THE DATA ADMINISTRATION NEWSLETTER – TDAN.com
ROBERT S. SEINER – PUBLISHER

When filling a position, simply creating a job title, or even an entire department, does not in itself provide business value. The true value proposition for the enterprise is achieved only after securing the right person in the right role. This is never more valid than when your organization is selecting an enterprise information architect, as the responsibilities of this position go well beyond mere knowledge of data and information, making the role notoriously difficult to fill. The enterprise information architect requires superb people skills, as well as a comprehensive understanding of the industry and the ways in which the business generates revenue.

From a high-level perspective, there are three key objectives that confront enterprise information architecture. These key objectives are: a) curing business pain, b) equipping business to compete, and c) adopting and instilling practices that promote the long term health and profitability of the enterprise.

Curing Business Pain

First and foremost, the strategy proposed by the enterprise information architect must address pain being felt within the business. But in order for recommendations to be credible to business stakeholders, the information architect must be able to articulate an in-depth understanding of the root cause and remedy in business terms.

Furthermore, no matter the industry, it is difficult to imagine any area of pain that does not involve data, whether that information involves manual procedures or ones that are automated.

For example, the business problem of being outpriced in the marketplace is often a result of high infrastructure costs that are, of course, passed on to the customer in price and to the shareholder in dividends and share price.

Infrastructural inefficiencies and waste manifest themselves in many forms, and a surprising number are in purview of the enterprise information architect. The usual suspects include a proliferation of databases, development teams and brands of products, and a computing infrastructure that must support a long list of application silos.

As a rule of thumb, if an enterprise has more databases than cubicles, then it is likely that the enterprise is overly burdened by the operation’s total cost of ownership (TCO).

But the actual TCO does not stop with many non-integrated operational applications, as downstream costs of business intelligence and data warehousing correspondingly advance. As each new island of automation is born, the need to integrate it into various consolidated views of data for the business emerges, each accomplished as a quick and dirty set of patchwork. At first, the costs of this approach are quite low; but as the complexity mounts, so do costs.

The silo approach is initially so attractive because it requires the least amount of planned architecture. The development of each consolidation must occur for each new silo and for each back-end analysis database. Once developed, the operational data from each silo must be consolidated into each analysis database as often as up-to-date information is required.

In some situations, the pain being felt by the business has existed so long that business stakeholders have become desensitized to it.

For example, business users record information and pass it to IT for processing. Then, when the business attempts to consume the data that has been processed, the condition of the data can be such that it is no longer understandable in business terms. The most obvious way to detect when data is no longer understandable to the business is when there is a need for a middleman, called gray people, to help business users interpret the data. Sometimes this happens with general ledger reports or other basic business intelligence reporting.

Even though IT may have unknowingly contributed to data quality problems over the course of time, interestingly IT does not possess the ability to correct data quality problems.

An IT organization can take a number of measures to prevent the quality issues from worsening, but the ability to remedy the existing damage requires real partnership with the business.

In a 2007 Gartner press release, “Dirty Data” is a Business Problem, Not an IT Problem, Gartner states, “Over the next two years, more than 25 percent of critical data in Fortune 1000 companies will continue to be flawed, that is, the information will be inaccurate, incomplete or duplicated…” As we can see, however, the problem is shared rather equally between IT and the business.

To approach the challenge of addressing pain, enterprise information architecture must effectively utilize everything it knows about the business, industry, organizational structures and people so that it may help suggest the appropriate incentives within the enterprise to reduce an expanding IT infrastructure. Until a reasonable set of steps is formulated to reward senior and middle management to embrace the desired behavior toward right-sizing the data and resulting landscape, the pain can only worsen over time, as market share in sales and revenue fail to reach their full potential or even decline.

Therefore, in order to successfully confront business pain, any cogent plan must clearly articulate the problem in business terms to business stakeholders so that they may properly appreciate the impact and the alternatives for treating the pain versus implementing a complete remedy.

