Larry Reader's patience had worn thin. It was becoming increasingly clear to him that too many financial decisions rested on homegrown spreadsheets packed with too much impenetrable data and too few answers. The result was a crippling condition that Reader calls "spreadsheet overload" but which many others refer to as "spreadsheet hell."
Reader, the CFO of Duke Manufacturing, a privately held manufacturer of commercial-food-service equipment for restaurants, hospitals, and schools, was about to join the I'm-fed-up-with-spreadsheets club. Membership often hinges on a sudden realization that, for all their virtues, spreadsheets are the IT equivalent of a screwdriver: useful for many things but too often employed for the wrong chore simply because they happen to be close at hand. And cheap.
But as anyone who has ever used a screwdriver to chisel wood, poke holes, scrape paint, or pry things apart knows all too well, sooner or later they break under the strain of all that misuse.
That was the case at Duke, where an already stretched finance department would assemble spreadsheet data, key it into enterprise resource planning (ERP) systems, and then spend far too much time tracing the sources of resulting errors. "It was difficult to determine where the original data even came from," says Reader. "Forecasting and budgeting were tough activities, to say the least."
That has long been a common lament, and has, over the years, pushed companies to explore a range of business intelligence (BI) software applications to either augment spreadsheets or, at bolder companies, virtually replace them. That slow migration got a boost from the Internet, as companies began to realize that Web-based BI software allowed a new freedom in how users input and shared the data at the heart of most budgeting-and-forecasting operations.
Now, companies may be "on the cusp of a new generation of business-intelligence tools," declares Boris Evelson, principal analyst in BI software at Forrester Research.
Just in time, too. By some estimates, corporate digital data volumes will grow 30 percent a year to reach 1 zettabyte (a zettabyte is a "1" followed by 21 zeroes) by 2010. As all that data ushers forth from ERP systems, companies have a choice: drown in it or leverage it for better decision-making. The latter is nearly impossible without BI and related categories of software (such as analytics), which are optimized to extract nuggets of gold from mounds of informational ore.
The new generation that Evelson alludes to is characterized by two things: vendor consolidation and new licensing options. While the BI market remains crowded, it has become noticeably less so over the past 18 months as top-tier vendors such as IBM, SAP, and Oracle have made large-scale acquisitions to beef up their BI offerings. And, from the high end to the low end, many vendors now offer "software as a service" (SaaS) licensing options that essentially allow customers to rent the software rather than spend large sums to buy it outright.
The latter development may provide many more companies with a cure for their spreadsheet headaches. "In the past, if there were any problem whatsoever with the spreadsheet we would spend all day figuring it out," says Bryan Rogers, vice president of finance at Chicago-based insurer Unitrin Direct.
A strategic plan to increase sales, for example, required so much rekeying, reloading, and rechecking of data that weeks were wasted, says Rogers. "It got to be that everyone had his or her own little Excel model, with underwriting having one and sales another," he says. "Excel is an excellent spreadsheet, but a poor platform for collaborative planning across an organization."
Rogers weighed several BI options but found the prices too steep, until he learned about the SaaS offering from Adaptive Planning, which allowed him to introduce BI for just $35,000 a year.
In addition to its core BI capabilities, "the tool was especially valuable after we acquired another insurance company," Rogers says, "since we could add their specific information into our own planning system and use our own assumptions, capitalizing rules, and per-unit cost drivers."
Duke Manufacturing also went the SaaS route, choosing Host Analytics's Host Budget/Host Forecaster program, which allows Reader to slice and dice performance data supplied by Duke's 90 different product groups. One upshot is greater accountability on a very granular level. "We can look at each product group in terms of fixed and variable material costs and labor costs, and then compare them against sales in previous years and against our current budget to see if we're budgeting properly," Reader explains.
At Homedics, a designer and distributor of personal-care products in Commerce Township, Michigan, "we used to joke that we had four definitions of net sales," says CFO Bill Carroll. That's a common problem at companies that rely on a hodgepodge of spreadsheets or multiple BI tools. So Homedics standardized on one product, QlikView, which (in conjunction with ERP software from Oracle) gave the company a set of standard definitions thanks to its use of a series of fields where commonly defined data converge. "The ability to get quality, transparent information quickly helps me to chart one course of action versus another much faster, and as a result leads to quicker decisions," Carroll says. "It's all about being decisive. And in the current economic climate, smart planning and execution have never been more important."
Homedics used the QlikView tool initially for sales analyses; salespeople could fire up their laptops and drill down into data that shows sales per category per customer per region, right down to the materials contained in each product. "Sure you can do this with Excel," Carroll says. "It's just a question of how much time you have to do it."
But for all their powers, BI tools of all vintages face a monumental challenge in making spreadsheets truly obsolete (see "Off Course" at the end of this article). Microsoft gauges the number of Excel users worldwide at more than 400 million, and Forrester Research estimates 50 to 80 percent of enterprises still use stand-alone spreadsheets for critical applications like financial reporting.
At security software giant Symantec, Excel spreadsheets still collect and manage data from operating entities, says CFO James Beer. Familiarity and ease of use win over his internal customers. "Frankly, I'm agnostic," he says. "We believe that the best-run companies typically have a heterogeneous technology environment, with the onus on IT to integrate it all seamlessly."
