Today, customer service consists of a complicated, bewildering and not so effective array of tech and non-tech solutions. But artificial intelligence, used the right way, can deliver a far better experience.
In recent years, the proliferation of digital technologies has created multiple customer service channels and touchpoints through which citizens can access online government services. Unfortunately, user experience is often overlooked in the design and deployment of these new digital services.
Citizens’ expectations of service are shaped not only by their interactions with government agencies, but also by their everyday digital experiences. For example, a recent Accenture survey of over 5,000 citizens from five countries found that as they encounter more user-friendly AI solutions in their daily lives, expectations for government use of these technologies increase. In this changing environment, the need for a convenient and seamless customer experience across all engagement channels has never been more pressing.
Citizens expect fast, accurate service when interacting with government online, and agencies have deployed many channels and technologies to meet these demands. However, citizens are often frustrated by what they view as complex processes and extended waiting times, and the problem is compounded when services or solutions — whether chatbots, mobile apps or websites — are not designed around their needs.
For example, when contacting government agencies, citizens often face a combination of complicated interactive voice or numeric-based response (IVR) systems, overextended customer service agents, and underperforming chatbots. These solutions are often unable to quickly resolve inquiries or, as in the case of IVRs and chatbots, to escalate requests to a human agent in a timely manner. The result? Poor overall customer service.
Problems also arise when government agencies cannot determine customer intent. Response systems at contact centers often send customers around in circles, transferring them between agents. This is both frustrating for citizens and extremely costly for organizations. Despite investing significant amounts of money in automation, agencies continue to spend heavily on recruiting and training personnel to perform basic administrative tasks that, with the right design and planning, could be automated. In fact, training front-line service staff remains one of the biggest expenses for some government agencies.
Customers can signal the same intent in many different ways. For example, a bank’s customers may have many ways of requesting their account balance. By developing a detailed library of customer intents, cataloguing how and why customers are reaching out, government can, with the right technology, respond more effectively and efficiently. Clearly understanding customer intent enables organizations to build a channel strategy for each customer contact. At the same time, high-performing self-service technologies should be able to quickly determine which calls are suitable for automation and which require management by a human agent.
The Oregon State Treasury's OregonSaves program is one example of a government agency putting citizen service needs first. The agency has developed a virtual agent and devised an intent library of questions that are best suited to be answered by a virtual agent or chatbot. The result: High-quality customer service and more than 10,000 customer messages answered by the virtual agent to date.
Government call center operations are ripe for transformation by AI in the near future. Voice recognition and natural language processing will enable intuitive and seamless digital experiences that surpass traditional interactions, improving customer satisfaction and reducing cost. Meanwhile, deep AI platform skills and data-driven applied intelligence technologies will enable game-changing organizational insights for management.
Today, only 5 percent of customer interactions with contact centers globally are powered by AI technologies, but Gartner estimates that by 2021, this will rise to nearly one in six. Fortunately, government agencies have access to a new generation of AI-enabled chatbots, capable of detecting caller intent and responding to the root causes of customer inquiries, considerably reducing call volumes.
Recently, a government call center sought to automate routine inquiries to deliver faster customer service while lowering costs. It developed an integrated approach, combining more contextual search for a public website, a digital assistant for Web chat and phone interactions, and an email handling agent. In taking these steps, the agency made personalized service available outside of standard operating hours. The result was a 30 percent improvement in service speed, with 60 percent of customer interactions handled without human intervention.
As chatbots become more intelligent and learn from citizens’ queries, they’ll also be able to shift more callers to self-service options, which are less expensive for agencies and preferable for citizens. Additionally, the information collected and analyzed by AI during these interactions will provide new insights to customer service agents and to the wider organization. AI won’t replace people. Instead, it will augment their capabilities and enable them to focus on higher-value work.
But before leaping to adopt new technologies, government agencies must think more holistically about the changes they’ll need to make to achieve the full potential of AI to enable a better customer experience. After all, a chatbot interface that delivers poor user experiences and unsatisfactory service benefits neither government nor citizens.
To capitalize on AI’s potential, government agencies need to consider more than just the technological capabilities of the solution they want to deploy. They must ensure that technology enhancements go hand-in-hand with developing people's skills and creating new organizational processes to support AI-powered citizen services. Only then will they be able to improve the overall citizen experience and reap the rewards of operational efficiency.