October 1998

Emerging Architectures For Customer Channel Management

BY GREG STACK, TECHNOLOGY SOLUTIONS COMPANY (TSC)

Today's customer management environment is characterized by rampant business change. This change stems from increased competition, frequent marketing strategy changes, new products, new campaigns and a strong emphasis on customer retention and cross-selling. Amid these changes, organizations have seemingly impossible objectives to quickly deploy new products and customer channels while improving customer retention and reducing budgets for information systems. These objectives are further complicated by fragmented systems architectures that have repetitive functions across systems and channels.

In this environment, it is common for businesses to duplicate customer relationship components, such as business rules, workflow, customer data and contact history, across legacy systems and emerging customer-contact channels [Web, voice response units (VRUs) and direct sales]. Because each system grew up around the product or channel, it can be difficult to consolidate customer policies and procedures.

This configuration creates customer management and relationship chaos. When service levels and customer views are fragmented across systems and access channels, it creates a major barrier to successful customer relationships.

Within this fragmented architecture, customer relationship initiatives often create conflict between the business and information systems (IS) groups. Seemingly simple business requests are, in fact, complex undertakings because business rules and customer data are scattered in numerous systems and channels. Because programmers must change and test even the simplest requests, businesses must plan technology releases across systems every four months, when what they really need is real-time, simultaneous change across all systems and channels.

How can companies escape this escalating cycle of high cost, slow response and fragmented customer views? It's possible with a well-planned systems architecture that employs reusable, process-based components across all systems and customer access channels.

Architectural Foundation
All channels should be treated equally. Business logic, services and databases cannot differentiate whether a customer request originated via Web, e-mail or VRU. Customer relationship management integrates voice, Web and VRU channels with intelligent request routing and customized customer presentation based on contact history, preference and current situation.

This approach is known as a component-based "N-Tier" client/server architecture. When applied to customer relationship management, it is called a "TERM" (Technology Enabled Relationship Management) architecture, an acronym coined by the Gartner Group.

At its simplest level, the TERM architecture is divided into three tiers. Each layer is separated from the others by an abstraction layer, which eliminates the need for any layer to recognize the database type or physical location of the other layers.

Database
The TERM architecture's top layer provides data services and includes multiple generic and application-specific databases, such as customer-contact, product, billing and reference.

IS can design customer-contact and preference databases so that users can define specific customer attributes, situations and business rules in data tables, without IS having to alter the data model. This allows business users to make critical business changes in real-time.

Magic Middle Tier
The middle TERM architecture tier is where the magic really takes place. This tier, which contains business logic, acts as a traffic cop and message translator. It receives, routes and sends messages from the data and presentation tiers to generic, reusable services within the middle tier. This is the communications mechanism between the client channels, the middle tier services/business logic and the corporate application and customer relationship data.

Middle tier services are based on discrete object-oriented components. These components provide an abstraction layer between tiers and services, enabling users to move, add or replace a service without impacting other components. Decoupling tiers and services minimizes the cost and impact of technology changes.

Middle tier components provide many additional benefits for customer relationship management. Since all channels use common components, this tier provides a cross-channel mechanism for completely capturing customer contact history, preference and transactions across both systems and channels. Similarly, all channels can retrieve customer-contact and preference information from a single source.
Examples of middle tier services include:

• Common business rules services with end-user-changeable rules;
• Common workflow queue and distribution;
• Log-on and security profiles;
• Data access services and transaction coordination across multiple databases and legacy systems;
• Legacy system emulation, message translation and gateways;
• Message brokering services between databases, applications and tiers;
• Support of diverse messaging and protocols across multiple systems and channels.

The middle tier can also contribute to process-based navigation by providing client channels with system access and database services. This capability facilitates customer relationship reengineering. For example, some legacy processes require users to enter similar information into multiple screens across one or many applications. The middle tier can be programmed to automatically update multiple back-end legacy systems from a single client-initiated transaction.

Channel Presentation Tier
The lowest level of the TERM architecture is the channel-specific client tier. Each channel provides its own presentation logic to the customer. This includes a workstation graphical user interface (GUI), Web page, VRU script, etc. The client level does not contain business logic, but serves as a presentation "shell" that is configured based on data.

