Manage Learn to apply best practices and optimize your operations.

Payback Time for Business Intelligence Software at Midsized Businesses

Business intelligence is finally paying off for midsized companies, especially those that nail down business needs and choose products wisely.

Irving "Bubba" Tyler had been quietly laying the data foundation for business intelligence software tools for years when suddenly the call for action came. In 1999, Quaker Chemicals, Tyler's employer, hit the BI hot button. The $400 million Conshohoken, Pa., maker of custom chemical compounds for the steel-making process was undergoing a massive structural reorganization. To pull it off, Quaker had to be able to spread information to employees around the world quickly, efficiently and accurately.

BI was the change agent of choice.

"Because of our change in focus from geographic to customer segmentation, because we need to manage the business from all over the world, and because people here often wear more than one hat, we had to deliver consistent information," says Tyler, who is CIO. "The only way for us to do that was to standardize using BI."

BI -- that oft-claimed territory encompassing everything from database to query technology -- has finally come of age, and not just for technical reasons. Yes, the number of tools geared toward and priced for the midmarket has increased. The Web-based technology needed to access and distribute the data is mature and readily available. But the real ace in the hole is that the business community is demanding it.

"CIOs are struggling to figure this out, especially since they are being asked to make BI 'happen,' " says Mark Smith, CEO of Ventana Research, a research and consulting firm in San Mateo, Calif. "The first step is to think about the foundation of BI, to look at this thing from a people, process and information perspective so that you understand the business need."

That need was palpable at Quaker Chemicals, which had reorganized itself along business rather than regional lines so it could run more efficiently, develop products quicker and solve problems once instead of multiple times. "We have labs all over the world and we knew that if we looked at global rather than just local needs, we would have a much more efficient research and development operation and transfer of knowledge," says Michael Barry, Quaker's vice president of industrial metalworking and coding. "We're selling our knowledge as much as we are our chemicals, so we had to be able to grab that and use it everywhere."

That meant harnessing and distributing the knowledge being generated by the company's 1,400 employees in 15 countries. Using a data model and data warehouse that he started developing in the mid-1990s, Tyler crafted a BI system based on technology from SAS Institute Inc. It delivers financial, sales, research and other data using a browser-based query tool. Static reports are available as well. And people are using it. In 2003, the system fielded 40,000 queries; in 2004, that number was well over 65,000.

Barry is convinced that without BI and Tyler's combination of business and technology skills, Quaker would never have pulled off its corporate overhaul. "We could not have done this without IT and these systems, and without a team that understood the business," he says.


BI Outsourcers Face Challenges, Too

CIOs Joe Sullivan and Tim Theinz may not know each other, but they share an unlikely bond. They both use business intelligence software to help their companies' clients increase loyalty.

Sullivan works for the loyalty division of Carrollton, Texas-based UniFocus, which helps hotels, restaurants and casinos hold on to their customers and employees for as long as possible. Theinz's position at Noel-Levitz Inc. places him in the equally volatile world of the college student. His job, and that of this Denver division of financial aid company Sallie Mae, is to help colleges and universities recruit and retain students better.

During the last four years, these two companies -- both essentially BI outsourcers -- have been moving their clients toward a more self-service analytics model. They've faced different challenges. UniFocus discovered that many of its customers weren't ready to embrace the analytics service and didn't want to pay extra for it. But the investment has delivered in other ways. "It hasn't measured up to our dreams of grandeur in terms of increased revenue, but it has improved our product and given us an edge on the sales front," Sullivan says.

For Noel-Levitz, the more self-service model means universities can get analysis of their student recruitment data in minutes instead of days. "We've gone from having to grab data from FTP sites and needing a week or more to deliver results to scoring the data in real-time," says Theinz. "It's pretty neat."

Both companies have had to make adjustments to their service based on the tech savvy of their varied customer base. They also must pay careful attention to the quality of the data and the questions analysts ask. Getting the data into the right format is more important than anything else, both agree.

Loyalty is a shifting commodity, which is why both companies are trying to better understand their client data. UniFocus is developing statistical models to help turn its clients' survey results into a set of action items, while Noel-Levitz is building a service that will help universities figure out how to keep their financial aid costs down. "It's all about running the business better," says Theinz. -- S.L.W.

The Business Intelligence High

Quaker Chemicals is in good company climbing the BI hill. Warn Industries, Emergency Medical Associates, United Pipe & Supply and many other midmarket companies are figuring out how to use analytics and more sophisticated data infrastructures to better understand their businesses. According to data from Cambridge, Mass.-based Forrester Research Inc., 39% of 925 midsized businesses surveyed plan to purchase BI software this year, up from 35% in 2004.

