• Current Issue • Ad. Rates • Writer's Guidelines • Archives • Circulation

Minnesota Technology Magazine - Fall/Winter 2006

Data Dimensions

Suffering from information frustration? A range of new technologies promise to help firms get the right data at the right time to the right people in your company.

BY MARY LAHR SCHIER
Mary Lahr Schier is a Northfield-based writer.

The Digital File Cabinet

Can "tiered storage" help solve data overload?

More data can make for better business decisions, but most data—up to 70 to 90 percent—is inactive most of the time. As data has grown, so has the need for better, more efficient storage of it.

For most IT professionals, information lifecycle management (ILM) means tiered storage-the assignment of data to different media depending on its use pattern and value to the organization. Tiered storage can save money by putting less-frequently used or less-valuable data on cheaper storage, such as a SATA device rather than a high-performance fiber channel. However, implementing tiered storage can require that IT departments make tricky, sometimes time-consuming decisions about where to place the data, as well as labor-intensive work making the shift.

A new product developed by Minneapolis-based Compellent Technologies Inc. takes the politics and a lot of the work out of storage decisions, says Bruce Kornfeld, the company's vice president of marketing. Compellent's software is one of several relatively new products that provide automated data migration. However, it is the only product that automatically moves data around at the block level, and moves it based not on an assigned notion of value but on how frequently it is used.

For instance, if you create a 50KB document, that document contains not one block of data, but 50,000. Compellent's Dynamic Block Architecture^TM tracks the usage of each block of data and moves it to better storage the more it is used. So, if the accounting department is doing the end-of-the-quarter books and needs to look at old purchase orders, those orders would automatically move to a higher, faster level of storage. After the quarterly reports are done and the p.o.s are idle, they move back to lower storage. "It's completely invisible to the end user," says Kornfeld, who explains that because more data can be kept on lower levels of storage, companies reduce their capital expenditures for storage-by up to 74 percent. Briggs & Morgan, a St. Paul law firm, found that two-thirds of its data—mostly e-mails and legal documents—could be moved to lower-level storage after implementing the Compellent system.

The system is being marketed primarily to mid-sized enterprise companies, particularly those with revenues of $25 million up to $1 billion, says Kornfeld. "The reason it has been especially attractive to companies with 50 to 200 employees is they don't have huge resources and don't have time to be moving data around all the time," he notes.
-M.L.S.

Face it: You're drowning in data. E-mails, Word documents, PowerPoint presentations. Reports about your finances, your inventory, your workforce, your competitors. Relational databases, production statistics, sales prospects—it all comes at you fast. Sometimes, you know you have the information, but can't find it. Sometimes you create more information, then discover someone else in your organization already had the same data at hand. Meanwhile, the information continues to pile up and you buy more storage to hold it all.

You're not alone either. A recent University of California at Berkeley report estimated that the amount of digital and print information produced each year will be eight times higher in 2009 than it was in 2001. By 2009, the amount of information created also will be enough to fill the equivalent of 300,000 Libraries of Congress. Not surprisingly, the ability to manage information from the moment it is created until it's archived or destroyed is increasingly a factor in business success. It's also a growing challenge. Only about 20 percent of the information business creates is in database format—that is, a traditional row-and-column listing that can be sliced and diced easily. The rest is what's called "unstructured" content—all of the e-mails, Web pages, digital images, and documents that you and your employees create every day. No wonder producers of content spend about 60 percent of their time locating, formatting, and structuring content they already have, according to Zapthink, an information technology research firm. To get better control of information and reduce the costs of storing it, companies are turning to a variety of technologies and processes called information lifecycle management or ILM.

For IT professionals, ILM requires a review of the company's storage needs and its ability to access data and migrate older, less valuable data to cheaper storage while keeping current and important data available. Several advances in storage technology have made it easier for a company to have the data it needs easily at hand without spending a fortune on storage. (See "Dashboard Basics," on page 24.) Yet, ILM is more than an IT department issue, and often combines corporate policies and procedures with technology to improve business decision making. "Information lifecycle management is really about helping organizations figure out what kinds of data they need to make meaningful decisions for their business," says Dan Grady, CEO of St. Paul-based Systems Consulting Group (SCG), which has provided management and technology consulting services to some of the Twin Cities' largest companies, including 3M, Best Buy, and Tennant Co. "The data changes over time, and so does where it is located."

Dan Grady and Charles Hanson, Systems Consulting Group, Inc.

PUTTING INFORMATION TO WORK

When Charles Hanson, SCG's manager of Microsoft consulting, talks with clients about information lifecycle management, he tells them to think of their business as a big, threetiered cake. "The bottom layer is the data that a company has," says Hanson. Taking a long look at the bottom layer of the data cake is the best way to determine what information you have to manage and how valuable it is to your business. Hanson asks company leaders and IT managers to think about such questions as:

What does the data look like?
Where is it stored?
How often is it accessed?
How does the business use the data?
What's its value? How can it be classified?
What are the rules and processes by which data is classified?
Where does data end up?
Is it archived? Destroyed?

