To manage construction on an oil-and-gas delivery system, you have to like to travel — and not always to places familiar from tourist handbooks. That means a lot of miles logged on back roads and a lot of time spent in places where Internet connections — if they exist — are not always fast and reliable.
Colonial Pipeline is a transport company that delivers gasoline, diesel fuel, and other petroleum-based fuels to communities and businesses stretching from the Gulf Coast to New York Harbor. Its 50 construction managers travel the 5,000-mile system — running projects that range from pipeline expansion to system integrity improvements, to road and river crossings. Snail’s pace E-mail connection times were a major problem for Colonial before it put its managers and everyone else who uses its financial data reporting system on a new Web-based platform two years ago.
Some of the facilities from which the construction managers found themselves forwarding their monthly capital expense reports “were really slow,” recalls Kelly Nodzak, manager of shared systems at Colonial. “They don’t even have T1 lines out there — they’re on phone lines.” On top of that, Colonial’s old homegrown COBOL-based AS/400 system dated from the 1980s and didn’t allow managers to work offline. But with the new system, they can simply download a template, add in the data on their laptops — “in their truck, if they want,” Nodzak says — then reconnect and upload it. “You don’t have the delay of retrieving and sending each time.”
The new performance management software Colonial adopted is a combination of relational databases and a multidimensional OLAP (online analytical processing) database, all on a Web-based platform. It collects data from 150 users, including those construction managers and finance staff. Because it’s not an old-fashioned batched system, it can consolidate all the data into a report to management in minutes instead of overnight, so that last-second corrections on the morning the report is due don’t throw the delivery schedule off by another day.
Best of all, says Nodzak, the people in accounting and budgeting can update and roll forward information just as easily and quickly for other types of reports as well. That means they can drill down to the location, district, department, or corporatewide level to provide real-time information, as needed, about the company’s performance that’s tailored for users at any of these levels.
Many companies have similar stories to tell about how their financial reporting has been transformed. Corporate performance management used to be one of those things every company had — but didn’t always know it had. At most companies, the components include budgeting, forecasting, business modeling, decision support (including dashboarding, traffic lighting, and balanced scorecards), strategic or long-range planning, and consolidation and reporting — along with the security measures needed to protect the data. Other functions that are often grouped with these as CPM include: activity-based management, Six Sigma, economic-value-added (EVA) incentive programs, the human resources information system, compliance, business intelligence, workflow, and approvals.
Until a new generation of sophisticated database applications was introduced and became available on Web-based platforms, however, many companies never thought of these functions as part of the same overarching system. Now, it’s possible to combine them all into one suite and, with some vendors, into a single integrated platform that allows users throughout the company to enter and extract reports from the same data pool in real time.
But the advantages run deeper. A simple spreadsheet-based system of financial reporting is similar to operating on multiple databases — one per user, in fact — which for large companies means that software must be installed, maintained, and upgraded on perhaps hundreds of PCs. It also means a greater likelihood of inaccurate reporting.
CPM systems running on a common platform eliminate such problems, since all users work off a single data warehouse; they also tend to let information flow more freely within the organization, breaking down the silos that departments and divisions set up to protect “their” data from others. At cable giant Adelphia Communications, users work off an integrated OLAP decision-support systems. They also tend to let information flow more freely within the organization, breaking down information silos —
and assuring that the process won’t turn and reverse. The finance department, too, can devote far less time to gathering data and more time to analyzing it.
Meanwhile, a hybrid of an OLAP database and one or more relational databases — commonly called a “ROLAP” system — gives users a more flexible tool. OLAP allows them to perform complex computations very fast. Relational databases, on the other hand, allow users to integrate the text notations and the aggregated transactional information most commonly found in the planning of employee-related expenses, fixed-asset additions, debt-instrument additions, and even supporting detail for key line items that require a zero-based planning approach (e.g., travel, professional fees, dues, and subscriptions). This provides a single repository for all of the company’s performance management components.
“OLAP is a bit like when we were moving systems off the mainframe computer due to the advent of the PC,” says Jim O’Connor, formerly IT operations and OLAP supervisor and currently integration design engineer at Adelphia. “OLAP is empowering users in just the same way.”
Faster and Easier to Use
Speed makes a huge difference. Robert Cox, senior director of financial planning and analysis at Erickson Retirement Communities, says that adopting a hybrid ROLAP system that combines an enterprise resource planning (ERP) system and Essbase has allowed users to “load some source data from the ERP [into Essbase] and it basically does the aggregations, calculates your metrics — all of these things on a scheduled basis.” “Then when people need the information,” he continues, “it’s blisteringly fast to get it. Before, you’d submit it before lunch and when you came back in the middle of the afternoon, your report would be ready. Running a simple report out of an ERP system that had to go through multiple departments and multiple time periods could take several hours and be a batch process. Now, the reports run in seconds.”
