Sifting Through Stored Data

Almost every company will tell you that its most valuable asset is its people. However, the value of a company's mission-critical data runs a close second. Any company can generate terabytes of data, but few can successfully manage that massive amount of information.

Customers must go to great lengths to simply store data — let alone manage it. They must sort through a multitude of storage options; each option has its own performance and cost benefits. Once the system is in place, data starts flooding in from employees and applications. Now, what to do with all this information?

The data that is generated and stored has a high value for companies. However, some companies can't take advantage of what they have stored in their systems. It's the proverbial needle in a haystack. A company knows a critical piece of data is stored in its system, but it can't access that data in a timely manner. Or, the data can't be presented in a usable format.

Companies can easily identify with this problem. But, FileTek was founded to solve this problem. FileTek, headquartered in Rockville, MD, is a storage integrator that has developed its own line of storage management software for Global 2000 companies. "Businesses have been good at storing data. They have been alright at managing data. They have not been very successful at making data available to business processes, applications, and decision makers," states Robert Simonds, vice president of marketing at FileTek.

Making data available in a usable form is the driving force behind the solutions created by the 130-employee FileTek. The integrator develops systems that are a key part of a total data warehousing solution. These data warehouses contain data from throughout a company and present decision makers with a clear picture of their business at one point in time. This technology is especially important to companies that have millions of customer contacts each day, such as retailers and telephone service providers.

Data Warehousing Gets The Most From Data
The ultimate goal for FileTek is to be one part of a complete data warehousing solution. But, data warehousing is composed of several technologies. One component is a relational database, which allows users to extract all related data stored in a system. Another data warehousing technology is a query generator, which lets users search databases for answers to particular questions. The information may then be presented in a report format, for example. Data mining is another important aspect of a data warehousing solution. With data mining tools, companies can extract information from their databases that may indicate sales trends or customer habits. Armed with this information, companies can be more proactive in their business planning.

"There are a couple of deficiencies with most data warehousing environments," says Simonds. One of those shortcomings is the total amount of information that companies can store. A retailer, for instance, may only store 65 weeks' worth of data instead of three years' worth. Or, a phone company may only store calls from the past two months — if at all. "Companies are forced to compromise," relays Simonds. "They truncate history and they aggregate their data into summaries. Rather than storing all the details in their databases, companies summarize their data and then store that information."

FileTek's software, StoreHouse/SM and StoreHouse/RM, allows companies to store and manage their data at minute levels. Thus, companies don't have to create summaries of their data that result in less detail. "We allow companies to manage atomic-level data, which are all the base transactions that take place. Our customers then move this atomic data from our system to a data mining tool, for example," says Simonds.

Solving A Horizontal Problem For Vertical Markets
Access to critical information is a horizontal problem, but there are several vertical markets where companies use atomic data to drive their business processes. This is the case with telephone companies, retailers, and financial institutions. Typically, these industries log millions of customer contacts on a daily basis. "Manufacturers of big-ticket items typically don't generate a large volume of atomic data through customer contacts," comments Simonds. "But, they may have large volumes of structured and unstructured data, such as parts and inventory information, in other areas of their businesses."

A telephone service provider is a prime example of a company that needs access to atomic-level data to run its business. FileTek will create a storage system for a phone company to store records of every customer call. At a minimum, that record is used in the company's billing process. "A mid-level phone company may have records for 50 million calls per day. A major long-distance company probably has records for 300 million calls per day," states Simonds. "In most cases, those records are stored on tape for at least 72 months."

With the records stored on tape, a phone company can respond to an audit request, or the FBI may need phone records for an investigation. Also, records need to be accessed to settle customer disputes over billing. "In each of these cases, the data is stored for defensive purposes. The data only allow a phone company to react to a situation," says David Beamer, president and COO of FileTek.

Help Clients Be Proactive
Most companies strive to be proactive and not reactive. Data warehousing technology allows them to accomplish this goal. In the case of a telephone service provider, the company can use stored data to market to its customers more effectively. Instead of simply storing customer call records, a phone company can use the data to create a profile of its customers. This profile can be used to understand customer behavior and determine when some customers may change providers.

Retailers who use data warehousing are also trying to get a complete picture of their customers. Retailers want to perform what Simonds calls "market-basket analysis." This allows each customer transaction at a store to be analyzed. For instance, a retailer can analyze sales data to determine that a particular store sold 200 cases of Pepsi. "Having instant access to this type of sales information was groundbreaking technology two to three years ago," says Simonds. "Now, retailers want to know more details about each customer transaction." For example, retailers want to know if placing Pepsi on sale attracted a certain type of customer, and what types of products customers bought in addition to the Pepsi.

"There are sophisticated data mining tools that allow companies to perform this type of analysis," says Simonds. "We allow companies to store this atomic data and integrate other vendors' products to create a data warehousing solution. The atomic data we store and manage is the fuel for data warehousing tools."

Data Can Shape A Company
Prospective FileTek customers have no problem recognizing that they need a solution to more effectively manage their data. However, sticking with the status quo is sometimes the stiffest competition that FileTek faces for an account. "I don't know how companies survive without data warehousing technology," says Simonds. "I don't understand why companies go through the trouble of collecting data through their operational systems and then purging it. Companies that don't invest in technology have to make compromises."

The goal of data warehousing technology is to put detailed data in the hands of decision makers in a timely manner. If this goal is met, then the data can be used to shape every company decision and process. It's what Simonds refers to as "zero latency." "Zero latency means that an executive can't wait for IT employees to write a program to extract data from a company database," explains Simonds. "Data warehousing technologies are a part of this zero-latency equation. Our storage management solutions are factors in the same equation."