How to Beat the Backlash From CRM
Spooked by runaway budgets and analysts citing failure rates higher than 50 percent, organizations are frustrated and wary of CRM. Many are realizing that it is more complex than they or the vendors who sell CRM would like to believe.
Done right, CRM helps companies wring more value from customers. But doing it right means more than deploying a few systems and throwing a few switches. Given the realities of entrenched, large-scale enterprise technology architectures and intractable data integration issues, it's no wonder CRM seems to stand in its own way, creating two new problems for every one it solves. But it doesn't have to be that way.
All of the pieces are available to make CRM work; no one's put them in one place yet. There are costly operational and analytic CRM systems that automate sales force processes and marketing campaigns, and there are chief information officers everywhere sweating their jobs because they're not sure whether their CRM deployments are going to have happy endings.
A new approach is emerging from the integration side with the potential to put CRM back on track.
The sure way to sidestep the CRM backlash, says the new wisdom, is to back sound deployment principles with extensible markup language, XML, and metadata management. An XML-based metadata framework for CRM will create an aggregation layer that lets companies pull together a meta view of who their customers are, based on what the company's business and marketing goals require they know about them.
Ultimately, this creates a dynamic CRM customer data foundation that serves all customer channels, unifying analytic and operational CRM systems. For this to work, organizations need to map standardized, reusable business definitions to the different CRM system schemas across transactional and legacy systems.
The ability to direct CRM data movement across systems is for the first time becoming available through third-party tools based on simple object access protocol and Java application programming interfaces. The practical result is companies finally can identify valuable upsell and cross-sell opportunities across products and business units and coordinate customer service across all channels.
Essential steps for CRM. What makes CRM so hard is that the information technology infrastructure is multidimensional, and all of the dimensions do not perform harmoniously. A successful CRM infrastructure integrates six elements: organization, strategy, workflow, application, data and metadata.
Metadata is the underlying definition or description of the actual data. It is "the data about the data." For example, in a database table, the metadata for a particular column might be "Billing_Address" and the actual data entry would be "16 E. Coker St." In the new wisdom, implementation of CRM means integrating all six of these elements.
o Organization. Many organizations taking a multichannel approach to sales and CRM have structured the call center, sales force and marketing groups into separate lines of business. These separate fiefdoms make it a challenge to implement a multichannel cross-sell campaign that can cross several layers and schemes of organizational matrices.
o Strategy.Companies need to integrate analytics platforms with operational systems. The ability to make sense of customer data is the jewel of CRM. But no less important is the ability to drive technology directly from analytics to operational systems.
o Workflow. As customer service employees interact with customers based on CRM initiatives, they need to adequately message, measure and capture results so CRM programs grow stronger and smarter and inspire new initiatives over time. When CRM processes require workflow across organizational boundaries, information and business events should be free to move across systems and business units.
o Application. Customer service reps need seamless integration among CRM applications to deliver the best service.
o Data. The ability to share data across applications, and thus across channel and product organizations, has traditionally vexed virtually every CRM initiative. Does the sales force automation tool have up-to-date information on the products and services customers have purchased? Does the call center have the ability to understand what the likely "next product to buy" is for each customer? Can the field service organization prioritize service calls based on customer value? The answer in every case is in good, timely data shared across the applications for sales, services and marketing.
o CRM metadata integration. CRM metadata management provides the essential link between strategy and technology. Metadata describes the "nouns" and "verbs" of strategy.
Consider these tactical programs: "identify high-value customers and offer them product X" and "when a high-value customer calls, offer them incentive Y." "High-value," "customer," "offer," "product" and "incentive" have meanings at each layer of the infrastructure.
The sum of these principles and the bottom-line requirement for CRM systems is their ability to respond to the dynamics of customer interaction no matter what form that interaction takes, be it e-mail, phone call, mail-in rebate or the click of a mouse.
Any CRM system needs to mesh with new sources, new information, new behaviors, new communication channels and new coordinated contact strategies. Most systems were not "designed for extensibility," but rather to meet static requirements.
It is difficult to document each functional requirement for these systems at implementation time, given the volume and complexity of market forces acting upon sales, service and marketing professionals trying to influence retention, acquisition and cross-sell. Therefore, the systems need to grow or whither. Architecturally, this means not only extensibility but reusability.
How to make it work. Assuming they have a solid business idea for deploying CRM, organizations will have to turn immediately to building a customer data foundation. This means creating an aggregate layer of data (i.e., metadata) flexible enough to change with each customer contact and universal enough to be available at any time at any point in the enterprise.
This data layer will provide CRM with common definitions without which enterprises cannot coordinate CRM across their various systems. Many systems fail simply because they do not have the right data in the right place in the right form.
Many large organizations have built data warehouses to integrate data and create "one version of the truth." This approach has drawbacks. First, when data must travel through the warehouse to get from application A to application B, latency is a problem.
Many companies have built marketing data marts that draw data directly from the source and bypass data warehouses because the data is too old by the time it makes it through the warehouse. Second, moving and storing lots of data is costly, in terms of infrastructure, initial set-up and ongoing maintenance. Third, centralizing the data sacrifices flexibility to the warehouse's architectural demands. Applications that can't connect, or that need to write back to a database, end up segregated. Customer-related data must be dynamic.
One way to coax the benefits out of data warehousing without these costs is to separate the metadata layer from the data warehouse. Instead of writing definitions of all of the "nouns" and "verbs" into extracted data and load processes or database-specific procedures, organizations need to store reusable versions in an externalized repository integrated with the processes that move data around.
Reconciling and reusing compatible definitions from the top down and the bottom up ensures that systems and employees focus on the objective at hand. Finance can still have a different definition of "customer" than sales as long as both definitions are mapped together and translated when CRM processes require workflow across those organizational boundaries.
Executing a customer interaction policy relies on the IT infrastructure and the ability to create a customer view across the enterprise. XML and metadata deliver the openness needed to build optimized and coordinated marketing-driven systems.
A comprehensive customer view is contingent on a unified understanding of key CRM business definitions or CRM classes. This all starts with metadata and classes. The classification "customer," may, at first glance, appear self-evident. But the data definition of "customer" may vary widely, depending on the nature of a promotion, the channel being used and/or the circumstances of the sale.
That's why organizations need standard metadata definitions that all systems and business units can use to identify upsell and cross-sell opportunities across products and business units. It's the only way marketers will ensure consistent offers and messages and enhance the likelihood of top-line revenue growth. It's the only way to sidestep the CRM backlash, which is only going to get worse before it gets better.