Putting the Customer Back in Customer Service
Three Principles for Customer-centric Support
Companies have long relied on investments in people, processes and technology to provide more effective customer service. They have implemented knowledge management practices to capture, organize, manage and analyze the content and resources used to resolve customer problems. Invariably, such support optimization initiatives are company-centric, focusing on the service variables within a company's control. They often miss the mark on incorporating the most important variable in every support interaction: the real-time need of customers. The implications speak for themselves.
A Disturbing Picture
Despite decades of investment, companies and customers are out of alignment in terms of what constitutes effective service. Consider:
- A recent study from the Customer Care Alliance found that just 16% of the respondents said they felt "completely satisfied" or "received more than they asked for."
- A study by Portland Research Group found that the average consumer must call a company 2.3 times before having his issue resolved. The same study showed that the average caller is transferred 1.5 times before speaking with the appropriate representative.
- A study by Dimension Data shows that agents spend more time wrapping up calls (an average of six minutes per call) than talking to customers (an average of four minutes per call). Average call-handle times increased by 18 seconds from two years ago.
The malaise confronting the support industry can, in part, be attributed to misguided strategies that focus inwardly on people, processes and technologies, but not outwardly on the critical fourth dimension, the customer. Datamonitor estimates that call-center investment in workforce optimization technologies will exceed $1 billion by 2006. If those investments continue to focus inwardly just on the support variables companies directly control, the performance statistics are not likely to improve.
Focus on the Customer
Few companies structure their support operations around optimizing each individual support interaction from the customer's perspective. The more prevalent approach is to structure support operations around technologies and processes that deflect more calls, reduce call-handle times and deliver pre-defined "solutions" to the greatest number of people. These are not misguided objectives; customer service is a cost center for most organizations, and delivering cost-effective service is the primary mandate for most service managers.
The Help Desk Institute (HDI) estimates support center median cost-per-incidents reported via phone, E-mail and self-service to be $20, $16 and $5, respectively. The rationale for moving to self-service channels is clear, even if the supporting strategies are incomplete. Without a more explicit customer- centric focus to how companies think through their support operations, no amount of investment is likely to reverse the troubling trends in customer service today.
Adopting the three principles of customer- centricity in your customer service and knowledge management practices will lead to an alignment of customers' expectations with corporate objectives:
1. Understand the customer's intent.
2. Personalize every interaction.
3. Be consistent across channels.
Understand the customer. To provide a positive support experience to a customer, you first have to understand that customer's specific need during each interaction, and then apply that understanding to marshal the information and resources that satisfy the customer's objective. In essence, to create a positive experience, the company must be able to capture and understand the customer's articulation of his problem, and then provide a highly personalized, one-to-one interaction in an automated, scalable, low-cost fashion.
The single most important technological innovation for automatically understanding customer needs is the development of natural language processing (NLP) technologies that analyze language for its semantic meaning. With NLP, companies have the ability to recognize the many different ways a customer can ask the same question. NLP can recognize that "which vehicles get the best mpg?" "which models get the most miles per gallon?" and "which cars get the best mileage?" are three ways of asking the same question. This enables questions to be categorized into intent categories. An intent category is a business classification for customer requests based on an understanding of the meaning behind the many ways customers can articulate their needs. This is different from traditional approaches, which seek to aggregate service requests and customer complaints into the attributes used to tag knowledge assets.
Research from InQuira, which packages industry-specific "intent libraries" for its intelligent customer interaction applications, shows that, in most industries, consumer needs can be aggregated into a manageable number of intents. In the telecommunications industry, for example, analysis of a random sample of 1,000 searches reveals that questions are rarely repeated, and that those 1,000 searches employ over 4,000 keywords. The implications are clear: optimizing customer experience and knowledge content around specific questions or keywords is unmanageable. As the graph illustrates, the "long tail" distribution of distinct questions reflects that people articulate their problems in different ways. However, when the questions and keywords are analyzed at the semantic level to determine the underlying intent of each request, the questions map to a manageable set of intent categories, effectively shifting the entire distribution curve up and to the left. Eightythree intent categories address between 72% and 74% of the 1,000 user inquiries. This powerful capability to reduce thousands of disparate, unique questions into fewer than 100 intent categories enables companies to focus resources on optimizing the user's experience by intent rather than by individual question or keyword. As a result, the company can define a more effective customer experience on an intent by intent basis, focusing on the intents most valuable to the organization.
Understanding the customer's intent or need is the key starting point for achieving true customer-centricity in support organizations. Understanding intent is the foundation from which companies can optimize all other customer service and knowledge management processes.
