Improved financial reporting and analysis. Efficient financial processes. Enhanced measurement and analysis of operations. A consistent compliance infrastructure. These are the measures of successful finance and performance management.

How can you achieve such an ambitious transformation? A data warehouse is a core component, but it's only a starting point. You also need to simplify your data sourcing, create consistent data definitions and take control of your reporting environment.

By integrating your financial and non-financial information, establishing consistent data definitions and consolidating your reporting and analytic tools, you can transform your finance organization and its contribution to the business. Armed with a better understanding of your enterprise data, where it came from, what it means and how to best take advantage of it, finance can help greatly improve the bottom line. The key to making it all work? Simplification. 

Untangling the spaghetti
Simplification needs to start with data sourcing. Enterprises dealing with multiple products or services create, maintain and utilize an abundance of financial and non-financial data—oftentimes within numerous operational systems. From those sources, information must be shared and analyzed, reports generated and audits conducted. Finance organizations within all companies must ensure their data sources are controlled and in compliance with relevant auditing and regulatory requirements such as Sarbanes Oxley, but the more complex their data infrastructure is, the harder it is to test and prove.

If not integrated within an enterprise data warehouse (EDW), managing the flow of data between multiple operational systems and multiple data marts isn't just complicated and costly to manage, but it can also result in a risky, spaghetti-like mess of information.

Controlling costs and chaos
Data marts are departmental data warehouses within a company. A single department may have multiple data marts, which they typically manage independently. For instance, the marketing department may have its own data mart to do campaign tracking; the new accounts department may have several data marts specifically for different types of products; and finance might have its own profitability system and another data mart to support its reporting requirements.

While this may make sense to the individual departments, these siloed marts can hamper analytical functionality within the enterprise. Data latency, poor data quality and lack of an integrated view of the business can result from individually administered data marts. And separately maintaining multiple data marts can be "astronomically costly," as John Van Decker, senior vice president and principal research fellow with the analysis firm Robert Frances Group, points out. Each mart requires its own hardware and software, as well as the additional expense of labor to maintain them.

While some independent data marts may be necessary, it's good to keep the number under control and make sure the data warehouse is used to update them on a regular basis. If you can control the proliferation of those multiple marts and ensure all of their data is regularly fed from a centralized EDW, you can significantly reduce maintenance costs by having a single central repository and less downstream marts to administer and maintain.

Likewise, with an integrated system, information can be universally defined and stored onto a common data repository and centrally maintained. The data output it offers can be easily delivered and comprehensively analyzed by multiple business units within the enterprise.

Data latency is another concern for businesses with nonintegrated systems. When stored in multiple marts, data that is accessed for analysis must go through several processes or "hops" between its source and the end user report. This, in effect, increases the time from request to retrieval. Streamlining your data flows and having reporting and analytic tools directly access the EDW reduce data latency and minimizes data movement and the creation of another reporting data mart. With an integrated system, your enterprise has the ability to access near-real-time data for critical enterprise-wide decision making.

Apples to apples
Even if you streamline your data-sourcing environment, you should be aware that other issues may arise, such as inconsistent data definitions. In this case, even simple transactions may be defined differently based on how users decide to classify them. For example, it is imperative that the finance organization is aligned and consistent in what is classified and included in areas as fundamental as sales, new products, service fees and revenue. While it may seem such basic financial terms are self-evident, if uniform data definitions aren't developed, you run the risk of producing incorrect or inconsistent management and financial reporting, with different staff divining conflicting versions of what they consider to be "the truth."

Additionally, departmental data marts oftentimes carry duplicate data, but with varying identifiers. Data that is, in effect, the same can be misconstrued or overlooked because it isn't labeled consistently across the enterprise. So now a department who wants a report indicating the number of customers who purchased a particular product may get false or incomplete information.

Creating a single data dictionary across the EDW—with each financial and non-financial term assigned a specific meaning—eliminates this danger. Reports will be consistent no matter the source used to populate them or the creator.

