PDF File: 

CFO2_wp_mcminn.pdf

Independent Consultant

Five practices can help large corporations to build high-value information architectures and to better manage business performance.

In the Digital Age, views on value creation are almost universally focused on the quality and use of information by business managers. Strategic acquisitions, new channel strategies, global sourcing arrangements, and personalized service offerings all depend on managers having the ability to proactively identify unmet needs, profit opportunities, and emerging customer behaviors or performance anomalies, and act before competitors to gain advantage.

It’s a simple fact that information matters. It’s equally true that many global companies have system architectures that inhibit the efficient flow of information, thus impacting management’s ability to manage performance.

Finding effective solutions to the deceptively simple problem of providing relevant, consistent, and timely information to decision-makers has proven to be particularly vexing for large companies. Markets and customer demand have reached unprecedented levels of volatility, while regulatory requirements have become increasingly onerous. Despite these dynamics, IT organizations struggle to develop limited-scope solutions to extract, distill, and distribute the most basic types of information buried within enterprise resource planning (ERP) and legacy systems. Executives, frustrated by the lack of useful information to support business decisions, are waking up to the fact that their value agendas are at risk.

This is evidenced by the considerable number of Accenture clients who are in process or actively planning initiatives to improve their information and performance management capabilities. Through our participation in many of these programs, Accenture has gained insights into important areas companies must consider and address to establish effective information architectures that support their business performance management needs. To better understand how certain leading practices have benefited our clients, it’s important to first understand the information environments and related issues that typically exist at large, diversified companies. From this baseline we can then highlight key practices for improving a company’s management of information.

Historical Perspective

Over the past three decades, four key technological developments have emerged to shape the topography of information architectures: mainframe applications, ERP, data warehousing and business intelligence tools, and the Internet.

Enter the Mainframe

Historically, companies deployed large numbers of internally developed applications to support critical business processes such as accounting, sales order management, and commissioning systems. As raw computing power was increasing rapidly during the '70s, the primary emphasis was on developing functionally rich applications optimized for a particular business purpose. As little thought was given to integrating these applications with existing systems, the stove-piped information architectures that resulted made it difficult for decision-makers to combine and glean insight from information housed in unrelated systems.

ERPs to the Rescue

Over the last decade, ERP systems have been widely adopted to replace aging legacy systems and better integrate core business processes. While these packages have been prized for their ability to scale to handle heavy transaction-processing loads, they’ve not been universally esteemed for their performance management features.¹ Although ERP packages are delivered with prebuilt reporting and analysis capabilities, the mind-numbing complexity involved to create new reports or to code changes has accelerated the adoption of third-party products that hide back-end complexity while simplifying reporting and analysis activities.² Furthermore, as the vast majority of large companies have not fully implemented ERP systems across all their business entities and/or have implemented multivendor ERPs or multi-instance, multiconfiguration deployments of a single vendor ERP, significant challenges remain in harmonizing information into a common view of the business.³

Accenture was recently asked to participate in a post-merger integration planning initiative to assist a global technology company prepare for a large acquisition. Management’s desire to move toward an integrated planning and performance management model was greatly complicated by the company’s more than 30 multivendor, multiconfiguration ERP instances and the correspondingly complex tangle of data repositories, reporting environments, and manually intensive support processes.

Birth of the Data Warehouse

As a direct response to the difficulty of extracting, combining, and analyzing information from multiple, disparate sources, the use of data warehousing and business intelligence tools has exploded. Because these tools have been relatively inexpensive to purchase and can be quickly deployed, literally hundreds of reporting and analysis solutions have sprung-up at both corporate and local levels to address specific information needs. As the vast majority of these initiatives have been independently conceived and sponsored, significant differences tend to exist between ostensibly similar information presented from different repositories.

For instance, it’s not uncommon to find that different business rules have been applied against a common data source (for example, a central general ledger). Moreover, discrepancies between definitions, descriptions, and hierarchies can create substantive differences between summary amounts and associated details that purport to represent very specific pieces of information, such as operating expense to date for a particular functional department. Unfortunately, all these differences can create significant confusion and added work to reconcile between various sources.

As part of a project to improve management reporting for a well-known media and information client, it was discovered that inaccurate financial information was being used to manage spending for a large department. Several months into the current year, these reports reflected a favorable budget variance, when in fact the group’s spending exceeded approved budget amounts. This was due to a calculation error stored in a data mart used to generate cost center reports used by this department.⁴

Arrival of the Digital Age

With the arrival of the Internet, the Digital Age was born. Today anyone, anywhere, anytime can access vast amounts of internal and external information through any computing device equipped with a browser. Not only has there been an explosion in the availability of information and knowledge assets made available via the Internet to corporate employees, there’s also been an unprecedented willingness to share valuable corporate intelligence with customers, suppliers, investors, regulators, and other parties along the value chain to improve service, reduce costs, enhance value, extend reach, and increase transparency.