Personalize the interaction. A company that understands a customer's specific need at the time of interaction is empowered to create a support experience that will delight that customer. Customers want their service experience to be relevant, dependable and seamless. Understanding the language your customers use—regardless of how or where the needs are communicated (i.e. Web, E-mail, call center)—empowers the company to resolve customer problems quickly, effectively and in a manner that builds trust and loyalty. Companies can define dynamic support experiences by intent. Rather than point customers to static content or simple FAQs, companies can define rules, by intent, that dynamically pull together the most relevant information to resolve the customer's need. In addition, related information can be returned to the customer that anticipates (and fulfills) subsequent needs stemming from the original intent. Personalizing the interaction in such a fashion accommodates the iterative nature of how people interact in a live environment.
Be consistent across channels. Customers will inevitably have problems that will require escalation or movement from self-service to assisted-service resolution channels. This need not be a source of frustration for the customer if solutions can capture Web self-service session history and automatically feed that information to the agent in the contact center, especially when the agent also applies the customer's profile context (from the CRM system) to the resolution process. Empower an assisted-service experience where contact center agents have the same sophisticated tools to retrieve resolution information, and (if new content is needed to resolve the problem) trigger the knowledge creation and management tasks and workflows directly from within the source service request, and you will ensure the seamless transition of all case-related information to every individual involved in responding to the service request.
Applying the Principles
Consider a few examples to distinguish customer-centric practices from more widely accepted company-centric practices.
A Web self-service example. One company- centric practice might be to offer a self-service search function on the support Website, only to have the underlying technology match—often poorly—the customer's specific real-time need to static content derived from the company's answer to a previous customer's question. The company's approach is to fit existing answers or solutions to new questions or problems, even when a match is tenuous. For example, a customer seeking "income limits for a Roth IRA" might be directed to a general purpose overview of Roth IRA, which he will then have to read through to find the answer to his question. This is clearly a company-centric approach.
If that customer were on a customer-centric Website, he might be delighted with a response that highlights an exact answer to his query, and includes helpful related information, like a definition of a Roth IRA and a link to an application —information dynamically retrieved and presented based on an understanding of the customer's intent for that interaction. A customer- centric alternative recognizes that each customer interaction is unique, and potentially iterative. To more effectively resolve customer problems, a customer-centric company responds to each unique customer service interaction with a personalized, dynamic experience based on the company's understanding of the customer's needs at that point in time. Each self-service search interaction is a unique oneto- one dialog between customer and company. The experience must be personalized to reflect the current needs of the customer,without product line or domain boundaries restricting the information presented as part of that experience. To be customer-centric, support managers must endeavor to first understand their customers' objectives; and second, use that intelligence to deliver personalized interaction tuned to the customer's specific real-time needs.
A multi-channel example. Think about how customer cases are escalated or routed across channels. A common frustration for many of us is to dial into an automated IVR system, enter an account number at the prompt, navigate through several submenus, and not be able to get to the information we need. When we opt out of the IVR menu to speak with a live agent, the first question he or she inevitably asks is "May I please have your account number?" Many companies have distinctly defined business processes for each interaction channel, which often results in redundancy and delay for the customer. Customers view each support incident as a single experience. Customer- centric support organizations think in terms of the customer's support experience, and implement technologies and processes that allow seamless escalation from self-service to live agent channels, such as when Web self-service search session history is logged with an online case submission so live agents can begin their efforts where the customer's self-service efforts left off.
To achieve customer-centricity, provide consistency across channels by investing in systems that capture the iterative nature of the service resolution process.
A knowledge management example. Even highly defined content creation and knowledge management processes are often biased by organizational structures, thereby limiting the usefulness of such knowledge in customer service scenarios. New content may be created by knowledge engineers, perhaps in response to analyses of customer feedback, and then tagged for future retrieval by product-specific or other company-defined attributes that are inconsistently applied by employees who do not interface with the knowledge consumer. To retrieve such knowledge to resolve subsequent customer problems requires expert skills and an understanding of how the information was originally tagged and organized in one or more knowledge repositories. A non-expert knowledge consumer, like an end customer or novice call center agent, may be unable to frame a question or problem in terms that match the tags applied to the content that contains the critical information he needs to resolve the problem. A preferred, customer-centric approach to knowledge management would be to organize knowledge so that it can be consistently retrieved by non-expert users. Considering global trends reflecting increased focus on self-service options, this is particularly important. For knowledge to be properly applied in customer service contexts, it must be retrievable using a customer's vernacular. A customer-centric approach to knowledge management practices would be to trigger knowledge creation, organization, management and analysis from an understanding of how that knowledge can be applied to customer-articulated situations.
Support organizations have traditionally focused inwardly on investments in people, processes and technologies to improve productivity and lower costs, and results have been mixed. The path to improvement must include an outward focus on the customer experience. Follow the three simple principles of customercentricity, and achieve customer service alignment with your customers.
The InQuira 7 suite of applications enables companies to provide customer-centric service, online and through assisted service channels, across marketing, sales, service and support business functions. More information is available at
www.inquira.com Organizing support interactions by underlying intent captures value from beneath the "long tail"