A consensus on definitions from departments across the enterprise is essential, both in creating and utilizing the dictionary. This will give you a consistent view of the data no matter which department is generating it. This is also critical to regulators, who want to make sure all data is defined in a common way—apples as apples and not oranges disguised as apples.

The key to achieving this goal, according to Van Decker, is cooperation. This applies not only to different departments within the enterprise, but also to business units and IT. "They need to partner up on efforts such as these," he says. "Any type of [data] warehousing project needs to include both operational areas and IT so that if changes are made to data definitions, they're all empowered to either make the technological changes, or at least tap on the shoulders of the IT people who manage those sacred cows."

It's critical that data sources are constantly monitored and data terminology updated and shared throughout their organizations.

A common information blueprint
By now it's been demonstrated that utilizing a single data repository is cost-effective, decreases data latency and lowers the risk of inaccurate regulatory and management reporting. We know that a common dictionary must be created and maintained, and that all departments must work together to help define specific terminology. But how are these processes architected so they're effective throughout the organization? Again, simplification is the key.

One way to manage data across the enterprise is through the use of logical data models (LDMs). An LDM is specific to a particular industry, thus providing the flexibility to meet users' diverse business needs. In essence, an LDM is a data warehouse development tool that defines how data is logically organized within a corporation and how those key information elements relate to each other.

Just as a blueprint helps to organize numerous workers on a construction project, an LDM can provide business and IT users a "map" to help navigate the many challenges of integrating multiple disparate data sources into a common data repository of consistent and meaningful information. By employing an LDM, your organization can gain greater control over finance through data stewardship and governance.

The right tool for the right user
An essential step to creating a valuable finance and performance management EDW is to determine how to best give business users access to information that meets their needs.

Enterprises can build rules and calculations into the data warehouse, allowing it to serve as both the originator of the data and the repository for the results. By having all data, business rules and calculations contained in one centralized repository, users have greater transparency into the details behind what is contained within financial reports, and can begin to use automation to get earlier insight into anomalies.

This transparency of business data gives those who need it—including auditors—a simple way to determine how financial results were attained and how they apply to the larger business. As Van Decker puts it, "Transparency ensures that those who need access to data can get it and that there are appropriate alerts and notifications in place when specific events impact the business."

Leading organizations are enabling self-service information delivery, allowing various levels of users—from executive management to power analysts—to gain timely access to the relevant information they need.

Not all information is appropriate for all levels within an organization. Financial analysis reports aren't necessary to human resource personnel; likewise, accountants shouldn't be privy to the company's personnel records. One way to address this issue is to create a reporting portal that gives individual users the appropriate access to all of the company's data to which they are authorized in a simple interface anyone can master.

It is also crucial that the data accessed by these various departments remains, when possible, within the central repository data warehouse and not moved to outside tools. Data movement must be, at the very least, kept to a minimum. Exporting and moving the data unnecessarily brings us full circle with issues of increased data latency, inaccurate or lost data and, consequently, the risk of regulatory noncompliance. Using the tools directly against the data in the warehouse minimizes these risks.

Just as certain data isn't appropriate for all departments, different users also have different reporting tools and requirements, and the system should be set up to allow these varying standards. Providing the right tools for specific levels of use within the EDW and controlling the various users' access to data appropriate to their level and needs are important steps in making business-critical decisions.

Enabling easy, consistent access to the appropriate information provides cost savings, improved control and reduced data latency. Overall, self-service reporting decreases the time users must spend collecting the data and, instead, increases their time spent on analyzing the data to make business-critical decisions.
 
Simply stated
Successful finance and performance management has many dimensions, but the foundation is simplification. In this world where bigger is not always better, simplicity remains the driving force behind information integration. T

Jim Utsler is a senior writer for MSP TechMedia.

Illustration by Jean Tuttle

� Teradata Magazine-September 2006