For over fifteen years companies are gradually shifting from providing customer service by live person-to-person channels to computer-based, automated and self-service modes. In the past three years the momentum seems to have even increased to replace bilateral human interactions with human-computer interactions — human on the customer’s side, computer on the company’s side. The trend is evident in a variety of sectors, including manufacturers and dealers providing maintenance and repair of goods, inherently service providers (e.g., mobile and Internet telecom, health, insurance, tourism), and retailers. The servicescape is definitely changing, and the repercussions are still unfolding (e.g., customer adaptation, social-related, customer-company relationships).
We can identify several stages a company may go through in reducing its direct human interaction with customers — transitioning from face-to-face to phone touchpoints, then to self-service on the Internet and by mobile applications. In-between companies have applied methods such as IVR on the phone and Web-based live chats with human customer service representatives (CSR). But the latest technological advance in computer-based self-service entails a potentially more extensive substitution of intelligent virtual agents for human service agents. It means that a larger variety of issues handled thus far via phone conversations and live chats on the Internet would be resolved by chats, at different levels of sophistication, with virtual agents.
Certainly a customer would not want to rush out to a physical branch of the serving company for resolving every problem with a product or service when it can be settled by a phone call to the company’s call centre. Some enquiries and technical issues can furthermore be resolved by means of e-mail or Web-based interfaces and resources without talking to anyone — many customers prefer nowadays to make a phone call to a company only after they have exhausted their options to solve their problem by means of self-service. There are clear advantages (e.g., convenience, control, independence) in using the computer-based modes of service. However, in cases where problems cannot be resolved effectively by self-service, and reaching a live representative of the company is made harder, it may engulf a wider gap between the company and its customer, possibly inducing frustration and anger. It may be even worse for consumers who are less computer-orientated and have difficulties using those online tools (e.g., tasks that may seem obvious or easy-to-do to Millennials [Generation Y], and to a large extent to Generation X consumers, are less likely to be so for earlier generations born before 1960).
Companies have a strong incentive of cost reduction to reduce or limit forms of human personal service — various estimates suggest that the cost of interaction may fall from $10-12 when face-to-face to $5-6 by phone to less than a dollar by e-mail, online live chats or social media, and even less in human-computer interactions that do not involve a human on the company’s side. The face-to-face channel seems to fare the worst. Some companies eliminate branches for meeting with customers, reduce their accessibility or span of services provided face-to-face, and generally de-motivate customers to come and see their representatives for service.
- A few examples: (a) A mobile telecom company that receives customers at its service centre only for acquiring new phones or leaving a phone for repair at the lab but not for issues related to changing a service package or billing; (b) An airline that prefers customers to arrange and order flights by phone and better on the Internet, and de-motivates them to come and make their travel arrangements face-to-face; (c) A ticket agency (e.g., live concerts) whose office is unaccessible, relying only on phone and online contacts.
While consumers are more willing to utilise computer-based self-service tools and resources, and are doing so more frequently, this does not mean they are ready to give up access to a person from the company. That is a wrong interpretation by certain companies who make it more difficult for customers to access their representatives, by phone or face-to-face. The last thing a company should do is to let its customers feel that it is not interested in hearing or seeing them in person. Consumers should have the privilege to choose how to receive their service. Otherwise, it is a slippery slope whereby a company may distance itself too far apart from its customers.
There are a number of cases where acquisition and service are intermingled. For instance, when (a) consultation is required prior to a purchase decision; (b) the buyer is a repeated customer; and (c) a purchase transaction is made online for a product or service consumed or experienced in the real physical world.
Intelligent virtual agents (IVA) may operate in several forms with regard to their level of exchange with users. They all rely on advanced methods of artificial intelligence and abilities to interpret natural language, and they may also utilise knowledge gained through Big Data analytics (e.g., of previous customer enquiries) to improve their quality of response (e.g., Watson by IBM, Siri by Apple, Optus by IntelliResponse). In one form, that may be described as less dynamic, a user poses a question in his or her own words, to which the IVA replies with the most accurate answer it could find from existing content in the company’s knowledge base (e.g., product profiles, service procedures, bill structure). The information, provided in text, is standard for any similar question on the same topic. The agent may assume the still image of a real person with a first name. A more sophisticated form of chatbot is an animated figure that behaves more like a live agent and can actually speak. The content may not differ from that given by the agent described formerly but it gives a more realistic “lively” feeling of speaking with another human being.
The advantages of this new breed of IVA are not to be underestimated. The IVA can save customers considerable time that is often needed in reviewing multiple results for a search query, referring the user to various pages from a company’s knowledge base. The IVA is also more flexible and efficient than the anachronistic method of pre-edited FAQ. The IVA can construct a relevant answer ad-hoc on a much larger variety of issues than a typical FAQ and it is much faster and more accurate in providing the correct relevant answer than a user searching the company’s resources. Yet, a virtual agent’s answers are based only on information that is pre-existent in the digital library of the company — if a customer asks for more details on a topic that are not available in advance, the agent may revert to repeat itself (links to related or additional details may be enclosed in an answer, thus excusing the user from posing the next question). The virtual agent also seems to provide standard answers not related to a specific personal problem described by the customer (e.g., particular monetary figures in the customer’s recent bill). For that purpose, the virtual agent should promptly escalate the call to a live CSR; the question remains, how readily IVAs are configured and able to do so.
