June 2001

BY PATRICK J. SMITH, YELLOWBRICK SOLUTIONS, INC.

Many companies will argue that customers are their most important asset, especially those whose future business is likely to appreciate. They also realize that electronic CRM (eCRM) is evolving in definition and practice and as a result, the associated technical challenges are numerous and difficult to overcome.

In the back office, customer data originates from multiple sources and is stored across disparate data stores, while front-office systems are often channel-specific and do not interact with the back-office systems, which results in inconsistent customer information that is difficult to collate. Furthermore, business intelligence tools are used to analyze information, but lack the ability to operationally drive interactions in the contact center and at other touchpoints.

Corporate efforts to be more effective in managing relationships and maximize sales opportunities on every communication channel are evident in the large investments in personalization technology for digital channels such as Web and e-mail as well as business intelligence tools designed to gain greater marketplace insight. However, for most companies, the contact center is the central point of eCRM, accounting for 80 percent of customer interactions. Ironically, the build-out of online channels has been largely conducted independent of the contact center communication strategy, leaving these digital channels out of sync.

The rate at which customers are adopting multichannel interactions is increasing, as is the expectation that when they change channels, they can pick up where they left off and close the loop. This increasingly popular perception is turning up the heat on companies to synchronize the call center dialog with online channels using strategic back-office data.

The goal of obtaining a single view of a customer is unlikely to be realized if the enterprise can only see a percentage of the picture. There is a new approach, however: the mid-office platform that can shorten the gap between the front and back offices. The way it works can be compared to other traditional avenues.

Multichannel Woes
Web and e-mail personalization applications are generally coupled to dedicated, if open, databases, each with customer-specific information. These data are separate from information available in the database that supports the contact center. This means that customer interactions on one channel do not carry over to other channels.

Deploying customer data warehouses to address the shift from products to consumers and centralized data provides more effective one-to-one cross-selling and upselling. But channel-specific personalization technology ultimately promotes collecting additional relevant customer data that are stored outside the warehouse -- a striking contrast to the advocacy of the single view of the customer. Effectively managing relationships requires three problems to be addressed:

• Customer information that is stored in a marketing data warehouse as well as other service, sales and billing databases, which does not provide a complete and timely view of the customer and prevents it from being used to aid real-time decisioning.
• Personalization rules to drive scripts and content must be crafted for each of the contact center, Web and other channels, leaving opportunities for gaps where customers can receive an offer at the contact center, which is unavailable on the Web since the site has not been updated yet.
• Touchpoints do not interrelate, which prevents interactions on one channel from having an impact on another.

Campaign Management: The Traditional Approach
During the boom of the late nineties, companies were looking for new ways to acquire customers and traditional campaign management technology was viewed as the engine of CRM to drive product-focused campaigns. This involved connecting a campaign management application to an existing customer data warehouse for offer management.

While this allowed the reuse of the existing corporate data warehouse, attempts to integrate these batch-oriented tools with the contact centers and other online (live) channels have had little success. Campaign management applications lack the ability to seamlessly synchronize inbound and outbound communications because they rely on the concept of list-pushing. This is the process of generating a list of individuals to receive a specific offer. The list is then forwarded to the touchpoint where it is updated by the channel application and then sent back to the database for reconciliation. This makes closed-loop marketing very difficult, resulting in such inconsistencies as trying to cross-sell a customer a DVD player to complement the television recently purchased but returned due to poor customer service.

The direct mail model upon which campaign management tools are based does not scale to fit the real-time multichannel model, and expensive software integration is required to tie in back-office data and front-office touchpoints. This has proven to have a high degree of risk in addition to being costly and time-consuming with limited ROI.

Multichannel Applications: The Enterprise Approach
Enterprise CRM application suites are newer, more mature and offer analytics, personalization, contact center and other touchpoint applications wrapped around a proprietary marketing database. The enterprise approach, when integrated with an existing data infrastructure, can provide customers with a more consistent experience across channels.

Dedicated analytical and campaign-management applications are used to segment customers and a campaign-manager app is used to configure rules that will be executed on each channel. The apps are coupled or "hard-wired" to a prebuilt data model that is included in the package. This single data model will run on most database platforms and provides each application the same view of customer data. The single view also allows all channels to present the same message to the customer and the campaign-management application provides real-time decisioning within a single channel to drive the dialog during a customer interaction.

