Sometimes, getting smarter requires a little dumbing down.
Take the case of Deltek Systems, the Herndon, Virginia-based application software provider. A few years back, management at Deltek wanted to improve the way workers made decisions. After looking at several approaches, Deltek executives decided to install business intelligence tools and applications across the entire company.
There was just one hitch: Enterprisewide rollouts are notoriously hard to implement in one fell swoop. So, rather than install the systems and software all at once, Deltek took it slow. “We decided to install it in finance first and then roll it out department by department,” explains CFO Lori Becker. “We just wanted to perfect the solution before we introduced it elsewhere.”
Perfection evidently was realized: Today Deltek staff in human resources, marketing, finance, and sales can go to a portal and pull up diverse financial, accounting, customer, and employee data for analysis. “All of it is integrated seamlessly and pushed to whomever may need it for decision-making purposes,” crows Becker. “We’ve now got a full-enterprise BI solution.”
That’s unusual. While the search for business intelligence is booming — sales of BI products rose to nearly $4 billion in 2002 — full-blown BI enterprise deployments are still relatively rare. Typically, BI tools from several vendors are intermingled and integrated with various enterprise resource planning (ERP) programs, which explains the entry of big time ERP players like SAP and PeopleSoft into the BI market.
In addition, business intelligence vendors are beginning to plant their flags in other parts of the BI galaxy. Integrated BI tool vendors like Business Objects and data warehouse builders like Informatica, for example, are getting into the applications game, while app providers are adding tool sets and data warehouses to their product offerings.
All this crossover has created a dizzying array of BI products on the market — leading many corporate clients to pick and chose BI products from multiple vendors.
Adding to the available options: Over the past few years, scores of new business intelligence vendors have arrived on the scene. Currently, there are more than 100 software makers hawking a wide range of BI-type stuff, from specific tools for data extraction, mining, and query to off-the-shelf, prepackaged analytical applications. Many BI vendors also offer data warehousing systems.
Given the great flux in the BI universe, industry consolidation seems inevitable — and fast-approaching. In fact, in mid-December, publicly traded Cognos announced that it was acquiring privately held BI vendor Adaytum in a deal valued at $160 million.
No wonder BI analysts at research firms like IDC, META Group, and Gartner advocate a phased-in implementation of BI software, rather than the whole-hog enterprise approach. Surprisingly, so do many BI vendors.
“Prior to the dotcom glut, everyone in the software business had the same dream — build big, which meant build complete enterprise solutions,” says Sanjay Poonen, vice president of worldwide marketing at Informatica Corp., a Redwood City, California-based BI solutions provider.
“We, too, suffered from the same ‘field of dreams’ illusion,” Poonen concedes. “But over the course of the last year, based on customer feedback, we’ve decided to market more of a modular, piece-by-piece, division-by-division BI offering that can be rolled out gradually to encompass the entire enterprise.”
By taking the step-by-step approach, businesses arguably take less of a financial gamble. Admittedly, they may encounter common IT implementation problems like change management and employee resistance. But those problems occur on a much smaller — and more manageable — scale.
Further, gradual deployment enables corporates to gauge return on investment — before doling out more dollars. “Most companies these days are loath to fund large all-encompassing efforts like a corporatewide BI installation,” says David Folger, vice president of Web and collaboration strategies at META Group. “They’re focused instead on solutions to critical, tactical problems where the ROI is clear.”
King of ROI
And when it comes to ROI, business intelligence appears to be king. Says IDC’s Miller, “The return is extraordinary.”
According to research conducted by IDC, the average analytics projects yield an average ROI of 431 percent. In the consultancy’s most recent survey, respondents said they had realized returns ranging from 17 percent to more than 2,000 percent on business intelligence initiatives. The average five-year ROI? 317 percent.
Of course, CFOs may view triple-digit numbers like that with skepticism. And many BI applications are not direct revenue-generators. Hence, computing the ROI of BI rollouts can be tricky.
Indeed, CFO Becker says she is still calculating the ROI from Deltek’s BI project, noting the difficulty of quantifying soft benefits like improved decision-making and more efficient communication. The company’s total BI investment, including the software, hardware, user licenses, staffing increases, and training was about $168,000. “We’ve estimated that Deltek’s account managers saved about five hours a week using the BI tools,” Becker notes. “Since we have 21 account managers, that works out to 5,250 hours per year of saved time, which is considerable.”
