[February 25, 2002]

How Current CRM Technologies Inhibit Good Relationships, And What Can Be Done To Help

BY BARRY BRIGGS

Customer relationship management (CRM) currently exists under a double standard. On one hand, it's supposed to help big companies overcome the impersonality of modern commerce by treating customers as the important individuals that they are. On the other, CRM isn't about service at all; it's really about sales. It's supposed to help companies retain their customers and sell more products.

This apparent double standard reflects two distinct worlds of CRM: operational and analytic. Operational CRM includes systems that directly interact with customers: Web sites, call centers, and sales force automation. Analytic CRM comprises the data mart and associated campaign management software. This dichotomy becomes problematic when one imperative, such as automating e-mail responses to reduce Web site management costs, conflicts with another, such as improving service center feedback. CRM breaks down when a short-term necessity of meeting a call center budget overrides a long-term goal of happy, repeat customers. Part of the problem is that CRM originally came from the service side, while marketing is aligned with sales. As a result, a new approach to CRM that can bring all these pieces together is essential.

Why? The promise of this type of integration is huge. A successful print-based marketing initiative, for example, immediately conjures images of a corresponding Web site or e-mail drive. The customer data from the print campaign is in the CRM system, so why not use it for an Internet-based promotion? Because the systems with the data are often separate from the systems that would run the campaigns, and the permissions associated with the data may not be shared. Most companies would have to pay a third party to run such a campaign because their e-mail, Web, and customer or prospect database systems won't work together. Moreover, the customer data likely doesn't reflect the results of the print-based effort. Marketing might like to organize an e-mail program around the prospects or the new customers, directing them to the Web site, where the company could learn more about them. But the data doesn't allow for it. Each CRM campaign winds up being an isolated event because the underlying foundation -- customer data -- is not accessible to the right systems in the right form.

Drifting Islands In A Stormy Sea
The traditional divisions between sales, service, CRM, marketing, and information technology leave businesses without a comprehensive view of the customer across all their customer channels, such as e-mail, Web, call center, direct mail, etc. In addition, too many CRM implementations simply automate current -- and flawed -- customer processes, resulting in, according to analyst estimations, a 40-70 percent failure rate in CRM implementations. Companies are asking, "How can we automate our processes?" rather than "How can we use technology to do things differently now?" So the magazine subscriber who responds to a poorly targeted e-mail marketing campaign with a repeated subscription cancellation request receives a notice demanding payment one week later. It's easy to see how customer care ends up just the opposite.

CRM developed on the operations side where companies automated processes within sales and customer service such as sales force automation and call center scripting. The Internet opened vast new possibilities for marketing, such as the ability to capture customer data and run targeted promotions and marketing efforts using that data. But marketers soon found that customer information collected over many years was locked in silos that didn't interoperate, and was so haphazardly structured that was virtually useless for marketing campaigns in its raw, "dirty" form. As a result, operational and analytic CRM are stranded on isolated islands separated by a stormy sea.

The time and expense involved in updating and cleansing this data prevents real-time decision making. Marketers yearn to improve their performance through better targeting, direct marketing activities, or outbound telemarketing. But they're finding it nearly impossible to create a timely, integrated, operationally coordinated marketing promotion using such disparate customer touch points. To some extent, most have accepted these limitations. Until now the players have largely maintained "staff" roles; advertising did not work with direct marketing, which did not work with partner marketing, which did not work with public relations and so on. Those days are gone.

Where Do We Go From Here?
Companies need to take a systematic approach -- at the data level -- to making CRM work. This starts with four steps:

1. Discovering all repositories where customer information resides;
2. Understanding what the repositories contain;
3. Making sure all systems have access to all of the information; and
4. Developing a common understanding across all systems.

In addition, all CRM systems need to recognize the same business events -- whether it's a sale on a retail Web site or a change of address in a financial company's database. This could happen through a combination of database triggers, enterprise application integration (EAI) products, or custom coding. The methodology comes down to three imperatives:

1. Gain control of all CRM information. The most important step to healing CRM is to ensure all points in the infrastructure understand the same key abstractions. All systems need to share a common definition of the customer, product, and service.
   
   
2. Understand what's happening at every point in the infrastructure. CRM architects need to ask what business events are most important to the company. Is it when a customer buys something or signs a contract? Is the company most interested in potential selling opportunities when, for example, a customer changes address or expresses interest in a given product?
   
   
3. Coordinate the interactions so the system works as a unified whole. Using triggers, EAI, or custom code, the system needs to let all channels know what the customer is doing.

Bringing It All Together
Companies need integrated hubs of customer data to be available across all customer channels, from e-mail to printed service receipt. Extensible markup language (XML) and metadata are going to bridge the gaps between sales and service systems and CRM. Universal, reusable customer and product definitions will let companies define, build, and reference a comprehensive customer snapshot that all can share. This will allow the call center, marketing, parts and accessories, and service and warranty departments, for example, to all work off the same constantly changing customer information through a simple graphical user interface (GUI).

The convergence between operational systems and marketing technology has only just begun: Web systems dynamically generate pages based on transaction history as a user clicks on a link. While waiting for cash to appear from an automatic teller machine (ATM), a customized product message appears on the touch screen based on the customer's banking preferences. A telesales representative waiting for an order confirmation asks questions that drive further personalization based on historic purchase data. A mobile phone is paged when an out-of-stock product is available again. It's increasingly difficult to separate rules that personalize content from rules that deliver content. This underlines the need to begin tying these systems together. It's clear this new lay of the land will require basic changes at technology, strategy, and architecture levels.

Understanding who the customer is, and how the computing infrastructure manages the customer, is the primary imperative of business in this decade. Today, the customer is virtual: islands of non-integrated, uncoordinated data -- all about the same customer -- reside in any number of systems in the enterprise. Leveraging this data collectively opens up new opportunities for marketing and sales.

Once the CRM system has cataloged the company's customer information, the next task is to keep it up to date -- in real time. If a bank, for example, learns a valued customer is buying a new home, the bank's rep needs to make that mortgage offer immediately. Coordinating and tuning a CRM infrastructure is a daunting IT and marketing challenge. But properly implemented and managed, CRM infrastructures can provide untold revenue opportunities, marketing efficiencies, and cost savings.

CRM integrations need to start with concrete ROI projections and carefully thought-out business goals wedded to a technology architecture that provides rapid-fire connections -- a kind of neural network -- between dynamic customer data and unified touch points. Only then will the right hand know that there is a left hand, let alone know what the other is doing. Only then will companies stop alienating customers with poorly integrated CRM technologies that do more harm than good.

Barry Briggs is chief technology officer and vice president of engineering at Wheelhouse Corp., a developer of CRM integration software that makes CRM work, where he is responsible for charting the company's technology course and managing its development. Earlier in his career, he was the chief architect and lead developer of Lotus 1-2-3.