% of consumers now expect a customer service email response time of one hour or less, with 15 minutes being considered “world-class.”

As well, increasing adoption of self-service technology has had the result of automating simple transactions and thus leaving the more complex interactions to the contact center. Agents now need to be armed with more information than ever before as every interaction now requires an “A-game” effort.

And let’s not forget the agents’ need to address operational requirements such as fraud mitigation and compliance.

Resolving Situational Whiplash

To understand how organizations are addressing the issue of situational whiplash, we can look to other scenarios where technology is being used to surface information, provide alerts and automate processes. These include jet cockpits where situational insights such as altitude, airspeed and target range information are constantly provided to the pilot. And the auto industry, where technology provides driver assistance with scenarios such as parallel parking, alerting when a driver drifts into a neighboring lane, and even braking to avoid a collision.

In the case of the contact center, organizations are leveraging AI-powered real-time agent assist solutions. These solutions can combine linguistic and acoustic insights analyzing what’s being said and how it’s being said with additional real-time input from the agent’s desktop, CRM system, knowledge management and more.

While analyzing all these inputs in real-time, these solutions can trigger guidance to the agent and alert a supervisor where necessary to improve the interaction outcome, reducing the cost to serve, increasing customer satisfaction and the overall value of the interaction to the customer and the organization.

By giving agents guidance and coaching “in flight,” call outcomes improve immediately, reducing the need for after-the-fact analysis, call-backs and rework.

Real-time agent assist has already provided a quantifiable return on investment in numerous deployments, including those launched during the pandemic.

  • One of the largest utility companies in North America used real-time agent guidance to prompt agents with qualifying questions before sending technicians out for home visits, ensuring that both their employees and the customers remain safe and healthy.
  • A global utility company is improving risk management in real-time for life-threatening scenarios when gas leaks are reported near critical public services such as schools or hospitals. Alerts are sent out to immediately dispatch on-the-ground resources for resolution.
  • A global international bank used real-time agent assist to monitor stressful situations and alert agents to reach out for support. In parallel, they also prompt agents for relevant upselling and cross-selling, increasing offer rates by 100% and phone sales closure rates by over 25% in just a few months.
  • A North American telecom company applies real-time agent assist to more than 10,000 agents—many now remote—to monitor potential customer complaints and escalations, significantly improving effective low-cost resolutions.
  • One of the largest North American insurance companies provides contextual knowledge to its agents based on real-time conversational intent.

Ironically, the introduction of AI technology to support assistive guidance helps agents to be less robotic and more human—not only improving the employee experience, but improving the customer experience, as well. For example, prompting agents to respond with empathy when needed, and streamlining quintessential processes, authentication and even fraud detection.

Readying Your Agent Assistive Approach

When equipping the enterprise with assistive agent technology, it’s important to avoid the risk of introducing another layer of distraction; the technology must be seamless yet effective, as user experience is key. At Verint, our approach is to present a unified, customizable and non-intrusive work assist app in the corner of the agent desktop.

Like most new technologies, it’s best to start with a small-scale pilot program—perhaps 10% of your agents—to see the impact of call outcomes and identify the best use cases. Start with a few focused use cases, such as identifying and helping to resolve complaints and escalations. Consider asking for volunteers and providing incentives to test-drive this new technological approach.

Keep measurement simple; use scorecards and reports to document improvement. Once you have deployed, you can apply lessons learned to fine-tune your efforts and expand your deployment.

Contact centers are increasingly facing an engagement capacity gap. They need to grow and improve customer engagement while addressing continuously changing interactions and increasing customer demands—all while working within budget constraints. An assistive agent initiative can help to address this exponential gap and provide a significant return on investment. Promoting the discipline of active listening and rapid response helps to enhance both agent and customer engagement.

An agent assistive approach has never been more critical as agents are scattered across the remote workplace and the need for employee engagement remains high. This type of technology is destined to become a key part of the digital workplace in 2021, as organizations look to support ongoing improvements in customer and employee engagement along with continued pressure for operational efficiencies.

D. Daniel Ziv is Vice President of Speech and Text Analytics, Global Product Strategy at Verint. Ziv has led Verint’s analytics offerings and strategy since 2002 and has helped implement AI solutions with many Global 1000 organizations to generate actionable insights that drive performance improvements and enhance customer engagement with quantifiable impact and ROI. Ziv is a published author, a frequent speaker at leading industry events, patent holder and speech analytics industry pioneer.