A related challenge also exists in terms of how to deal with the terabytes of externally generated customer, supplier, and competitor information companies have collected over the past few years. As if the task of managing internally generated information were not enough, business managers and IT executives must now focus their limited attention, resources, and budgets on integrating and leveraging this rich, new source of information.

Key Challenges

System architectures that support information flows for most businesses are surprisingly similar (see Figure 1). In general terms, these architectures consist of four primary layers; (1) a source systems layer (commonly known as transactional systems or systems of record), (2) an integration layer (used to transport and harmonize data to predefined standards), (3) a data warehouse layer (for storage of information in formats optimized for reporting and analysis demands), and (4) an access and delivery layer (which delivers information tailored to specific business processes and user’s requirements). Although these components are common to all organizations, three factors contribute most to differences between businesses and overall effectiveness of these architectures: ownership and governance of information and systems; degree of integration; and scope of capability deployment.

 

Figure 1: Typical System Architecture Supporting Information Flows

(See Larger Image)

Ownership and Governance

Many large corporations manage their business entities in a decentralized fashion to promote local ownership of business decisions and accountability for results — certainly an appropriate choice for many companies and a governance approach that’s favored by most investors. However, in the absence of a strong and committed corporate center to set and enforce technology and information standards (like general ledger accounts, customer numbers, product codes, etc.), this local ownership creates highly fragmented, duplicative information architectures that complicate information integration and increase ongoing support and maintenance costs.⁵

Degree of Integration

The most significant challenge all companies face is information integration.⁶ Many companies lack effective systems and information management strategies that promote the use of common codes, definitions, hierarchies, and technical infrastructure. This creates very significant obstacles in combining information from across the business.

Different companies have attempted to address this integration challenge through different means with varying degrees of success. While adopting a commonly configured, single vendor ERP is an attractive option for achieving integration, the enormous initial implementation effort, associated business distraction, and ongoing maintenance costs, not to mention the feasibility of fitting a common solution within a highly diversified business, make this option impractical for most companies.⁷

An alternative typically considered is the deployment of an enterprise-wide data warehouse. While establishing a centralized, trusted source of enterprise information is desirable, many companies have been surprised by the cost and difficulty involved in creating a reporting and analysis environment that can adequately meet the needs of a large, diverse user community. Moreover, managing and reflecting the continuous waves of business change within these environments make this type of solution difficult to sustain for many companies.

Scope of Capability

Due to the low budgets associated with most reporting and decision-support projects, approval and funding has been owned at a departmental level. Consequently, the vast majority of these initiatives, not being visible to senior IT and business leaders, have not been well-coordinated, resulting in large numbers of limited-scope point solutions that are time-consuming and costly to maintain and difficult to leverage across the business.

Five Leading Practices

Practically speaking, decision-makers create value for their organizations by improving business processes. For example, ensuring compliance with statutory reporting requirements, handling customer orders in a way that reduces the possibility of fulfillment errors, and accurately calculating tax payment amounts to minimize penalty and interest charges all enhance cash flows, reduce financial and operational risk, and contribute to increased business value.

By having a comprehensive understanding of key business processes and policies, and their contribution to enterprise value, managers are able to focus available resources on addressing the highest-value-at-stake issues. However, critical pieces of information are needed to enlighten management decisions that translate into sustainable improvements. Hence, the need for efficient access to information to facilitate rapid decision processes. By establishing a common business language, supported by effective data management practices and common technical infrastructure, information can more efficiently flow to decision-makers.

Through our experience working with some of the world’s leading organizations, Accenture has observed certain practices common to clients who have successfully tackled their information management challenges to create effective value-creation capabilities. These observations can be summarized into the following key practice elements: focus on high-value management processes; agreement on critical information needs; availability of information; effective data-governance practices; and use of common technology infrastructure (see Figure 2).

Figure 2: Layers of Information Architecture

In this way, information architecture can generally be thought of as interdependent practice elements that must properly align to enable effective corporate performance management processes.

Focus on High-Value Management Processes

As companies have grown and become more complex, awareness of end-to-end processes and their impact on business value has diminished. While this situation is more pronounced at companies or within industries that have undergone significant, recent change, all organizations have experienced the consequences of detrimental practices and/or policy exceptions that have not been sufficiently visible at key levels in the organization.

