Have you traveled by air recently? Maybe for a vacation? Or maybe you are a frequent traveler like me. Sitting in Chicago's O'Hare International Airport during yet another layover I started to think about the baggage handlers, maintenance crews, gate agents, flight attendants, pilots, air traffic controllers and many others in the airline industry who, along with myriad complex systems and technologies, provide an infrastructure that makes air travel possible.

The ultimate purpose for all of these people and systems is to deliver you and me to our destination safely, in relative comfort and, we hope, on time. Their purpose is to serve the air travelers. They may be the best at what they do, but there is no reason for them to do it if travelers don't fly.

In a similar fashion, IT organizations, both large and small, exist to serve the needs of the business. Whether we are database administrators (DBAs); system administrators; data modelers; extract, transform and load (ETL) experts; systems architects; procurement managers; or CIOs, our purpose is to provide business organizations with the capabilities, services and information they need to fulfill the primary mission of their business. Without business users there is no reason for IT to do what it does.

Eliminating disconnects
Departments within an organization must work together to achieve success. Most people would consider this conventional wisdom. Yet there are many examples of disconnects between business and IT.

A few years ago, I had a person on my staff who had been the chief architect on an integrated, multi-application user interface utilized by hundreds of customer service representatives. During the three years she spent designing this system she had never spoken with an end user. When we later asked her to spend a month with these representatives, she had a very enlightening experience. Although the system was quite sophisticated with respect to the underlying technology components, it was far more difficult to use than she had imagined.

I also remember meeting with the IT directors of a 200-person information services team (data warehousing and decision support) a few years ago. They were angry and frustrated that, in spite of all of IT's effort and the dozens of data marts they had provided, the marketing organization was looking outside of the company for help since it did not feel IT was meeting its needs. There was certainly a disconnect between what the business side believed they needed and what IT was delivering. These disconnects frustrate both IT and business users. They waste time and money and can put a strain on a company's ability to reach its goals.

Making the connection
To avoid disconnects, IT and business must work together to develop a shared, business-oriented vision for the data warehouse. This requires executive sponsorship and participation by all of the stakeholders. It does not have to be perfect and it will need to change, but the process will help get everyone on the same page. Without this shared, strategic vision, any self-serving tactical needs from throughout the organization will prevent companies from moving their decision-support capabilities to new levels.

There are a number of important processes and activities required to design and build a data warehouse. Yet, after all of the work to configure and set up the technology platform, and to model, source and load data into the data warehouse, how much business value has been delivered? None. Consider "Data warehousing value hierarchy," below. As with airlines and travelers, business value is not realized until business organizations use the data warehouse to answer questions, perform analysis, influence decisions and act on those decisions.

Furthermore, both IT and business should be involved in the design and preparation of the data warehouse. It is in the execution of these processes that both sides gain an understanding of the challenges and constraints that each faces. When IT and business work together, there is shared ownership—shared responsibility for unproductive decisions and shared credit for the successes.

Going the distance
Airlines have metrics around safety, on-time arrivals, lost luggage and other factors. Business and IT should work together to establish business-oriented metrics of progress and success. Metrics of return on investment (ROI) are valuable, but often difficult to define and measure. As a proxy for business value, companies might also consider less direct metrics such as increased end-user activity, new business questions, new applications and improved data quality and timeliness.

As always, metrics should be kept in balance and prioritized. What gets measured gets done. That can be a double-edged sword. For example, data quality metrics can be very useful when assessing progress. On the other hand, service level agreements (SLAs), which demand high availability but are not balanced against other SLAs and measures, are easy to meet if IT severely limits changes to the system.

For those of you in IT, how about spending a couple of months sitting alongside the end users? Once in a while, embed yourself into the business side of your enterprise. Help your end users answer their business questions, develop reports or perform analyses. Observe their needs, their priorities, their use of the data and tools. What is being demanded of them by their management? What schedules and milestones drive them? What frustrates them about the systems they use? What do they appreciate about those systems? Are they doing things the way you envisioned? Could they do their jobs easier or better? Can you help make that happen?

Spend time with the users. You may see the data and the capabilities you are providing in a new light. After all, the users are your clients. Your mission is to serve them and help them be successful. T

Data warehousing value hierarchy It is important for organizations to understand and consider how value is derived from a data warehouse. To build a successful data warehouse, first you need to provide a robust, scalable, flexible, high-performance platform. Then you must model, extract, transform, cleanse and load data onto that platform. Yet no business value is realized until business users are able to use the data warehouse to get answers to questions, to support decisions and business processes, and to build applications. The bottom line is that data means little until it's put to use. It is obvious, therefore, that IT must deliver a usable product in order for business users to access and analyze the information they need to make intelligent business decisions.

Teradata Magazine-March 2007