Customer Care Must Transform
Optimizing customer profitability over the life of an account stems from a multi-industry transformation of customer care. This initiative is being driven, in part, by eye-opening data that reveal a need to begin managing not just customer relationships but also the customer experience.
A recent American Customer Satisfaction Index (ACSI) study shows that customer satisfaction levels across multiple sectors—including fixed and wireless communications, cable television, finance, insurance, retail, banking and airlines—on average trended downward from 1995 to 2006. These findings undercut the otherwise glowing customer satisfaction reports that companies in these sectors generate from their own data.
For example, a contact center report might indicate that a customer's question was answered in 13 minutes and that she was "highly satisfied" with the response time. That same report would not reveal, however, if the customer spent up to an hour before calling trying to find the answer in documentation. It would not reflect the customer's overall experience with the company.
The truth is that customer service is today a largely reactive business catering to processes rather than customers. Many tasks are carried out manually, and customers are homogenized into a standard set of practices. They are also expected to identify problems explicitly and then wait while the service provider attempts to respond.
In an environment designed to optimize customer value, service providers can reverse this reactive scenario and realize the elusive trifecta of increased customer retention and loyalty, profitable revenue growth and reduced operating expenses.
Optimizing Customer Value
In the world of customer experience management, value optimization comes into play each time a customer receives a bill, makes a payment, uses a service, makes a call or downloads content. Every such event is an opportunity to improve the interaction. The goal is to know customers and learn about them continuously so that data can be applied to boost profitability over the full customer life cycle. Though the premise is simple, execution is quite complex.
To operate on a customer-by-customer basis requires centralized policy management of all customer data or customer-affecting data, with automated, real-time and proactive policy enforcement across all channels. A policy management engine has to cover all traditional customer touch points such as agents, IVRs, kiosks and storefronts, as well as all billing and service delivery touch points.
This policy management approach can then be used to support proactive cross- or up-selling aimed at high-value customers. It can also be used to offer credits or special promotions that boost satisfaction and prevent churn. The same policy management approach can also provide appropriate and cost-effective levels of service for mid-value and low-value customers. Ideally, this set of capabilities will enable proactive customer care, useful automation and enhanced agent effectiveness.
Proactive service becomes possible when companies can anticipate customer needs and then carry out actions that save time and money yet improve customer relationship equity. Improving the quality and accuracy of automation can ensure that customers receive appropriate treatment based on their relative value. Agents become more effective in this scenario, because their actions can be guided by integrated process flows that deliver knowledge and contextual awareness to their desktops in real time, giving them real insight into the individual customers and their experiences with the company.
That said, managing customer data across dispersed legacy systems in different regions and formats remains one of the greatest obstacles to delivering quad play offerings. The question is how a service provider can access and synchronize the right data in real time to drive proactive service through all channels.
Leveraging ‘Value Drivers'
One of the first to attack the customer data management problem was the consulting firm McKinsey & Co. Working with Tier 1 operators in North America and Europe, McKinsey developed a process for managing customer lifetime value called customer lifecycle management.
McKinsey applied a three-step process to the data management issue. It began by collecting customer data in consistent, predefined categories. Next it applied special algorithms to this data to calculate the value of each customer. Lastly, it examined key process areas to determine the effectiveness of existing business systems.
McKinsey called these key areas "value drivers," such as recurring revenue, cash cost to serve, cross-sell/up-sell, credits and adjustments, renewal promotions, migrations and churn. Making manual policy improvements in each area helped operators improve financial performance dramatically (see figure).
The downside to CLM was its manually developed processes and point solutions, which could quickly become outdated and expensive. With CLM, a company might not create new policies to keep improving lifetime value and EBITDA.
As the market evolved toward convergent services, it became evident that operators would need an automated solution that could handle quad play's complexity, scalability and speed to market requirements. Customer lifetime value optimization aims to accomplish this goal by combining the principles and basic methodology of CLM with advanced automation.
Knowing What to Ask
Optimizing customer value involves two major components: internal diagnosis and strategy, and selecting and implementing a comprehensive solution.
Because every operator is different, it is essential to begin with a detailed diagnosis of current systems and problems. The CLM process provides an established path for determining the necessary customer data and the value drivers that will impact lifetime value. Initial fact-finding generally includes:
- Determining the essential data needed to optimize the customer experience and to understand total account value
- Identifying the processes and policies that need to be put in place
- Answering how the operator will create policies for each touch point, and how it will enforce these policies consistently across all channels
- Evaluating whether reactive policies for incoming events are sufficient and account for all sources of data regarding interactions with a given customer
- Creating a plan for keeping new policies up to speed with the introduction and evolution of quad play services.
