ARTICLE 3

Do You Trust Your Information?

Data pours in from many applications, but ensuring that it's accurate, timely and easily available to users is the key to business value.
By Bob Violino
October 23, 2008

Enterprises today face unprecedented volumes of information coming in from a multitude of sources. Applications such as enterprise resource planning (ERP), customer relationship management (CRM) and supply chain management (SCM) generate huge amounts of data. So do online applications, mobile devices, radio frequency identification readers and point-of-sale systems.

Much of the information organizations have is stored in departmental databases and data marts that can be accessed by users within those departments.

But just how trustworthy is the information that’s gathered? Can organizations ensure that the data is accurate, consistent across departments, and timely? If it is, information can be an invaluable resource that provides insight into customer demand and market trends, boosts productivity and helps enterprises compete better. If it isn’t, the information can actually hurt business performance and result in dissatisfied customers, frustrated business partners and unproductive employees.

Unfortunately, for a lot of organizations, much of the information available to business users is neither accurate nor timely, and is stored in “silos” rather than broadly available to users. “Most information in most companies fits that description,” says James Kobielus, senior analyst at Forrester Research.

Ideally, organizations would load most of their operational data into an all-encompassing enterprise data warehouse (EDW) that is accessible from a single business intelligence (BI) environment, and that BI environment would be available to all the company’s users.
But research by Forrester shows that most organizations have loaded only a fraction of their structured operational data (for example, ERP, CRM, financial, supply chain management) into an EDW, Kobielus says. Furthermore, he says, many companies use two or more BI environments, often in conjunction with diverse EDWs. And these BI environments are available to and used only by a minority of information workers — fewer than 10 percent of the employees at many companies.

“Consequently, most data is maintained in diverse silos: different line-of-business online transaction processing applications, different EDWs, different operational data stores, different data marts and so forth,” Kobielus says. “Even for the information that is maintained in EDWs — of which many companies have several, sometimes dozens — that data is usually batch updated overnight, and sometimes only weekly or monthly.” As a result, much of this data is not up to date.

“[Organizations] vary widely in the degree to which they have matched, merged, deduplicated, cleansed, corrected and standardized the data prior to loading it into EDWs or delivering to BI reports, dashboards and other applications,” Kobielus adds.

So most data remains in silos, not in a single authoritative EDW. Most is not cleansed consistently with corporate-standard, data-quality tools. And the data is not kept fresh through real-time propagation to the BI environment from source applications, Kobielus says.
In this kind of environment, one business user might have information about a customer that differs from what another user has. How can enterprises ensure that users are getting the most accurate, timely information that will actually be of use to them, regardless of where they’re working?

The trustworthiness of data depends in large part on having a single, authoritative EDW and cleansing data prior to loading it into the data warehouse. The latest, freshest data should be loaded directly from systems of record, Kobielus says.

“Ensuring this trustworthiness of all data sets demands that both IT people [for example, database administrators, line-of-business application administrators], and data owners [businesspeople responsible for managing the accuracy of all data sets] cooperate in an administrative workflow known as ‘data governance,’” Kobielus says.

A big part of this workflow includes a collaborative designation of master source-data sets for each data domain, and a definition of policies and rules for merging, matching, standardizing and correcting data prior to loading it into the EDW and delivering it to BI applications.

If all of this sounds like a lot of work, consider the consequences of not providing accurate information: business failure or bad decisions that keep an organization from realizing its potential, Kobielus says. “There’s a reason why the EDW is often referred to as the repository for a ‘single version of the truth,’” he says. “If the EDW holds trustworthy data — as determined by the cleansing policies/rules defined under data-governance workflows — then businesspeople can have confidence in the reports, queries, dashboards and analytics that incorporate the data.”

Comment

This article are present about the strategies initiatives that organization undertaken to help it gain competitive advantages and business efficiencies. The organization information is stored in a database. Database is maintaining information about various types of objects, event, people and place. If the information stored in a database, the information will safe. Information is an organization asset. Databases offer many security features including password, access levels and access control. Databases can increases personal security as well as information security. Then businesspeople can have confidence in the reports, queries, dashboards and analytics that incorporate the data