Vendors have developed simple and intuitive front ends for CPM systems, resembling the Excel spreadsheets that they replace. “It’s like Excel on the Web,” says Greg Johnson, financial analysis director at insurance giant Aetna, which recently converted from a system largely based on E-mailing spreadsheets back and forth to a combination of OLAP and relational databases from Clarity Systems. Together, these enable the finance department and its counterparts in the company’s various divisions to create customized reports more easily than in the past, making the company less dependent on help from the information technology desk.
That means greater freedom for users to stretch out, developing new analytic approaches tailored to their company’s specific, evolving needs. Some companies have used their new capabilities to develop new charge-back models for pricing the overhead costs of, for example, real estate or information technology’s contribution to their product offerings. The deeper-drill analysis that they can do with a combination of OLAP and relational databases has also helped users to develop multi-year budgeting models that quantify an individual salesperson’s contribution to the bottom line. That helps in making decisions about hiring and firing, and which market or geographic sectors deserve the most attention.
The net result: more accurate and reliable data, greater ownership of data for its users, and greater room for users to apply some real analytic intelligence in telling the story of how the company is performing. “The ability to retrieve data, look at it, decide if it is or isn’t what you want, modify it, retrieve it again — just really changes the dynamics of how the work gets done,” says Erickson’s Cox, “because you’re no longer waiting for information. I think that allows people to be a lot more creative in their processes and try, through trial and error, things that they would never have even tried to do in the past.”
This article is excerpted and adapted from Best Practices in Selecting Performance Management Software: Finance Searches for Increased Flexibility and Control, a report that summarizes suggestions and recommendations from senior finance executives who have successfully implemented CPM systems within their organizations. CFO Research Services and software provider Clarity Systems developed the hypotheses for the research jointly. Clarity funded the research and the publication of the findings; CFO Research Services produced the final report. You may download a copy of the full report by filling out a brief form.
Moving your company’s budgeting, general ledger, and other financial reporting systems to sophisticated databases operating on a Web-based platform is a major step. But for most companies, it’s just the first step in a progression that could see more and more data applications moving into the new environment. In deciding to go ahead with a corporate performance management overhaul, top executives need to make sure that their up-front decisions make it easier, not harder, to expand the system and add users.
Lessons learned by companies that have already gone through the process include:
Take small steps. Don’t try to do too much, too soon. Many companies start with budgeting — the single most critical financial activity — which can serve as the hub of a system that expands to include other activities as well. But even budgeting itself can be broken into sub-components such as HR planning, capital planning, expense planning, revenue planning, and balance sheet and cash flow forecasting. Many companies will choose to start with P&L planning to get a quick, early “win” for their new system by tackling the piece that takes most of their time and is the fastest to integrate.
Bring your users into the planning. Find out what they want to be able to do on a new, Web-based system, and build their comments and suggestions into the vendor selection process. And include users of your company’s business intelligence systems — they might find advantages to using your CPM databases as well.
Take great care in selecting a consultant. Avoid those with extensive technical background but little experience in the finance area. Ideally, your consultant will understand finance and database modeling.
Take plenty of time in designing the system (“blueprinting”) and planning the “go-live.” In particular, make sure your data is cleaned up and the hierarchies that organize it are firmly in place before you replicate them into the new system. And keep top management informed each step of the way, in case a change in direction becomes necessary. Often a consultant will be able to present a few alternative design approaches. Build a prototype to make sure that the design will handle your reporting requirements, allocation methodology, and user-input views, and verify that user responsibility can be assigned the way you need it to be.
Keep your database bite-size. A large OLAP cube can quickly become too big to operate quickly and efficiently. Implement partitions that break the cube up into individual units for different functions before going live — not after.
Plan to add more users sooner rather than later. The point of Web-based CPM is to empower users. If it does, the demand for access will only grow.
Keep the application simple. In particular, make the interface as much as possible like those of systems the users are already familiar with, such as an Excel spreadsheet.
Test vendor candidates with real business cases. This will give you a clear idea of how they address the specific needs of your company’s users. Make sure you see enough of your own business case demonstrated to be comfortable with the vendor’s approach.
Look for single integrated suites that meet your business requirement. A single integrated suite means one installation, one interface to learn, one application to maintain and upgrade, and no requirement to perform complex mapping between independent applications sitting on different database technologies. It also assures that all information users are accessing all information from one common source. A single integrated suite also means a shorter delay before adopting additional performance management components, since you already own and are trained on the technology.
Look for the vendor to use the best database for the process you are looking to solve. OLAP databases are great for performing complex calculations, making changes to your organizational structure, and achieving rapid response times. Relational databases are optimal for the transactional components within the performance management process.