Message delivery is API based and can be complex to integrate with third-party touchpoint tools, albeit superior to pushing lists. The data model is proprietary, however, and usually cannot be extended since the applications are designed to work with the data in a specific format. ETL (extract, transformation and load) tools must be used to populate the database as a scheduled batch process and the batch nature of data access prevents the database from having a real-time view of customer data across corporate divisions. The campaign application can only change a call center script, e-mail or banner ad after the analytical database is updated, which is fine for offer management, but may not take into account the affect of other information about customer activity elsewhere in the enterprise, such as in sales or service.

Eager vendors of enterprise CRM suites have wisely employed IPO proceeds to assemble a collection of nonintegrated applications. While these remain campaign focused, the combination of tools into a suite more closely addresses the shift in thinking from campaigns to message consistency on multiple channels, so if a high-risk customer is appropriate for a retention offer, the same message can be delivered on every channel.

Campaigns To Context: The Platform Approach
Market pressures are accelerating a shift in thinking from pure campaigns to strategic dialogs or customer context management. This is the ability to remember the customer in the context of the relationship as it spans product marketing, sales and service and build on the previous interaction.

The mid-office marketing platform is a unique approach that combines customer context rules engine technology deployed on top of a software platform. The platform is a framework to which analytical tools, back-office databases and front-office touchpoints, including the high-traffic contact center, are connected without costly integration. The mid-office marketing platform delivers all the proven capability of campaign management, but extends to context management by allowing the back office and the front office to operate in sync.

Platforms sit between the databases in the back office and touchpoint systems and are typically equipped with a series of software adapters that serve two purposes -- data access and message delivery.

Much information can be found in the marketing data warehouse, but additional customer information can be found in contact center databases for service, sales and billing as well. The platform connects to multiple databases via back-office adapters that combine, for example, marketing information, service history and billing data to provide a virtual real-time view of customer data -- a real-time complete view can be used for decisioning.

The customer context engine executes business rules on this real-time view and can make decisions on individual data elements that are available through any database on the enterprise network. Based on the outcome of rules, the appropriate messages for the customer are sent to the current channel with which the customer is interacting.

Since touchpoint technologies vary greatly in implementation (contact centers use scripting whereas Web sites deliver graphical content), the exact rendering of the message must be performed differently depending on the channel. Touchpoint adapters in the platform engage during a customer session to translate relevant messages from the customer context engine into a format specifically displayable at the touchpoint. This can be a contact center script or an e-mail message.

The mid-office marketing platform allows companies to engage with customers across channels by:

• Employing analytics to create business intelligence from the myriad customer information that spans customer service, sales, billing and other areas.
• Creating relationship rule sets that govern the dialog and maximize service, sales and marketing opportunities.
• Connecting rule sets to business intelligence and to multiple back-office databases with real-time, native access.
• Employing a real-time interaction engine across channels that delivers content or messages and collects responses in real-time to close the loop on the dialog at every opportunity.
• Connecting to the front office via a set of touchpoint adapters.

In this paradigm, companies would automatically recognize a customer's position in the value chain at the contact center and other channels, and would engage in an appropriate real-time dialog. Context management allows companies to go beyond the coveted cross-sell/upsell credit card offer to providing customers with a plethora of relevant information such as an opportunity to check the status of their most recent service request or an opportunity to sign up for online billing.

There are several advantages to the platform approach. The first advantage is the ability to use the customer context (the complete picture of the relationship) in the decisioning process that drives the conversation: multiple datasets, one customer, data access on the fly. The only other way to achieve the complete view is to ensure that every actionable piece of information is available in the warehouse, and that is unrealistic. The second advantage is the ability to configure a single set of business rules used drive the relationship, regardless of the channel. This can provide contact center representatives with information used to follow-up with customers on their latest Web interactions, in real-time. Perhaps the most appealing thing about the platform approach is the ability to scale, as the enterprise needs dictate, even if this means starting with one channel and expanding by connecting new touchpoints or databases to the platform. This is a low-cost-of-entry and low-risk approach to CRM.

Look At The Risks And Choose An Approach
At the end of the day, effective CRM will add value to the bottom line, and synchronizing the contact center with other channels is a key element. The investment and timeframe to implement a complete solution depends on the requirements and projected enhancements to your business model. Integrating a campaign management system is a perfect way to more accurately target market, and while integration can be expensive, the ROI is a factor of the current marketing expense. The enterprise plan delivers a lot of functionality, but this is an investment in a technology as well as a vendor. The tools are maturing, with the mid-office platform being the newest evolution and a good strategy to control cost and risk.

Patrick J. Smith is vice president of Product Management for YellowBrick Solutions, Inc. He oversees the company's product management team responsible for product vision, definition and market positioning.