The better the question
Could the most personal customer service start with AI?
COVID-19 prompted a flood of calls from customers, patients and constituents with urgent questions.
In March 2020, the COVID-19 crisis struck the United States. Organizations were flooded with questions from concerned customers, patients, and constituents. Healthcare centers and government organizations urgently needed to respond to the deluge of calls from people desperately seeking answers.
These organizations were also dealing with their own challenges. They suddenly had to run call centers with agents working from home and balancing the demands of work and family.
The sheer number of callers threatened to overwhelm the human agents – yet each individual’s question needed an answer, and quickly.
The better the answer
Virtual agents handle individual queries at mass scale
Virtual agents’ ability to intuit customer needs delivers intelligent and reassuring interactions.
Most organizations use a combination of call centers and chatbots to respond to consumer queries. Although many people ask the same questions, organizations still struggle to provide the most up-to-date information in quickly changing environments.
For example, one of EY’s healthcare clients was receiving a huge number of calls to its emergency lines from distressed patients who were worried if they were showing COVID-19 symptoms or had a question about the outbreak.
Dan Diasio, EY Americas Artificial intelligence Consulting Leader, explains: "Our virtual agent is different from a basic chatbot because of its ability to be hyper-personalized, intuit customer’s needs and actually perform important transactions."
Our virtual agent is different from a basic chatbot because of its ability to be hyper-personalized, intuit customer’s needs and actually perform important transactions.
Technology at speed: A customizable solution implemented at a pace
EY was able to help the healthcare client deploy a virtual agent solution on its website, to field the vast majority of queries with a basic assessment of symptoms and conditions and appropriate guidance for treatment. This helps provide instant and accurate information to the organization’s patient population.
EY had been developing the solution in-house when the COVID-19 outbreak struck and was able to upskill the virtual agent to meet the client’s specific needs very quickly – deployment took just five days.
Diasio explains: "People were waiting for eight hours to speak to somebody on the phone. We implemented the Virtual Agent solution rapidly to alleviate this pressing demand."
People were waiting for eight hours to speak to somebody on the phone. We implemented the Virtual Agent solution rapidly to alleviate this pressing demand.
Humans at center: Human-centered interactions are driven by personalization
In the early stages of the pandemic in the United States, a state government experienced a huge upswing in the number of calls to its unemployment information line, particularly requesting details of how to file and how to set up a request for unemployment benefits.
The state government asked EY to help them handle these calls with a virtual agent solution, and it quickly went live with the ability to answer the top 50 questions. During the course of further analysis, the virtual agent was upskilled to answer a further 30+ questions that were coming in frequently.
The virtual agent helped people understand their unemployment eligibility, how to apply for a stimulus check, and what to do if they hadn't received their check.
In this way, the virtual agent could constantly be upskilled to better meet users’ needs – and engage in complex dialog. For example, where a basic chatbot could help explain how to file a claim, the virtual agent could enable the user to file that claim directly.
Ash Kaduskar, Senior Manager for Artificial Intelligence at EY Americas, explains: "The virtual agent helped people understand their unemployment eligibility, how to apply for a stimulus check, and what to do if they hadn't received their check."
The better the world works
Better customer experiences drive long term value
Virtual agents were customized to deliver value and relief for specific customer needs.
Next-wave virtual agents like this can be deployed on websites, via text messaging, in apps, or on social media platforms – wherever customers would prefer to use them.
Not only are we reducing the call volume that was coming into the state department, we’re quickly alleviating the anxiety felt by many of the people needing unemployment assistance.
In these two deployments, virtual agents helped to massively reduce wait times – for example, they provided instant answers on unemployment queries when previously citizens had to wait for eight to ten hours for a human to call them back.
Kaduskar explains: "Not only are we reducing the call volume that was coming into the state department, we’re quickly alleviating the anxiety felt by many of the people needing unemployment assistance."
These virtual agents can be skilled to handle a wide and growing set of user interactions in multiple business functions, and are built upon human-centered design and trusted AI principles.
Virtual agents can now also include enhanced customizations, such as the ability to process transactions. By integrating backend systems, virtual agents can access relevant data and provide a recommended solution to a specific problem – just as contact center agents do.
Dean Sauer, Managing Director for Artificial Intelligence at EY Americas, explains: "These virtual agents can be skilled to handle a wide and growing set of user interactions in multiple business functions, and are built upon human-centered design and trusted AI principles."
Innovation at scale: A solution to transform customer service across industries
For the state government client, the virtual agent reduced call volume by 15-20% and handled interactions with more than 110,000 individuals between late April and late June. And for EY’s healthcare client, the virtual agent solution has handled over 225,000 customer interactions. Both deployments continue to support patients and citizens.
“We live in a world now where people want answers and they want them right now,” says Diasio. “And in particular, with the disruption that we've had through COVID-19, if people have questions about when they're going to get their unemployment check or what kind of support they can file for, the last thing they want to do is dial a phone number and wait for hours for somebody to answer.”
“It was already a stressful environment, and this was just going to amplify stress,” he continues. “So what the virtual agent does is helps those people get answers to their questions almost instantly.”
EY has since developed a suite of virtual agents for clients across multiple sectors – all hyper-personalized, constantly maintained, and paying special attention to data privacy.
As the world returns to work, these multi-dimensional virtual agents can help organizations and their customers quickly access resources and prepare for what comes next. Future iterations of these virtual agents will include new domain-specific conversation skills, enhanced personalization based on consumer profiles, new content and dialog recommendations, and increased access to information.
By operationalizing high-performing virtual agents, EY has helped numerous organizations save time and money so that they can focus their resources on larger digital transformations to help them prepare for a post-COVID-19 world.