Typically, operators find that selecting the right data is the toughest of these challenges, because they have to decide what makes specific types of data, out of the vast volume available, most important. CLM standardizes this task by reviewing data from 19 customer data areas, with 70 percent drawn from billing. Those 19 customer data areas include, for example, factors such as account histories, rate plans and daily revenue. More broadly, the goal—and the base algorithm for determining lifetime value—is to compare the sum of all revenues associated with a customer against the sum of all costs to provide service to that customer.
The resulting data are batch shipped to a data warehouse, where algorithms are then applied to determine the lifetime value of each customer. Based on the relative lifetime values, policies for both reactive and proactive interactions can be defined. Reactive policies are those, for example, that address how agents handle incoming calls, questions and complaints from customers. Proactive policies address situations where a problem has been detected and actions are taken to alert and accommodate the customer, but appropriately based on lifetime value.
Eight Essentials
The next step is to select and implement a solution that will put the data to work. Because customer value optimization hinges on automating CLM practices, this phase is critical. Operators know from experience that software alone will not necessarily cure customer data management issues, and that claims can outstrip actual performance. In reviewing available options, operators should at least investigate whether a solution meets eight key criteria:
1. Automation. The system must have rules-driven, repeatable processes for increasing customer value, thus moving beyond costly and time-consuming multiple point solutions.
2. Centralization. Business policies must be unified or centralized to ensure consistent treatment of customers in all interactions and all channels.
3. Personalization. Because no two customers are alike, it is essential to personalize the experience for each one, taking into account individual attributes, behaviors and the interaction context to make the customer feel as though he or she is being treated uniquely, as an individual.
4. Integration. To have a notable, positive impact on the customer experience, a holistic view of the customer relationship is necessary. This is accomplished, in part, through integration that will pull only needed data from defined sources to create a multi-faceted lifetime view of the customer. Web services and SOA are the most common techniques used today to enable this type of integration and to restrain its cost.
5. Immediacy. The system must operate in real time. It must evaluate care, billing and service events as they happen, and it should respond instantly to maximize the positive impact, or minimize a negative impact.
6. Scalability. It must scale to support millions of customers and hundreds of millions of transactions, offering fast responses and ensuring that the system never creates a bottleneck that degrades the customer's experience or profit potential.
7. Greater agent effectiveness. Automation of policy selection and enforcement is vital to eliminate trial and error, so that agents are guided through the interaction with business policies that determine—based on the holistic view of the customer—the appropriate time to make an added offer and what it would be.
8. Value added to existing systems. The system should leverage existing front- and back-office systems to detect real-time events, and to implement real-time actions generated by policy. It also must use existing databases for source data needed to calculate each customer's lifetime value.
The last point—integration with existing systems—is critical. The goal of optimizing customer value should not come at the price of decommissioning legacy systems, which often represent multimillion-dollar investments and are difficult to replace. Instead, the system should complement and coexist with all vested systems, data, and processes, including OSS, BSS, CRM and product catalogs.
Dealing With Product Catalogs
Mergers and silo product catalogs have resulted in hundreds of products with different definitions and pricing. The software engines that optimize customer value work best in a centralized product management environment because the more product versions, the more policies required for different customer groups. Operators are well aware of the problem and are moving toward centralized product management with uniform definitions. During the migration, however, multiple product catalogs should be included in the greater legacy integration scheme that focuses on identifying and capturing needed information.
Customer Care That ‘Thinks'
Through CLM-driven self-analysis and by selecting a solution that satisfies the eight criteria, the quad play provider can begin to benefit from optimizing customer value. One of the hallmarks of this process is a standardized methodology that produces consistent results. The same five-step sequence applies for any customer interaction, whether to prevent a wireless churn event or to seize the opportunity to up-sell movies:
1. An event is detected and sent to a rules engine that can trigger multiple policies.
2. The event is analyzed in context. Various policies may consider pre-calculated value, demographics, age group, past buying habits, likelihood to churn, family status and recent customer events.
3. The rules engine evaluates applicable policies and makes recommendations.
4. In the case of multiple recommended actions, optimal choices are identified.
5. Actions are executed via messages or commands to the appropriate channel.
These five steps need to happen in real time. One characteristic to look for in a solution is the ability to self-improve by collecting feedback on actions taken with customers and adjusting policies accordingly. Actions that trigger positive feedback are more likely to be repeated in the future. Those that get a regular thumbs-down can be moved lower in the scale of prioritized actions. This can create a feedback loop that improves processes and policies over time in a lifetime value context.
