The proliferation of technology makes this an exciting time to be in customer service. Fundamental assumptions, such as the fact that customers are human, are being challenged by advances in technology. This is heralding a new wave of customer service:
- Wave 1: Person to Person — If a customer had a customer service issue, then she would typically speak directly with a company representative.
- Wave 2: Person to Multichannel Interface — The introduction of channels (telephone, mobile, internet, email, etc.) introduced variety into customer service interactions. Customers could now choose how they wanted to interact with organizations. They could also choose to resolve issues on their own instead of through a representative.
- Wave 3: Things as Customers —Due to advances in virtual personal assistants (VPAs) and artificial intelligence (AI) technology (such as machine learning, data analytics, and natural language processing), the customer seeking issue resolution might not be human. By 2020, Gartner predicts that at least 20 billion things will be connected via the Internet of Things (IoT) and 5 percent of all digital commerce transactions will come from smart devices. Therefore, a portion of customer service interactions will be performed by VPAs (also referred to as voice assistants or virtual customers) rather than customers themselves.
This third wave of customer interactions is still nascent and limited to specific, low-complexity and rules-bound interactions. However, progressive companies have started investigating what is required to allow virtual customers to seek resolution on behalf of human customers. Customer service and support leaders will begin to encounter more intelligent virtual customers in their operations but not at scale until the following five challenges are overcome:
Challenge 1: Technology Capacity and Capability
Virtual customers need a lot of information about the customers they represent. They must recognize a customer's preferences and understand the breadth of factors that might influence a customer's actions. Without this knowledge, the virtual customer cannot appropriately navigate the trade-offs that are inherent in any service interaction. For example, a VPA making a hair appointment for a customer will struggle if it does not understand whether the time of the appointment or access to a specific stylist is more important to the customer. AI as it currently stands cannot independently navigate this type of interaction; human intervention is still necessary.
Another challenge is machine-to-machine communication. Many phone inquiries start with an interactive voice response (IVR) system. A VPA confronted by an automated system could be slowed down due to system compatibility issues.
Challenge 2: Data Privacy Regulatory Environment
There are a few factors that make the current regulatory and operating environment unfriendly to virtual customers. Current legislation is unclear, and most organizations have insufficient technology and lack clear internal policies to ensure privacy rights are upheld. Operationally, organizations face challenges in ensuring virtual customers are authorized to act on a customer's behalf and can effectively interact with an organization's channels.
Organizations must, therefore, consider how to authenticate virtual customers and determine that they are actually acting on behalf of the customer they claim to represent. In addition, organizations will need to decide when and where they are comfortable serving virtual customers and whether they are willing to accept personally identifiable information (PII) from them.
Challenge 3: Determination of Legal Liability
Most organizations have no clear policy for determining liability in virtual customer interactions. If something goes wrong with the transaction, (for example the VPA gave the wrong information or ordered the wrong item), who is liable? Additionally, in the case of crimes, organizations and legal institutions must also determine if user data collected by VPAs is protected under constitutional law.
Challenge 4: Brand Strategy
A customer service interaction, while usually unwanted, is still an opportunity for the customer to interact with the company. When customers use VPAs for customer service, for better or worse, they no longer have the same experience with the company. Gartner predicts that 75 of the top 100 global consumer brands will lose 20 percent of their brand equity value due to decreasing brand loyalty and the increasing influence of digital gatekeepers. This could be challenging for marketing and customer experience leaders as they need to determine how to keep control of the relationship with the consumer without having the benefit of direct contact.
Rather than emotions, algorithms derived from user preferences drive the choices that virtual customers make. Instead of focusing on fostering emotional connections with customers, service organizations will need to determine how to engage virtual customers' algorithms.
The Final Challenge: Human Acceptance
Even though more customers are employing connected devices and/or VPAs in their own homes, there is also a growing distrust of technology. According to Gartner research, an increasing number of consumers would trade convenience for the assurance of data privacy. With VPAs taking the helm, customer service organizations will have to work even harder to foster human trust in the interaction.
We are still at the beginning of the third wave of customer interactions, but virtual customers have already started acting on behalf of human customers in a limited sales and service capacity. Customer service and support leaders must, therefore, be aware of how these virtual customers behave and how they could impact future operations.
Tiffany Fountain is a vice president and team manager, and Emily Potosky is a senior specialist in Gartner's Customer Service and Support Practice.