Get Your Customer In FocusGet Your Customer In Focus

Mike Biwer, vice president of product and technology services at Power Information Network, has learned a lesson that many organizations are just now discovering: The more strategic CRM becomes, the more it's about data.

An affiliate of J.D. Power and Associates, the research firm best known for its automotive surveys, Power Information Network gathers daily point-of-sale transaction data from more than 6,000 car dealers in 40 major markets in North America. Biwer's organization has to clean, manage and deploy all that data, and Biwer has become a veritable evangelist. "Data is one of the pillars on which our company was founded," he says. "Our solutions are critically dependent on it."

For some time now, data has been bulging in data warehouses, OLAP servers, packaged applications and other technology platforms. The terabyte threshold, once the exclusive domain of a few big banks, retailers and telecommunications providers, hardly draws notice anymore. A good deal of this data continues to populate enterprise dashboards that enable companies to monitor sales, track profitability and understand customer segments. All the while, CRM has been blossoming in front-line sales force automation, customer service and marketing management.

Many now see BI as the essential component of successful CRM. But what's been a boon to many companies' bottom lines may represent a headache for CRM developers unaccustomed to the challenges of data. Let's examine what kind of data management and integration problems you'll confront in bringing BI into CRM and where you should turn for the required technology. The companion article "Intelligence in the Here and Now," considers how IT organizations are addressing key issues of performance, data integration and predictive analytics to support CRM objectives.

From Tactical to Strategic

In the early days of CRM (circa 1998), "CRM Secrets of Success" and its cousin, "The Seven Deadly Sins of CRM," were two staple presentations at nearly every related conference. They were variations on a theme emphasizing CRM as a business initiative, enriching customer-facing business processes with more relevant interactions or better service, while at the same time driving efficiencies through automation. "People, process, technology" was the catchphrase back then, and business sponsors and practitioners put the secrets into practice (see "Secrets of CRM Success, Debunked").

Today, everybody's automating customer-facing processes in sales, marketing and customer care, and e-business has entered the mainstream. The enticement was not only happier customers, it was cost savings, and thus so-called operational CRM has become standard practice. Those who have delivered operational CRM are now confronting the difficult truth that their new CRM systems not only process data but also actually generate new data. The accelerated creation of customer data presents a host of new business opportunities, not the least of which is greater customer insight. As priorities shift from tactical to strategic, the need to transcend the canned reports that accompany most CRM products is growing more urgent. CRM experts coined the term "analytic CRM," but I'd argue that it's business intelligence, replete with BI's challenges.

Here's a typical scenario. A sales executive is delighted that salespeople are maintaining their lead lists and managing their pipelines with the company's new CRM tool. As customer activities make their way into a database, the executive starts thinking: With the canned reports telling us which salespeople are above quota, wouldn't it be great to factor in their gross margin for this year? And for last year? A report on who closed the most new accounts would be nice.... And who made the most prospect calls.... Are the two related?

So begins grassroots BI. The sales executive asks the CRM team to build some reports. And the team confronts a crop of brand-new problems that have likely been mowed elsewhere in the enterprise.

Sowing the Seeds of BI

In their quest to establish the infrastructure to support sustained customer analytics, most development teams accustomed to operational CRM find thorny BI issues lurking in the tall grass:

When it comes to CRM reporting, many IT organizations still struggle with the system contention issues inherent in growing data volumes and more complex analytics. The temptation is to reduce the quantity of reports, cut back on user access or trim data volumes to avoid performance problems. Users assert that BI is less about efficient processing and more about efficient decision-making. But as the diagram below shows, the more sophisticated the reporting and analysis capabilities, the more burdened the system resources.

System resource limitations aren't the only hobgoblin of BI. Poor data quality can eclipse all other issues. "We gave ourselves six months to get our customer data in good shape," says Chad Wright, IS application manager at Avid Technology, a developer of digital media creation tools for film, video, audio, animation, games and broadcast professionals.

Wright and his team understood early in the process that integrating customer data from SAP and their legacy CRM system would trip up their usual manual and batch data cleansing processes. "After working through the initial data integration, we felt we needed a more proactive, real-time approach to acquiring and cleaning data," Wright says. "Our goal is to [address] data quality as close to the point of entry as possible, whether that's a Web site capturing customer forms or an employee inputting data." Avid now uses software from Firstlogic to apply duplicate checking and address cleansing in real time.

But examples such as Avid are still rare. Most IT organizations are rewarded for keeping costs down, so they're reluctant to add system resources or acquire enabling technologies. Meanwhile, users witness the business benefits of analytics and lobby for more data, more flexible queries and information from ever-more disparate sources — prompting arguments over ROI, cost sharing and data ownership. As CRM systems are increasingly taxed by reporting and integration demands, so is the patience of businesspeople keen on getting their hands on new customer insights.

Avoid Distortion

Notwithstanding the cost and effort involved, managers have begun considering the business cost of siloed data and inflexible reporting. "If we hadn't addressed our customer data problems early, we'd be seriously challenged to deliver the benefits we're seeing with CRM and BI today," says Chad Wright, who credits Avid's more automated data cleansing to an innovative IT management team that understood the risks of redundant customer data and its impact on frustrated business users.

Indeed, many fabled CRM failures prove the belief that partial, duplicate or contradictory customer information is often worse than no information at all. Efforts to deliver customer lifetime value, return on customer and return on marketing investment analyses have forced companies to confront the painful truth that profitability, cost-to-serve, on-time payment and other value metrics all originate from geographically and technologically disparate systems. Proactive companies recognize that the cost of forfeiting these insights can vastly exceed the BI investment.

The good news is that software vendors are stepping up to the challenge. Enterprise information integration (EII) products from vendors such as Centerboard and Composite ease navigation between disparate systems. Database vendors including Oracle and IBM have built-in extensions making it easier to link their CRM databases to external flat files. Siebel's acquisition of nQuire in 2001 transcended the transactional mindset; the resulting Siebel Analytics tool simplifies integration of non-Siebel data, including both real-time and historical sources, easing on-demand ad hoc reporting for data hungry business users.

Many of the new CRM solutions have built such capabilities into their native operational environments. "We saw it coming before it happened," says Sundip R. Doshi, CEO of Surado Solutions, a CRM vendor targeting the midmarket. "We knew better than to build the tool on a proprietary database. And we knew that dashboards and dynamic workflows coupled with back-office integration would be critical to our customers looking to use the information generated by our tool." Doshi cites Surado's ability to capture and bridge data sets from other legacy systems as one of its key values.

Form a Guiding Perspective

CRM development teams looking to embrace BI should consider two question sets:

If CRM is focused on a narrow business area, the incumbent CRM vendor probably offers extensions to meet data integration and analysis needs. Conversely, if the CRM data sources and user population are broad and disparate, buying additional CRM tool licenses is ill advised. Better to invest in a database, tool and ETL infrastructure to bolster BI capabilities for the long term.

Having such an infrastructure makes it easier to glean data from CRM systems and other data sources with the aim of deploying regular and meaningful customer intelligence across departments and user communities, and, as Power Information Network does, directly to customers. "The incoming raw data we receive may be the messiest on the planet!" Biwer says. "Our investment in the infrastructure to transform and support that data and deliver critical information to our clients has been significant. And, he adds, "it's been worth every penny."

Jill Dyche is a partner with Baseline Consulting and the author of The CRM Handbook (Addison Wesley, 2002). Write to her at [email protected].