Not that Symantec isn't brimming with BI tools; it just insists that they all be able to coexist. An Oracle data warehouse serves as the hub, supported by analytical tools from BusinessObjects and Hyperion. A Star Analytics integration server moves data out to operating units.
The company fully acknowledges that while spreadsheets are comfortable for many employees, they pose challenges beyond the difficulties of doing rapid, sophisticated data analysis. Security is one; BI systems can keep critical data confined to well-protected servers in a way that easily circulated spreadsheets simply can't.
Another advantage to BI, harking back to the multiple definitions of net sales that Homedics grappled with, is its ability to get everyone on the same page. "We all want one source of truth," says Jeff Brobst, vice president of finance at Symantec. Excel's vast capacity and its design as a personal productivity tool make data integrity and ownership very hard to verify. As Brobst warns, "You can go down a rat's nest."
Kurt Schlegel, vice president at technology research firm Gartner, offers an example. "Say you have a product-recall issue and need to find out fast which of your plants produced the tainted component," he says. Could a spreadsheet-based system help you quickly answer questions such as, If consumers are at risk, how much product must be recalled? Is it on shelves, in warehouses, in transit? What are the consequences of doing too little? Too much? What's the bottom line? "With million-row spreadsheets not uncommon and different users mashing up the data," says Schlegel, "you might not know where the data came from, in what form, and whether it is trustworthy."
This tug-of-war between data reliability and control will plague users of stand-alone spreadsheets for the foreseeable future, Evelson warns. He is adamant that in the current economic climate, BI tools "are absolutely essential, because companies compete based on the decisions they make."
The cost depends largely on scope. At the high end it's entirely possible to spend several million dollars to roll out a traditionally licensed system across a global organization. At the low end, the SaaS model allows companies to pay on a per-user, per-month basis or a similar arrangement, and may allow a toe in the water for only a few thousand dollars.
The cost-benefit analysis has to take into account just what BI systems are for. In the past, they provided a slicker way to produce reports and develop budgets. Today, as Evelson says, they can become a competitive differentiator when they are used to analyze customer demographic data, sales trends, macroeconomic indicators, and a host of other measures.
Even an evangelist such as Evelson has no doubts that spreadsheets are here to stay. They are "a permanent fixture in enterprises because no other analytical application is as ubiquitous and familiar," he says. And for all their advances, BI tools are still criticized for being too rigid and hard to learn, too IT-intensive and difficult to mix and match.
But as finance departments grow more frustrated with spreadsheets and more interested in tapping the powers of analytics, as IT departments devote more resources to solving parts of the BI puzzle, and as vendors channel more and more resources into their applications, these complaints may abate. As Rogers says, "The time saved in finance thanks to BI tools is now devoted to more-strategic imperatives."
Russ Banham is a contributing editor to CFO.
Cutting Through the Static
When Jeff Nemy joined Northern California Public Broadcasting as CFO in 2006, the company was approximately two months away from acquiring two TV stations. That would complicate budgeting in the best of times, but Nemy soon found out it was worse than he feared. When he asked for a copy of the current forecast, he learned that there wasn't one.
Worse, creating one would require the consolidation of approximately 600 spreadsheets from nearly 200 business units. "Consolidating all those spreadsheets would take up to six weeks," Nemy says. "It almost made me walk out the door." But he didn't. A budget was produced and then a search was launched, for business planning and consolidation software (virtually synonymous with BI). Nemy ended up implementing OutlookSoft (now owned by SAP) for everything from budget consolidations to dashboard analysis to the establishment of three-year rolling forecasts. "We've cut many weeks off our budgeting-and-forecasting process, which provided important time for value-added analysis," he says. "I get reports now in a 'cube' that I can slice and dice by department and project."
The software cost less than $250,000, which Nemy says the company recouped in the first year, thanks in part to the system flagging a looming budget shortfall in time to cut discretionary spend and boost revenue with added pledge days. In its second year, says Nemy, "I would estimate the software produced a fourfold ROI."
Although spreadsheets haven't disappeared (in fact, Excel serves as the front end to OutlookSoft for departmental users), Nemy says he has firm control over the templates. "They're locked, so people can put in numbers only where I want them," he explains. "My job is remarkably easy compared with what it was before. Spreadsheets alone are folly for organizations with any level of complexity." — R.B.
Blame the spreadsheet or the user?
The spreadsheet is a wonderful tool that no one ever learns how to use properly, says Richard Block, a CFO leadership partner at Tatum LLC and adjunct professor of management accounting at Babson College, where he teaches courses in managing operations and cost. "No one seems to take a course in it, but instead typically learns it from the person sitting next to them in the office." Combine that lack of knowledge with bad practices and the spreadsheet's inherent power, he says, and "the potential for misuse is high."
To ensure documentation is accurate and complete, Block, a former CFO, advises that users clearly identify the purpose of the spreadsheet; note who wrote it, when, and how calculations were performed; clearly label specific columns and rows; and avoid a common breakdown — long comments typed into cells that block data in adjoining cells — by clicking the "Wrap Text" feature.
Most important, he cautions against dumping a massive spreadsheet at the CFO's cyber-doorway. "The smart thing is to delete all tabs that are not relevant to what you need the CFO to look at, copying and pasting [into a new file] only those tabs that require perusal," he says. "If more scrutiny is requested, you can always send the whole file later." — R.B.