In our experience, those organizations that have been most successful at developing effective decision-support competencies recognize that certain business processes have measurably greater influence on business value. Consequently, these companies exercise considerable discipline in directing their information management investments into areas where substantial ongoing benefits can reasonably be expected.

In an effort to enhance stockholder returns, a well-known consumer products client developed an overriding objective to outgrow its competitors in global markets. To execute this bold strategy, it was felt that new capabilities would be required to better understand customers’ buying behaviors and to structure and manage global procurement programs. By developing the ability to view customer purchasing activities on a global basis, it was discovered that certain large, multinational retailers had been cherry-picking product purchases from the company’s affiliates in an effort to obtain the lowest possible price at favorable terms. This previously undetected practice had put unwarranted, downward margin pressure on key product categories.

Agreement on Critical Information Needs

One of the most contentious and hotly debated areas of any performance management project centers around what information should be monitored to measure success, inform critical decisions, or focus management’s attention. In an effort to promote buy-in at various levels throughout an organization, all too often, an inordinate amount of time and consideration are devoted to identifying and gaining agreement on information requirements. Invariably, this approach leads to unexpected project delays and overly complex, large, and costly solutions that are of little use to decision-makers and provide limited returns to the business.⁸

A preferable practice adopted by many of our clients involves following a more incremental approach in the determination of user requirements. Project executives closely adhere to the 80-20 rule, both in terms of the depth and extent of requirements gathered and the scope of capabilities targeted for deployment. This is most practically achieved by limiting the amount of time devoted to determining information needs for specific applications and restricting the number of participants involved in the requirements-definition process to more experienced business leaders. This approach, together with the use of an industry-specific data model:⁹ accelerates the identification and prioritization of information necessary to meet basic needs, increases the likelihood that delivered capabilities are sufficiently flexible and extendable, and minimizes the change management effort to build and sustain project support.

Recently, Accenture assisted a Fortune Global 50 client in executing a strategy to reverse its sagging stock price. Management felt an enterprise-wide information-management and performance-tracking solution would be required to achieve its desired goal of increased revenue and profit growth. To build credibility with investors, the company’s board established an overriding program objective of delivering performance improvements as quickly as possible. As such, initial project efforts focused on compressing project definition activities without sacrificing the quality and effectiveness of the desired foundational solution (which would need to support more than 300 business units operating in 80 countries). By adopting an incremental approach to determining information needs, combined with rigorous project scope management, the program’s aggressive timelines and cost budgets were met while expected benefits were realized ahead of schedule.

Efficient Information Access and Delivery Mechanisms

Executives increasingly find themselves overwhelmed by the daily flow of information they receive from multiple sources. A typical complaint of managers at all levels is that the information they receive is incomplete and presented in a layout that requires manual manipulation to convert to a form usable for decision-making. To turn this mountain of data into information is no trivial matter. Recent studies suggest that the majority of a manager’s time is spent devoted to assembling, massaging, and reconciling rather than analyzing and leveraging enterprise information.¹⁰ However, of even greater concern is the impact these manual processes have on the timeliness, and therefore relevance, of information used.

To automate and streamline information flows, more and more companies are integrating data warehousing solutions into their architectures. These technologies have been used to establish validated and reconciled information sources to better link performance management applications. Concurrent with these data warehousing initiatives, most organizations institute aggressive rationalization programs to promote usage of these centrally owned and controlled information sources while eliminating redundant reporting environments and associated support costs.

For instance, one of our major energy clients historically developed financial plans and budgets at one level and measured results at another. This situation created a significant amount of manual effort to pull together reports necessary to analyze and explain performance variances. By implementing a central information repository containing common structures for capturing and reporting plan and actual financial data, effort and focus was shifted toward analyzing business results and developing corrective action plans rather than assembling and reconciling reports.

While data warehousing has been an important technical enabler for improving access to business information, a more revolutionary technology is the use of portals. As Internet capabilities have matured, the importance and prominence of so-called enterprise portals has grown.¹¹ Having the ability to log on once through a browser and have seamless, secure, and personalized access to all relevant information resources (such as reporting, analytics, visualization, and workflow capabilities) is a revolutionary idea that’s changing the way and means by which enterprise resources and knowledge assets are accessed and leveraged across global organizations. Portals are also viewed as a key enabler for businesses leaders who desire to move toward a self-serve information delivery model to eliminate redundant reporting-related support and maintenance costs.

Common Data-Management Practices

Ineffective data-management practices are the most significant obstacles companies must address to extract greater value from enterprise information. Different numbering schemes, inconsistent definitions and disparate hierarchies greatly complicate management’s ability to combine similar information from across the company and ultimately impact performance management effectiveness.