Newer storage tools, such as automated data management, which moves data from one class of storage to another automatically based on company policies, and hierarchical storage management, which moves data into archival storage based on usage, have given IT professionals greater ability to manage storage economically while keeping data easily accessible.

The second layer of the cake is the business intelligence buried in "that big mess of data we have," says Hanson. This layer essentially marks the intersection between physical storage of data and its display. From this layer, companies determine key performance indicators and create the rules that determine how data is displayed or stored and who gets to see it.

The top layer of the cake represents the point at which companies "convert data from geek-speak to business language," says Hanson. At this top level, companies use their data to determine strategy and manage operations essentially in real time. Many times this involves creating portals for various aspects of the business such as human resources, partners or vendors, contract management, or Sarbanes-Oxley compliance. For example, members of your sales force may have a portal to use to enter data about potential customers. Back at headquarters, executives can use that data to tweak marketing strategies or determine how likely the company is to meet performance goals. The flow of information occurs continually and in real time rather than in periodic "strobe light" spurts, says Hanson. "Information is flowing between the layers all the time," he notes. "Companies are seeing a lot of organizational efficiencies when they streamline communications."

To ensure the information is flowing well in this top layer, more companies are turning to content-management systems. Properly done, content management can create "a single version of the truth," says Andy McMillan, vice president for product development of Stellent Inc., an Eden Prairie-based firm that provides content-management products to Fortune 1000 companies. Recently acquired by database giant Oracle Corp. of Palo Alto, Calif., Stellent's products use a Web-based architecture to store, categorize, and control access to business content. Stellent's system involves a repository for storing data and a variety of interfaces that control access to information and promote communication and collaboration. Stellent's Universal Content Manager converts documents to formats that can be viewed even without the original application. "This kind of system avoids a lot of repetition of formats and applications," says McMillan. "It means that not everyone has to have every application in which documents may be created."

Content management programs can control who can change or view documents. They also can prevent repetition of effort and errors, says McMillan. For example, if a product's price changes, content-management systems will take the old list off-line immediately, guaranteeing that no customers are given the wrong price. Collaboration is also easier. The Stellent system offers the options of creating dedicated Web space for collaboration team members. Information is stored and manipulated there, rather than in a string of e-mails which may be difficult to decipher later.

John Woodburn, The Woodburn Group

DASHBOARD DECISION-MAKING

The potential to make decision making more data driven is one reason many executives are turning to information-management displays—or dashboards—to keep track of key performance indicators. A 2004 survey by CFO.com found that about one-third of executives were using performance dashboards, and nearly 40 percent were considering them. Dashboards provide a more visual way to see how your company is performing. Displays are often set up as gauges or charts to indicate current conditions. Many systems send alerts to appropriate managers when an indicator is out of line.

Dashboards and the business intelligence systems underneath them allow executives to monitor key metrics in real time. "If a metric gets out of an acceptable range, the executive can drill down and analyze why," says John Woodburn, founder of the St. Louis Park-based Woodburn Group, a business intelligence consulting firm.

While dashboards, content management, and other ILM tools traditionally have been used in larger companies, Woodburn and Grady agree that more small and mid-sized firms are seeing value in these systems today. "They have the same needs as the big companies," says Grady. "They just don't have the resources."

Companies that are considering implementing dashboards or other business intelligence tools should begin with three questions, says Woodburn. First, what is the quality of your data? Do you capture the right kinds of data about your business and is it in a form that's available? Second, do you have the infrastructure to implement a business intelligence system? Finally, does the business operate off of metrics? If so, dashboards help executives and managers identify potential problems more quickly.

"As people do more and more work in a digital format, there tends to be information overload," says McMillan, who adds that with ILM tools, companies can move from digging through information to managing knowledge.

Dashboard Basics

How to get the most out of a new tool.

Considering using a performance dashboard to manage your company? Matt Christensen of Minneapolis-based PRAGMATEK Consulting Group recently offered readers of DM Review, an industry journal, some tips on getting the most from a dashboard.

Stay aligned with strategy. "Any dashboard should echo the strategic direction and priorities of the organization," says Christensen. Dashboard projects should start with a reexamination of the company's strategy, goals, and mission statement.
Define and choose metrics carefully. Make sure those mining and viewing data understand what is being measured. Try to avoid functional bias. That is, the director of finance should see more than just the numbers and historical data. Information on customer satisfaction, production, and sales are also useful. However, if dashboards are going to be applied throughout an organization, do not clutter them with irrelevant data.
Focus on leading indicators. While historical data is useful and easy to gather, Christensen recommends that metrics be developed and displayed in ways that focus on leading indicators. For instance, a manufacturing company changed the metrics displayed to its sales force from "sales year to date" to "daily sales required to achieve target."
Ease into using an executive dashboard. It may take several phases to implement a dashboard project. Start slowly and spend time understanding what the graphical interfaces mean and how to get the most from them.

-M.L.S.