Configurable customer presentation is a powerful concept for two reasons. First, since the presentation is stored as data, users can make basic changes to the VRU script or workstation GUI in real-time and without IS intervention. This includes changing items such as information display, scripting and process- and procedure-based navigation. Second, users can customize the VRU, Web and GUI in real-time using business rules based on customer preferences, values and events. This provides a powerful tool for managing and enhancing customer relationships.

For example, customers with outstanding requests could be routed directly to the agent, script or Web page pertinent to their request. Special offers, scripts or cross-selling could be made available to customers based on their profiles, preferences or value to the company.

The presentation shell concept can be leveraged further to create "process-based" navigation, which crosses existing product-based systems. As businesses evolve, it's not unusual for reengineered processes to require access to multiple systems. Sales and service are examples of combined processes as businesses strive for "one and done" customer-contact scenarios.

Process-based navigation can abstract legacy systems from new or reengineered processes that can now be stored as configurable data rather than making systems coding modifications. Process-based navigation can take many forms, but usually involves a tree or tab metaphor from which users can select major processes. Once chosen, a process' subtasks are displayed in the proper order. Each subtask is stored as a system screen, allowing agents to navigate between systems with a simple mouse click.

Each client channel shell maintains common customer data that all systems can use, eliminating data reentry. Sales and agent help scripting also ties data into a task. A generic scripting service can be used for help, training, process scripting or telesales operations. If a process is redefined, the business process owner can make a simple data modification to the process and task table. The process-based presentation approach allows users to engineer processes from a customer's perspective and also makes systems easier for agents to learn and use.

TERM Benefits
Component-based TERM architectures provide benefits in several areas, including channel and customer relationship management; business control and responsiveness to change; and system and technology transition and cost reduction.

Channel And Customer Relationship Management:

• One view of the customer across all systems and access channels provides true valuation, status and preference information, helping companies create and maintain personalized customer relationships.
  
• Customer loyalty is improved because customer interaction information is consistent across all access channels.
  
• The ability to quickly incorporate or modify customer access/revenue channels results in consistent information delivery on specific customers and customer types across all access channels.
  
• Contact scripts based on customer relationship history and rules enable companies to route customers to the appropriate agent.

Business Control And Responsiveness To Change:

• End users can directly configure processes, customer presentations, business rules, customer values and call routing in real-time. This capability enables companies to quickly respond to business changes and initiatives.
  
• The business-controlled presentation approach enables users to engineer processes from a customer's perspective, making the system easier for agents to learn and use.

System And Technology Transition And Cost Reduction:

• By reducing the number of components and system complexity, businesses can reduce training costs across channels and outsourcers.
  
• Reduced complexity and components decreases the cost of application support and future changes and enhancements.
  
• Improved cross-system navigation, simplified processes and redundant data-entry elimination reduces agent handle times.
  
• New channels can be easily added by using common services on the middle tier.
  
• Abstraction layers between services and architecture tiers minimizes the impact, cost and duration of technology changes. This allows businesses to phase-in changes to systems and components without massive change and testing across system and channel infrastructure.
  
• By sharing common services on the middle tier instead of the client channel tier, software distribution complexity and cost are dramatically reduced.

Evolving With Change
TERM architectures are best viewed as an evolution because they undergo continuous changes based on customers, business requirements and technology changes. The ability of a TERM architecture to handle change is its greatest strength.

TERM-based architectures are strategic win initiatives because they provide consistent and customized customer interaction while reducing cost, improving business control and providing a migration path to accelerate technological change.

Greg Stack is a senior vice president for TSC with more than 15 years of experience specializing in telecommunications client/server architecture, call center and mobile data systems. His vast technical knowledge, coupled with management expertise, has enabled him to create world-class call centers for Fortune 500 companies, including Whirlpool and Pepsi. For his clients, Stack has led the effort to design and implement multisite call center systems, imaging, artificial intelligence, ISDN, GUI-based applications and a host of viable technologies to meet dynamic call center requirements.