Business leaders have become increasingly aware of the high cost associated with not having a common business language to describe and understand performance. Consequently, many of our clients have recently begun initiatives to develop information management strategies and introduced data and reference standards to ensure that key business information can be efficiently and consistently merged. These initiatives typically involve setting standards for a limited number of business dimensions, establishing data maintenance practices and assigning ownership and responsibility for maintaining and enforcing compliance. Furthermore, many companies have created a corporate level committee, chaired by a senior business executive, to oversee management and coordination of enterprise-wide information management activities.

Common Technical Infrastructure

To address competitive threats and capture fleeting opportunities, responsibility for implementing new technologies has generally been left to local level business managers. Predictably, redundant and overly complex systems and information architectures have resulted — composed of multivendor ERP, middleware, data warehousing, and reporting tools.

To address the obvious issues associated with overlapping technical infrastructure, even highly decentralized businesses are instituting technical product standards and initiating aggressive infrastructure-standardization programs to establish and sustain more effective information management capabilities.

Summary

Today’s fast-paced business environment is infinitely more dynamic than just a few short years ago. The need for information to develop strategies and inform business decisions has increased dramatically. In spite of significant technological advances, decision-makers are still frustrated by the lack of timely and relevant information to optimize business processes and focus management actions. Decentralized governance structures combined with the lack of effective information management strategies have resulted in highly fragmented systems architectures that complicate performance management processes. Although the barriers to streamlining information flows can be significant, our experience indicates that large, global companies can drastically improve the use and value derived from enterprise information by adopting key practices that enhance performance management competencies and ultimately lead to increased business value.

Endnotes

1 ERP vendors are aware of this perception and have invested heavily within the last few years in data warehousing and reporting capabilities that tightly integrate with their transactional systems. SAP, Oracle, PeopleSoft, and others have developed integrated modules that utilize their ERP-centric data warehouses and incorporate their pre-delivered content.

2 These large, complex ERP applications such as SAP’s R/3 package require more than 50,000 tables to house configuration and transactional details for all of their associated process modules.

3 Our experience indicates that a common ERP deployment for large, multidivision organizations is truly a rare occurrence. Studies also suggest that 60 percent of information that is needed to manage the business is outside of ERP systems. According to Hackett, world-class finance operations still rely on an average of 1.7 ERP systems for their information. It adds that while this is significantly less than average companies, which use 2.7 ERP systems, the lack of a single companywide ERP system represents a roadblock that can handicap critical business processes and decision-making. And, META Group analyst Michael Doane says, “Few operations groups are prepared to manage 80 to 190 instances of the ERP system running in different divisions and geographies, but that is what some Global 2000 companies have experienced. It becomes a management nightmare."

4 A data mart is a database that is organized to meet a specific department or function’s requirements.

5 Information integration can be thought of in two dimensions, vertical and horizontal. Vertical integration can be thought of as combining information hierarchically from bottom to top or vice versa. Horizontal integration enables a consistent view of key business dimensions (such as customer, vendor, part number, etc.) across functions and/or business entities.

6 Integration is key to sustained competitive advantage. An integrated business is one whose pieces fit each other in terms of its positioning and its ability to execute from that position. Research in several industries has shown that integration is strongly and positively correlated with superior long-term financial performance. Organizations with this alignment are more stable long-term and more successful than organizations whose positioning and execution are misaligned.

7 Only 17 percent of companies have 50 percent or more of their systems in a single ERP, and only 36 percent have 25 to 50 percent of their systems in a single ERP. Source: Hackett 2002 Book of Numbers.

8 In a survey by Cutter Consortium, 41 percent of respondents said their organization had experienced at least one data warehouse project failure, and only 15 percent claimed that their data-warehousing efforts to date have been a major success.

9 Over the years, Accenture has worked with a large number of businesses that participate in similar industries. As one might expect, similar businesses that operate within particular industries tend to have similar informational needs. As such, industry-specific information models (data entities, dimensions, measures, metrics, etc.) have tended to emerge over time.

10 A 2001 Hackett study found that finance managers spend 74 percent of their time on activities that are essential but that add no value to the business. Even more distressing is the fact that managers are spending less time on decision support — just 16 percent, down from 18 percent in 1996.

11 The Gartner Group defines enterprise portals as portals deployed for the benefit of an individual enterprise in order to deal with the touch points of the enterprise. These touch points may be internal or external, and because they deal with a variety of different communities of interest they face multiple directions. They can be horizontal in nature, covering the breadth of information, applications and processes, function, or application.

About the Author

Independent Consultant

Steve McMinn was a partner in the Accenture Finance & Performance Management service line until recently. He specialized in finance function strategy, business planning, business intelligence, and performance management solutions. Mr. McMinn is now an independent management consultant.