Hence, IVAs at least at this stage may be able to promise consistency of relevant answers but not real ingenuity. Other aspects that also remain debatable are, for example, the ability of IVAs to identify the correct context of questions posed in natural language and their sensitivity to the mood of customers as a chat proceeds. These capabilities call upon a combination of experience and intuition that human representatives should still have the advantage in exercising over intelligent virtual agents.
In a main feature article in Fortune magazine (August 2015), Geoff Colvin discusses the impact that 21st century’s technological changes, particularly advance of automated computer and robotic systems, have on members of society, whether as employees or as consumers (1). He is critical of a spiral of underrating humans versus computers which may lead further to degrading human touch. In response, Colvin proposes areas of activities that humans should insist on continuing to perform, no matter the abilities of computers: remain in charge (e.g., be accountable to others, making judicial decisions); work together to set collective goals; sustain an advantage in satisfying deep interpersonal needs (e.g., in doctor-patient relations).
Colvin refers to a study by a research firm where employers were asked about the skills they expect to seek in five to ten years. We may predict those would be mainly analytic, business and financial-related (e.g., note warnings of a shortage in decision scientists). Yet, according to the study cited the future skills more demanded by employers include relationship building, teaming, co-creativity, brainstorming, cultural sensitivity, and ability to manage diverse employees. These stated priorities are partly at odds with employers’ own inclination to be more reliant on computer systems for service and allowing less leeway to customer-facing employees to act on their own judgements. Social interaction and empathy are expected to be in high demand in the 21st century. However, social interaction may regress when people become increasingly occupied with their smartphones and invest more in interacting with others through social media networks; and empathy, as Colvin shows, appears to be actually in decline among college students since 1990. Colvin concludes in suggesting that people should take the challenge by computers as an opportunity and work harder on their social skills and value as humans.
Forrester Research issued recently a brief report on the changed characteristics of young contact centre agents from the Millennial generation and how to accommodate them in the workplace (2). It is a new breed of (live) agents who are well-seasoned users of computer devices and computer-based tools and applications, an experience that shapes their approach to digital technology in leisure as well as at work. They have their own “philosophy”: any information they may need is stored in some repository these days (online or offline) and their skills should be directed to finding it. There is therefore no need for them to memorise facts and procedures. At work, they seem reluctant to learn details of products and services the way workers of previous cohorts have done. They prefer to learn where to find the information, being free of memorising details of product support. That clearly poses a challenge to professionals who develop the applications that agents should use for delivering computer-assisted service. Forrester proposes going towards the new agents with tools that reenforce their information search and navigation capabilities (e.g., improved knowledge management, context-wise tools). Additionally, it is advisable to provide them hardware such as touch screens which they are so familiar with and comfortable operating (e.g,, as on their smartphones), and compatible graphic interface.
The focus on new information skills is welcome and in due time, and companies are most justified to enhance them in the young service agents. But Millennials, and others in the same mind, should realise that their approach could be self-defeating. In order to excel at work, such as delivering an exceptional customer service, one should utilise in the best way his or her rich declarative knowledge in a domain and the practical experience one accumulated. Memorising information cannot be discarded because with expertise it means the CSR is better able to quickly provide the most effective solution to a customer’s problem. Can it be done equally well by looking up the solution or clues to it in a company’s knowledge repository? This is yet to be proven.
In the realm of keeping a sensible balance between human competence and computer technology, customer-facing employees are required to demonstrate professional aptitude (e.g., domain knowledge, proficiency in using information, responsiveness) and certain personality traits that can contribute to dialogue (e.g., reaching-out, courteous, open-minded)(3). Domain knowledge resides in one’s head (brain), not by sole reference to knowledge management systems. Thereby the human agent can develop the proficiency of using information retrieved from both own-memory and the information system as the task calls for. Companies are expected to reward exceptional CSRs. Even more advanced computer technologies may offer the agents the opposite — greater dependence on and integration with the computer system. Forrester suggests that live agents should be reserved for more complex context-sensitive conversations. If human service agents cannot demonstrate exceptional capabilities, companies will be encouraged to replace them with even-greater-intelligent virtual agents in future.
Companies as well as customers and customer-facing employees may perceive benefits in greater reliance on advanced computer technologies, for preferences or interests of each party. But there is a price to pay in company-customer relationships. What indeed is a relationship without a human factor, engaged on both sides? Companies should find it very hard to talk of a bond with their customers if they have little or no human contact with them. They should not expect too much loyalty from their customers in such conditions. The three parties have much to gain from preserving and supporting live person-to-person service.
Ron Ventura, Ph.D. (Marketing)
(1) Humans Are Underrated; Geoff Colvin; Fortune (Europe Edition), 1 August 2015 (Vol. 172, No. 2) , pp. 34-41.
(2) Brief: Retool for a New Workforce Reality — New Technology for a New Breed of Agent, Forrester Research Inc., December 2014.
(3) Adopted from a 2011 post: The Human Shortages of Relationship Marketing