informa
/
Feature

Competing On Analytics

What good is a lot of customer data if you can't make sense of it? Most companies are lagging in analytical capability; here's how to use it to reach the highest stages of competitiveness.
Building the analytical skills of employees and managers is part of the cultural change that analytical competitors must undertake. However, it's unlikely that all skills for complex analytics can be broadly distributed throughout the company. Most companies will need centralized groups of analysts to perform more sophisticated analyses and set up detailed experiments. It's unlikely, for example, that a "single-echelon, incapacitated, nonstationary inventory management algorithm," employed by one analytical competitor I studied, would be developed by an amateur.

There are two alternatives for these high-powered analysts. One is to have them work in the business function that's the company's primary competitive thrust. For example, Harrah's keeps most of its "rocket scientists" in the marketing department, because customer-loyalty programs are the primary orientation of its analytics.

The other logical home is under the CIO. Such analysts make extensive use of IT and online data, and they're similar in temperament to other IT people. Companies at which analytical groups report to the office of the CIO include Procter & Gamble, the trucking company Schneider National, and Marriott. Procter & Gamble, for example, has recently consolidated its analytical organizations for operations and supply chain, marketing, and other functions. This will allow a critical mass of analytical expertise to be deployed to address P&G's most critical business issues. The group reports to the CIO and is part of an overall emphasis within the IT function on information and decision-making. In fact, P&G's IT function has been renamed "Information and Decision Solutions."

Many of the companies I interviewed stressed the importance of a close and trusting relationship between quantitative analysts and decision-makers. The need is for analytical experts who also understand the business in general, and the particular business need of a specific decision-maker.

In IT-management terms, this is a variation on the IT-alignment theme that has been a focus of CIOs since their inception. Firms that compete analytically, however, have gone beyond traditional alignment to a closer form of integration. For example, at Marriott, where there's a strong analytical group within the IT function, company executives argue it's difficult to distinguish analytical IT people from analytical marketing or finance staffs, since the groups work so closely together on projects.

A consumer-products company with an IT-based analytical group hires what it calls "Ph.D.'s with personality"—individuals who have heavy quantitative skills, yet who can speak the language of the business and market their work to internal customers and perhaps even external ones. At Quaker Chemical, each business unit has a "business adviser," or analytical specialist. Reporting to the unit head, the adviser acts as an intermediary between the suppliers of data and analysis—normally, the IT organization—and the users—that is, the executives. The adviser not only stimulates demand by showing the business unit how analysis can be useful, but also explains business needs to the suppliers to ensure that business-relevant data and analysis will be provided. In the course of my research, managers frequently commented that a relationship of trust between the analyst and the executive decision-maker is critical to the success of an analytical strategy.

No less important than analytical skills and relationships is an analytical architecture. Historically, analytical technology at many companies has been highly dispersed, with many different tools, models, and spreadsheets. However, one researcher suggests that 20% to 40% of spreadsheets contain errors. Furthermore, the proliferation of user-developed spreadsheets and databases inevitably leads to multiple versions of key indicators within an organization.

Because of these problems, many companies are attempting to consolidate and integrate their technologies for business analytics. Such an approach requires IT organizations to develop new and broader capabilities for extracting and cleaning data, loading and maintaining data warehouses, conducting data mining, and engaging in queries and reporting. In the past, the tools have come from separate vendors and have been difficult to integrate. However, the leading vendors of BI tools and applications are themselves beginning to broaden and integrate their offerings, and to market and sell them at the enterprise level.

The most important factor in being prepared for sophisticated analytics is the availability of sufficient volumes of high-quality data. Many companies have made substantial progress in accumulating transaction data over the past several years, whether the data comes from ERP systems, point-of-sale systems, or Internet transactions. The difficulty is primarily in ensuring data quality, integrating and reconciling the data across different systems, and deciding which subsets of data to make easily available in data warehouses. Many companies remain highly fragmented, and have integration issues across diverse business functions and units. Even such basic points as agreeing on the definition of who is a customer can be problematic across lines of business. As I've noted, the lowest-ranking firms on the scale of analytical competition still face significant difficulties with these basic data issues. The leading firms, however, have largely overcome them.

To take advantage of good data, an organization also needs a capable hardware and software environment. Complex analytics chew up a good deal of processing power, so the workstations and servers used for this purpose must be substantially more powerful than those used for conventional office tasks. Apex Management Group, for example, a health-care actuarial firm, is transitioning to a 64-bit computing environment to deal with the complex and data-intensive statistical analyses it performs for its clients. For Apex and others focused on real-time analytics like real-time pricing and yield management, the speed of data management and analysis is a critical factor for software and hardware.

BI applications have often been managed at the departmental level, with analytically oriented functions selecting their own tools, managing their own data warehouses, and training their own people. However, if analytics are to be a company's basis for competition, and if they are to be broadly adopted across the organization, it makes more sense to manage them at the enterprise level. This ensures that critical data and other resources are protected, and that data from multiple business functions can be correlated. The enterprise approach may include both organizational and technical capabilities for BI.

Thomas Davenport holds the president's chair in IT and management at Babson College, and is director of research for Babson Executive Education.

What stage of analytical competition are you at? Tell us at [email protected].