The payoffs can be big as companies are able to recognize and act on industry trends more quickly; gain a firmer grasp on their financials; unearth potential problems related to their raw materials, manufacturing processes, or distributor networks; or just make it easier for users to get information.

Warn Industries Inc., a $58 million manufacturer of off-road equipment and accessories in Clackamas, Ore., figures its BI investment has paid for itself several times over simply by enabling employees to access data without having to ask IT to generate a report. "We knew it would help with the problem of data access, but we had no idea how much, or how popular the system would be," says Travis Pierce, the company's IT operations manager and head of IT strategy.

Indeed, enabling a more self-service culture is one of the potential upsides of BI, particularly for resource-strapped organizations like Warn Industries. "If you can deliver fast, consistent information, the investment in BI will pay off," says Smith of Ventana Research. "Users can get what they need much more independently."

But those benefits come with a price. There are still technology, cultural and change management hurdles to cross. And CIOs looking to invest in new BI technology must wrestle with the nuances of a shifting vendor landscape. "BI is hot right now, so many vendors are claiming it," says William McKnight, president of McKnight Associates, a BI-focused consulting and market intelligence firm in Dallas. "That makes it easy to get caught up in the hype and hoopla."

For larger enterprises, BI typically consists of online analytical processing (OLAP) cubes, predictive modeling, a vast reporting infrastructure and high-end query tools used by a core group of business analysts. At smaller companies, the BI system may comprise nothing more than a monthly sales report generated in a spreadsheet and distributed via hard copy. For most midmarket organizations, BI systems typically consist of static and dynamic reporting, query and analytic tools and a data warehouse or several data marts.

Some vendors are now staking a specific claim in the midmarket, including QlikTech International AB, Databeacon Inc., Vanguard Software Corp., Corporate Radar and Microsoft Corp. All provide analytic tools that are designed for relatively quick and less expensive implementations than enterprise-level products require. Products from QlikTech and Databeacon, for example, allow companies to analyze data directly from operational systems, such as human resources or finance applications. "Many of the midsized organizations we talk to are looking for something they can add into current environments without a big IT project, so that's our model," says Hans Galldin, president and CEO of Databeacon.

Microsoft is pursuing the midmarket with its own products as well as via partnerships with vendors like Databeacon. As Ventana's Smith notes, the Microsoft BI option has a definite appeal for midsized companies, particularly ones that are already "leveraging Excel and SharePoint and that have limited IT resources and minimal BI budget."

Then there are the enterprise vendors heading downstream to the expanding midmarket. "Many tech vendors view the vast number of U.S. small and midsized businesses as a bastion of growth in an otherwise bland, modestly growing market for IT," says Meredith Morris, an analyst with Forrester Research.

Topping the list of enterprise players are SAS Institute, IBM, Business Objects SA, Cognos Inc. and Microstrategy Inc. All offer comprehensive BI product suites that include data warehousing, data extraction, data mining, query, analytics and reporting tools. Some of them are also developing products packaged specifically for the midmarket. In January, for example, Business Objects unveiled its Crystal Reports Server XI, a $7,500 midmarket platform for distributing and managing reports over the Internet.


Tips for Success

Building a BI system involves steps in three areas:


  • Work with the business side to understand the kinds of information users need to make better decisions.
  • Be willing to take a stand when demands regarding reports and analytic tools don't match business objectives. You and your people will pay the price if you don't.
  • Establish yourself as a partner in making the company better, and offer practical alternatives to pie-in-the-sky requests.


  • Understand your data, from where it resides to whether it's clean enough to analyze.
  • Pick an analytics solution that matches your business -- is a data warehouse overkill, or an absolute necessity?
  • Leverage, leverage, leverage by building interfaces and report structures that can be re-used.


  • Be willing to bend -- the best BI system in the world is no good if users don't like it or use it.
  • Build success by starting with a small number of query options or reports and making sure they work.
  • Find a BI evangelist who is outside of IT; he will bring more trust and legitimacy than you'd imagine possible. --S.L.W.

Building a BI System: Pressure for Progress

Adding to the pressure of deciphering a complicated vendor market, some CIOs fear BI will turn into a high-profile, high-dollar failure. Building a BI system can be a long, slow process that requires replacing or consolidating aging systems, developing in-house expertise and courting a user community that may be less than willing to try something new.

"What midmarket CIOs need to remember is that BI can be successful if they keep it simple, at least at first," says McKnight of McKnight Associates.

Simplicity is a principle that Pierce has applied since delving into BI at Warn Industries more than two years ago. At the time, a significant portion of Warn Industries' 600-plus employees, particularly in finance and sales, were frustrated by both the quality of their business reports and the amount of time it took to get them. "If a finance person needed to get sales totals for a particular region or product, it would take us up to two weeks to pull that data," says Pierce. "No one was happy."