More to the point is CFO Lawrence Jankovic, who spearheaded the BI strategy at Bedford, Freeman & Worth, a New York-based higher education publishing house. Jankovic says his data shows the publisher has realized a 1,163 percent ROI from its BI project, which was implemented in 2001. “I was the advocate here and had to get buy-in from our project managers, which wasn’t easy,” says Jankovic. “I had to continuously sell the idea that BI is a decision-making tool that would provide immediate utility.” He pauses. “Given the ROI, I think I proved my point.”
Bedford, Freeman & Worth was faced with a difficult financial picture and, says Jankovic, needed to improve the quality of its decision-making. Each of the company’s project managers oversee a group of textbooks and related supplementary material, such as videos and multimedia, that support teaching the textbook content in class.
It’s a tall task. “They’ve got to figure out how much it will cost to develop the project and promote it, including author royalties and other editorial expenses, calculate how much cash will be produced over the course of time, including sales of the ancillary material,” notes Jankovic. “Then they have to give senior management a forecast of expected profitability.”
The problem was pulling all this data together to get a snapshot of the entire enterprise. “The project managers were spending an enormous amount of time entering data to support the budgeting and planning process,” says Jankovic, “and yet we were still unable to get a complete picture of all the elements that affected a project’s profitability.”
The company turned to business intelligence, deploying a system designed by Adaytum. “They needed a way to roll up data in such a way as to determine whether a particular project made sense from a profitability standpoint,” says Mark Stimson, vice president of international marketing at Adaytum. “Then, they needed to analyze this information across the spectrum of projects.”
CFO Jankovic says BI software helps with the analyzing. Project managers feed data and enter queries into the Adaytum packaged application to extract answers asking questions like, “What if revenues are below expectations for this project — how will that affect projected costs for the project and across the enterprise?”
“Finance does less from the top now, as far as asking for information,” says Jankovic. “We’re more like financial advisers now.”
Living in a Spreadsheet World
If sellers of BI software and users of BI software recommend a phased-in deployment strategy, that raises the obvious question: where to begin the phasing-in?
The answer is also obvious. “Finance is the natural entry point,” asserts Stimson, “since most of the planning and performance management processes in an organization essentially filter up and feed to finance.”
In fact, IDC estimates that 70 percent of the performance metrics a business tracks are financial in nature — and therefore reside in finance. Of course, if finance is the starting block for a BI enterprise strategy, then a company’s CFO is the likely choice to be the BI advocate.
“CFOs have been doing data analyses for as long as there have been enterprises,” points out Howard Dresner, vice president and research director at Gartner. “They live in a spreadsheet world, understand BI better than anyone, are usually tech savvy as well as business savvy, and have proven via all those ERP installations in the ’90s that they’re great project facilitators.”
By implementing business intelligence first in finance, a CFO can extract value right away. As in Jankovic’s case, that gives the finance chief ammunition to champion BI across the enterprise. “Success breeds success,” Dresner says. “It’s either a modular approach or the Big Bang, where you hope to incorporate all the data across the enterprise through a single effort. That takes enormous time and money and involves tremendous change management and people issues.”
Others agree. “In today’s economic climate,” says Henry Morris, group vice president for applications at IDC, “companies want to show ROI more rapidly. By focusing BI on one area and proving the return fits the organization’s physics, you have an argument to move it to other parts of the enterprise.”
The key to a phased-in approach, experts say, is remembering that it is just the beginning of the journey. “A particular department can extract value from BI,” says Paul Fitzpatrick, senior manager of product marketing at Cognos. “But the real value proposition is in leveraging it across the enterprise.”
Rolling out a BI system across several departments can foster better communication between employees. “Often, departments speak different languages and define things differently,” says Fitzpatrick. “For example, HR may define ‘customer’ in a different way than marketing and finance, and consequently might be under or over-allocating personnel resources to a particular customer segment. BI helps organizations act in a coordinated way.”
That’s what happened at Consolidated Container Co. The Atlanta-based plastic container manufacturer, with 2001 revenues of $786 million, came into being as an amalgamation of several blow-molding companies. “We had tremendous fragmentation,” notes Joe Baird, vice president of IT. “There just wasn’t a single source of good information to support decisions.”
Like other companies, Consolidated Container followed a modular, phased-in approach of its BI software. The application, designed by Business Objects, codifies, integrates, and streamlines disparate data from different data sources. “Everyone can reach beyond their own silos to analyze the same information or create reports,” Baird says. “And the data is being pulled across the enterprise.”
The benefit of all this pushing and pulling of data? Good business intelligence, claims Stimson, “like the intelligence military commanders receive from the field to make decisions.”