Talking to your CIO to get the data you need. (Workforce Optimization).
Financial executives need data, and lots of it. They need to have that information in a spreadsheet or other format, so that they can analyze it to make strategic decisions.
However, as most CFOs have discovered, getting data on demand and in a familiar format is far from easy. In most cases, the information technology (IT) department controls access to the information, and CFOs wait for batch reports to be run. If they need more data, or need the data in another format, they have to wait for a (probably overburdened) mainframe staff to prepare that information. All this can be inconvenient and time-consuming, and potentially result in lost opportunities.
One basic thing to understand is that CFOs and chief information officers (CIOs) live in different universes. CIOs do not concern themselves with quantitative analysis, pivot tables, regression analysis and correlation. If they do "what-if" calculations at all, it is infrequently.
The finance department needs data no one else cares about, and probably no one else appreciates how important the data is. In the post-Enron world, the finance department is undergoing unprecedented scrutiny, but the IT department is affected little, if any. And the IT department is not required to have instant answers to questions from the CEO and board that might range from the impact of changing the accounting method and restructuring the sales force to determining the ROI on some new purchase.
In short, financial executives deal with content, while CIOs deal with processes and keeping the infrastructure going 24 hours a day, every day of the year. Their world consists of protocols, operating systems, applications software and infrastructure requirements. Understandably, they prefer to have as many things scheduled as possible, and they like doing regular reports on a regular basis. Ad hoc requests represent one more disruption in their day.
Both departments are often overworked, but their duties are so different that there is often a gulf of understanding -- even suspicion -- between them. It is pretty safe to assume that the CIO might not understand just how important getting timely data, in an appropriate format, is to the financial operation.
The IT staff might not realize how dynamic the financial environment is, and how important ad hoc requests can be. In their minds, regularly scheduled batch reports should satisfy the financial department, which should be responsible for integrating data from various sources. In most cases, they are probably right, but in today's environment, financial executives must deal with the unexpected, as well as with routine matters. Such situations require non-routine responses.
Viewing the IT World
To help bridge the division between the departments, it's essential to "speak a little geek," to be conversant enough in technology to command the IT department's respect. That's especially true when the department puts in a funding request for some new technological initiative or when the finance department needs information quickly.
Some history is in order: The IT shop probably started life on the mainframe, using esoteric formats and languages such as VSAM (Virtual Storage Access Method), QSAM (Queued Sequential Access Method) and COBOL (Common Business Oriented Language). Most likely the majority of the older financial data is still in COBOL/VSAM applications.
The IT department probably remains wedded to the mainframe, for good reason. Estimates are that up to 70 percent of the world's business data is still on mainframes. The mainframe's reliability, security and performance are unparalleled in the technological universe. And, in tight economic times, it's essential that companies leverage investments in this proven technology.
However, the IT infrastructure now includes servers (of all shapes, sizes and platforms) and data (in a dozen different formats). The problem is that mainframes don't "speak" Windows or any of the Internet technologies such as HyperText Markup Language (HTML) or eXtensible Markup Language (XML). And those technologies don't "speak" mainframe.
With all that, the CIO is faced with a complicated environment. Getting data to the desktop quickly -- and in a desktop application -- isn't easy. Yet that data is exactly what the CFO needs. Consider the hypothetical case of a large national banking firm that had undergone a merger a few years previously. The CFO was asked to finance the merger of two IT departments, and to find a way to provide single statements for all accounts that individual customers had with the bank. Single reporting was obviously desirable, since it offered the potential of significant cost savings and improved customer service.
To help accomplish these goals, the CFO wanted to be able to access information on demand, without having to wait for a batch process. And, because he did some traveling, he wanted to be able to access the data not only from his desktop, but also from over the Internet.
All this is easier said than done. But it is possible. In trying to meet the CFO's needs, the CIO would probably recommend at least one of five major approaches: an enterprise resource planning (ERP) system, a data warehouse/data mart, an all-relational database, some ad hoc approaches or a direct, real-time method.
The alternatives could be used singly or in various combinations to achieve the desired result. Unfortunately, most of the alternatives are expensive and time-consuming, and suffer from technological and implementation constraints.
An ERP system attempts to integrate all departments and functions across a business in a single computer system. The particular needs of various departments -- including finance, manufacturing and human resources -- are impressed on the system. ERP gets very complicated because each area of the company has its own regard for the importance of the data and how it needs to use that data.
Organizations can choose to develop their own ERP system or purchase one from companies such as SAP, PeopleSoft or Siebel Systems. Whether purchased or built, however, research shows that an ERP system can prove expensive in time, effort, skills and, most of all, money.
The META Group, a market-analyst firm headquartered in Stamford, Conn., researched the total cost of ownership (TCO) of ERP, including hardware, software, professional services and internal staff costs at 63 companies, including small, medium-sized and large companies in a range of industries. The average TCO was $15 million (the highest was $300 million and the lowest $400,000).
The TCO numbers include figures for getting the software installed and the two years afterwards, which is when the real costs of maintaining, upgrading and optimizing the system are incurred. META came up with one statistic showing that ERP is expensive no matter what kind of company uses it, with the TCO for each user over that period at $53,320.