Equipping Business to Compete

The ability to equip the business with a competitive arsenal is the second key objective of enterprise information architecture. First and foremost, however, any competitive arsenal that is under consideration must also be an arsenal that the business can realistically afford and make good use of.

To begin with, enterprise information architecture should identify the data and information, including metadata, that is most pertinent to a well-defined business process as well as supporting IT processes. Any combination of these two can represent a valuable competitive advantage to cultivate.

An example of just one category of metadata is metrics, which can be collected from an automated workflow providing a powerful view of what is happening across the enterprise.

Efficient mechanisms for collecting metrics are like having radar and sonar that operate within a business to quickly illustrate what is happening within and around your enterprise.

More specifically, metrics can be used to reveal immediately how many transactions of each type are currently being processed across the enterprise, as compared with a prior period. Metrics are good at pointing out where an overall process is bogged down and delayed, and metrics also highlight the efficiency with which departments, teams or individuals support their activities.

Whether the process is in the business or deep within IT, only the availability of metrics can lead to metrics-driven management efficiencies.

Equipping the business with a competitive arsenal requires a good amount of business knowledge to determine which data, information and metadata will provide the greatest benefit. After all, there are few businesses that can afford everything available. Being armed with only a small amount of business knowledge is sometimes all that is needed to convey the potential value of metrics from across the enterprise to business stakeholders. The more business stakeholders understand the latest tools that are available, the more likely any gap in business knowledge on the part of enterprise information architecture may be overcome.

It should be noted that by no means are metrics the only valuable form of information that can empower business stakeholders. There are other valuable forms of data and metadata that can improve the quality and accessibility of business data and information, such as understanding the sparseness of certain data values, inconsistencies in data value format, or data values that are properly associated with one another in a database. These areas of metadata can create competitive advantages in leaps and bounds by making the data more usable to business users.

Ultimately, if the business does not benefit, or if the business is not empowered by IT, then it is important to address your priorities. The business should be able to audit and improve the quality of data that supports their business processes. The first challenge is, however, dealing with redundant data that is sometimes unfathomable for business stakeholders to muster the resources to tackle. When this occurs, it is a clear sign that IT has grown without the benefits of enterprise information architecture and that the business either knowingly or unknowingly has a competitive drag.

For example, continually fixing the data at the back end is like a production line that turns out defective cars, and then adds a repair shop onto the end of the assembly line to fix the problem.

With respect to additional areas of data, information and metadata, there are numerous areas of business data, such as those that are particularly useful for recognizing changing business conditions.

A core competitive advantage for any business is empowering business with information essential to help them properly adapt to changing business conditions.

The more agile even the largest enterprise can become, the better it will be able to maneuver to meet competitive and economic challenges. Even so, it is surprising how many businesses choose to operate in the open market without any form of radar or sonar.

The enterprise that optimizes its operational efficiencies with metrics is buying valuable insurance to know when and where it must adapt its course and activities to address the competition.

Adopting Practices for Long-Term Health

Adopting practices that promote the long-term health and profitability of the enterprise is the third key objective of enterprise information architecture.

Although this enterprise information architecture objective may involve adoption of a number of best practices across IT involving the use of business models, metadata management, data quality, business intelligence and data warehousing, a healthy regimen for long-term health is one that prescribes a wholesome dose of education and quality communication.

Simply put, the more enterprise information architecture understands the vocabulary and value of business information and the more business stakeholders understand the value of what information architecture can do for them, then the healthier and more profitable the enterprise can become.

Long-term value is derived from business architecture and the resulting business blueprint that serves as documentation to communicate the data and information used by the business to IT. The more IT is kept in the dark, the less useful your IT organization will become, as communication is essential.

It is going well when business stakeholders see their issues addressed in the architectures of the enterprise.

The challenge of enterprise information architecture is similar for large, mid-sized and small organizations. Data, information and metadata are the life-blood of good decision making, and enterprise information architecture is at the heart of it.