Still, the money for high-end BI software wasn't available, nor were the people needed to maintain it. The good news was that most of the data that needed to be analyzed was in one system -- the company's JD Edwards ERP system -- and the bulk of Warn Industries' employees work at one site.

Warn Industries ultimately chose QlikTech's QlikView technology, which Pierce says cost about 75% less than some of the enterprise-level systems the company considered. QlikView uses extraction technology to pull data from the ERP system. Users then access the data via a browser-based analytics application that's customized for each user community. So far Warn Industries has about 20 of these analytics applets, which deliver information on everything from parts-delivery performance to detailed general ledger reporting to manufacturing costs.

Next up is using the tools to enable the company's Six Sigma project. "We're looking to improve our overall performance, and our BI system is going to be critical for that," says Pierce.

BI Project Barrier: Culture Clash

Regardless of decisions made on the technology front, the biggest hurdles BI projects face are often cultural ones. "BI is really about creating a culture of information as opposed to relying on a top-down view of the facts," says McKnight. "That means staying just a little ahead of what the business community requires."

At United Pipe & Supply, CIO Mike Green has been chipping away at cultural change for almost five years. After modernizing the architecture and applications at the $150 million Portland, Ore., distributor of irrigation, pumps and waterworks systems, Green set his sights on enabling better analysis of the company's data and helping users understand the value of that. Green and other senior executives have been driving these initiatives to help United Pipe & Supply employees get a tighter grip on the company's business operations, particularly pertaining to product distribution and related costs.

"It can be tough because there's a group of managers that has worked almost their entire career at this company, and none of them have ever had this kind of access to information," says Green. "I'm delivering something to them that they don't recognize and that they don't necessarily want."

In Green's case, help has come in the form of a manager in the company's purchasing department who is greasing the skids for analytics on the user side. He works with managers outside his department and gets other people excited about the potential of BI for improving the business. "Most CIOs are in the position of being an arbiter of change in the organization, and they look to technology to help with that," says Green. "But if you have someone outside of IT who gets it and helps lead the way on the concept, it's a wonderful thing."

Spearheading cultural change is also about becoming a true partner in the business. Listening carefully to those on the business side and delivering small but substantial victories is a formula Scott Richards has applied as CIO of Emergency Medical Associates (EMA), an $80 million physician-owned group that provides emergency department services and staff to hospitals in New York and New Jersey.

Back in the late 1990s, EMA delivered monthly paper reports, most of which were built on 50-day-old data, to its physicians. Those reports contained bird's-eye-view information about contract management, billing, patient and hospital satisfaction, the frequency of diagnostic procedures, and even patient waiting times. "The doctors started to call and question the data in these reports, wanting more details about certain metrics," says Richards. "So we got their feedback, incorporated that incrementally, and built from there."

Now EMA is taking its BI system, which runs on software from Business Objects, to the next level by making it more dynamic. Richards and his staff have again worked with doctors to figure out which metrics should be assessed in daily, weekly and monthly increments, which data points should be trended, and even areas where performance-threshold alerts are needed. Physicians are accessing this information through EMA's secure portal and querying it themselves. "They're comfortable now doing this online, they tell their colleagues how great the tools are, and it has grown organically because of that," Richards says. The physician owners believe this will ultimately make the company more profitable because it will allow both trends and problems to surface and then be dealt with more quickly.

Yet not all users embrace the self-service approach so emphatically, and sometimes a hybrid approach must be employed. When many of Quaker Chemicals' field sales people refused to use the BI query tools, essentially claiming they were too difficult to navigate, Tyler grudgingly decided to e-mail the sales people weekly reports in addition to providing them with the Web-based query option. "We tried our best to get them to use the existing system, but ultimately we had to do what was good for the business," says Tyler. "Helping our people sell more products is more important than winning a technology debate."

Indeed, the BI bottom line is all about enabling the visibility that will enhance top-line growth or other metrics critical to the business. Today, the technology is at the eye of a perfect storm, where business need, product availability and, for the most part, user readiness have come together all at once. The keys for riding out that storm include manageable projects, user champions and leadership so that cost, cultural issues and project-scope creep don't capsize important initiatives.

As Quaker's Tyler sums up the BI journey: "My advice is to have courage, keep it simple and remember that technology is the easiest part. It's understanding the business need and helping people understand what they're asking for that's tough."

Sarah L. Roberts-Witt is a freelance writer in Raleigh, N.C.

Dig Deeper on Small-business infrastructure and operations

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.