Developers Must Smash the Glass House

Computerworld - The term glass house connotes a data processing department holding corporate data hostage behind nearly impenetrable barriers. Anyone outside the glass house struggles to get access to key business data stored there.
Glass-house software developers often forget the reasons their organizations have invested so much in computers, data processing and IT infrastructure: to boost productivity so that the output of goods and services costs less, and to create new business opportunities that would be impossible to exploit without IT tools and capabilities.
New business opportunities are forcing change on organizations, and glass-house developers need to change too -- and fast -- before their companies are so hamstrung that they can't compete effectively. One of the most important of these new business initiatives, a focus on quality and service, has already begun to emerge in those organizations that have committed to making investments in IT and business integration.
Enterprises that integrate their IT capabilities and revamp their processes can be so responsive to changing business conditions -- such as the particulars of customer demand -- that they can provide products and services to order, customized the way buyers want and delivered precisely when they want.
A perfect example of this is the automobile industry's shift from a uniform assembly-line (mass-production-driven) approach to manufacturing cars to a special-order-delivered-to-your-front-door (service-driven) approach. Similarly, the most successful producers of PCs build to each customer's order, and they do it faster than ever: Making a PC used to take six weeks; now it's done in just 24 hours.
Becoming quality- and service-oriented puts pressure on an organization and its IT staff to ensure that data access is abundant and simple and that the data being delivered is consistent and accurate, regardless of where it originates. This can be a difficult challenge because, too often, key data has been trapped in layers of applications developed over the years, making access problematic. Just like in archaeology or geology, the data trapped in earlier layers is harder to get at than the newer data, which has been deposited on top.
To help business users meet the needs of today's connected, on-demand customers, developers will need to get at the trapped data and push it out beyond the glass house and into the hands of business users. They can achieve this by taking the following three-step approach:

• Align data with the business. In many organizations, information about customers, products, suppliers and employees has been identified differently, either literally or structurally, in various databases. Data redundancies and inaccuracies abound. Until the data tangle in your organization gets sorted out, your chances of successful integration are minimal. The decisions made about what to do with data -- what to keep, what to change, etc. -- have a context: the business.
  Aligning data with the business can be a complex process, but not an exclusively technical one. It begins with identifying the data, articulating the related operational rules necessary to support the business and then choosing a terminology that appropriately describes that data.
  Once alignment is complete, you can map the business terms and rules against the physical data sources, producing a logical view of the company's data that is represented by commonly used business terms. Nontechnical staff can then use these tools to operate at a higher level. This means that line-of-business analysts can use common process automation modeling tools to design processes that generate greater productivity out of the front office.


• Use metadata to find the data. Metadata -- that is, data about data -- is essential to your ability to unlock the lazy assets of your organization, your data. However, metadata is often ignored in IT strategies because it is not well understood. Metadata is data that helps both people and machines come to agreement on what is being analyzed or discussed. It represents the locations, structures and controls surrounding the data. When all this information is catalogued and made available as an aggregate set, it can be used to simplify the unlocking process.
  The metadata about your existing systems and the data itself provide the details needed to generate a map between your business and technological views of your organization. This metadata can be used to "link" the logical view to the physical data. These links then also become an important source of metadata.


• Break the stranglehold on data access. In unlocking your data, developers will increasingly rely on Web services to provide programmatic access to data using common Internet protocols. Why opt for Web services when it comes time to unlock, align and interact to achieve integration?
  Web services don't suffer from the high levels of complexity that other distributed computing technologies do, such as CORBA (Common Object Request Broker Architecture) or Microsoft's DCOM (Distributed Common Object Model). What's more, Web services share a much greater ubiquity as a distributed computing technology than either CORBA or DCOM.
  For instance, Web services can now be used from within Microsoft Office 2003, which allows a technically savvy front-office worker to leverage existing Web services to complete jobs faster and with less effort.