Even if the expense is not an issue, time might be. It takes a long time to set up an ERP system. Implementation requires changes to the entire system currently in place. Hardware, operating systems, databases and applications may all have to be rewritten, returned or expanded.
Data Warehouse/Data Mart
Data warehouses or data marts are copies of data periodically extracted from older legacy applications, converted to common standards and made accessible for user analysis, querying and reporting. In practice, data warehousing focuses on a single large server or mainframe that provides a consolidation point for enterprise data from diverse systems. Data marting, conversely, deals almost exclusively with servicing a distinct community of knowledge workers.
A company considering a data warehouse would have to factor in the maintenance of the system and the data, and managing the capacity and performance of the warehouse. Keeping the data current and consistent is a never-ending task. (Data marts frequently grow larger than the company's original concept of the entire data warehouse.) When considering a data warehouse or data mart, an organization needs to have the long view. It typically is far more expensive (and complex) to maintain a data warehouse than it is to build one.
A relational database is middleware that links files together as required. Relationships between files are created by comparing data, such as account numbers and names. IBM's DB2 is an example of a relational database for a multi-user environment.
While an all-relational database would be one step to providing the integrated statements the bank in this example desired, moving to an all relational data approach has several drawbacks, including:
* Performance. (DB2, for instance, is not as fast as native VSAM.)
* Time. (Most applications have to be rewritten to run under the relational database.)
* Expense. (Licenses are priced based on CPU size.)
* Hardware. (Servers and other computers may have to be upgraded or replaced.)
* Security. (Security can be an extensive headache.)
Ad Hoc Processes
File transfer protocol (FTP), network file systems (NFS), screen-scraping or "sneakernet" are all techniques that IT organizations have adapted to extract data from the mainframe for use on the desktop. Ad hoc processes are typically stopgap measures designed to solve immediate problems as quickly and easily as possible. Companies use ad hoc methods because they cannot, or will not, commit to a more elegant and effective solution.
Although ad hoc approaches might appear cost-effective at first glance, in virtually every case, the user has to ask the mainframe staff for the data. The request is typically acted upon in batch mode, and the data is forwarded the next day. The data is usually outdated, inapproriate or is not in the desired format. By themselves, ad hoc methods could not meet the finance department's need for data-on-demand or for integrated statements.
Direct, Real-Time Access
A direct, real-time access method probably holds the most promise for getting real-time data to the desktop -- ideally, enabling Windows(r) applications and browsers to access mainframe data in real time, and to use that data in a familiar off-the-shelf application (Excel, Word) meeting the CFO's need for current data.
The direct approach can also simultaneously retrieve data from multiple files and from multiple mainframes and join the data in Windows or browser format, making the single account statement possible. This approach is unique in that it operates without the traditional process of brokering, duplicating or migrating the data. This method provides just the specific data records or fields required. The requesting application can access the data -- on demand -- through intranets, extranets or the Internet without requiring additional IT resources, making Web access possible.
Choosing the Right Methodology
Whether to produce a single report for multiple accounts, or any of the multitude of reports and documents needed in the financial world, financial executives need to understand the CIO's world a little.
There are alternative methods for accessing data. Many are long, arduous and possibly expensive. Even the emerging technologies, such as the much-hyped XML, require significant development efforts as well as an in-depth understanding of the company's business processes.
Choosing the right methodology is an exercise in understanding the currency and accuracy required for the given tasks. The CFO needs to understand the ramifications of redeveloping applications using a more modem infrastructure, moving applications to an all-relational database, building a data warehouse or data mart, relying on ad hoc approaches or utilizing a real-time, direct methodology.
By asking the appropriate questions, CFOs can help assure that any IT changes meet the finance department's needs for delivering current data to the desktop, when needed, in a format that the user can manipulate. By being aware of the alternatives, the financial executive will also be in a much better position to assure that any IT changes are good investments for the organization's long-term viability.
RELATED ARTICLE: Data Access: Questions To Ask
* What's the best way for me to get the information I need, when I need it?
* Can I get real-time data?
* Can I get information from various sources integrated into a single report?
* Can I filter the request so I get just what I need?
* Can I get the data myself, without waiting for a batch process?
* Can I get information in a spreadsheet format?
* Will I be able to access data using my browser?
* How much will the alternatives cost?
* Are there any hidden costs?
* How much data is likely to remain unchanged, regardless of the alternative chosen?
* What do the approaches take in terms of ongoing support and maintenance?
* Do the approaches truly integrate the data from different sources and present it in a single format?
* How quickly can we implement the various approaches?
Joe Cote is vice president of marketing for Xbridge Systems Inc. (www.Xbridgesystems.com), a Los Gatos, Calif., firm that develops and delivers solutions to provide real-time access to mainframe data.
|Printer friendly Cite/link Email Feedback|
|Date:||Jul 1, 2002|
|Previous Article:||What's normal in derivatives accounting. (Derivatives).|
|Next Article:||Buyers (and sellers): beware of new rules. (Accounting).|