Last year I purchased a new bicycle so I could join a friend for a charity ride in Colorado. At the bike shop, they assured me that this model would be perfect for climbing steep mountain terrain. So with visions of flying over Colorado's mountain passes, I attempted a gravel slope near my home.

Standing on the pedals for power, I lost traction, the bike stalled and, with my feet firmly locked in the pedals, I crashed to the ground. Immediately, I returned to the bike shop and accused them of misleading me on the capabilities of the bike.

The manager gently asked me about my fitness and technique. Fitness? "Oh, I ride maybe a dozen times a year for about 30 minutes." Technique? "I've been riding bikes since I was a kid, so I mastered the technique long ago."

Well, not exactly. What I found out was that although I have a top-notch bike, the equipment alone will not make me a top-notch rider. To be successful, it's important to match my skill with that of the equipment by learning the techniques for the type of riding I'll do and by logging relevant saddle time.

The analogy holds true in data warehousing as well. I occasionally talk with new Teradata customers who do not understand why they are not achieving the same benefits realized by more experienced Teradata customers. They assume that if they have identical technology, it will make them equally as productive and successful as those industry-leading data warehousing users.

Again, not exactly. While technology is absolutely critical, the human factor is equally important. Getting training and talking to consultants can help you to prepare, but eventually you have to "get on the bike and ride." And you may even need to experience some falls along the way.

Here are some tips to consider if you are in the early stages of your data warehousing program:

	Learn new techniques. A mountain bike is not the same as a road bike, and the riding technique is different for each. Similarly, a good data warehousing platform will be fundamentally different from an online transaction processing platform.
	Architecture and design methods, data modeling approaches, and organizational best practices vary for each. If you are new to data warehousing, you will need time to understand, implement and become proficient in new techniques.	Tips to making the grade Climbing or riding a steep terrain can be quite a balancing act. These tips can help you gain momentum both in riding that steep grade and in your data warehousing implementation: > Anticipate the coming terrain. > Avoid a big burst of power. > Pace yourself. > It's OK to glance to the top of the hill, but always stay focused on the trail in front of you.
	Take training rides. Consider your first projects to be training courses. The scope for your initial projects should be determined by your team's fitness, your tolerance for missteps and your eventual goals.
	You will be learning about the technology and tools, working out new processes, and discovering and uncovering things you had not known about your data. But as data issues are resolved, once-unfamiliar processes become second nature, and as your technical skills are honed, you will be able to deliver greater business capability more effectively and at an increasingly faster pace.
	Don't over-train. Begin with projects that have a high probability of success. Only you can assess the abilities of your organization, but it never hurts to err on the safe side by keeping your first data warehousing project relatively simple.
	It is very likely that you will make some mistakes. That is a good thing, provided you learn something and have managed the exposure. Success breeds confidence, and confidence breeds more success.
	Establish long-term goals. Recognizing your first phases are training sessions for ensuring a strong learning curve, and a growing success rate does not negate the need for ambitious and aggressive long-term goals.
	I recently heard of an organization that switched into high gear after a six-month learning curve in which two projects were implemented on its Teradata system. In just four years, 300 applications have been implemented on the data warehouse, and sometimes as many as 30 projects are being developed in parallel.
	Your data warehousing goals, too, should reach well beyond the early learning phase or even the first couple of years.
	Review and monitor your progress. You can monitor data quality, application performance, user experience, the actual time to develop and deploy new capabilities, and sometimes the direct or indirect business benefit of a new capability.
	Using metrics and periodic progress reviews can help your organization improve, and it can also provide a clear view of successes and places where there is room to improve.

Data warehousing is not a sprint. Training, pace and commitment are critical to success.

And by the way, that gravel slope where I crashed on my first ride with my new bike is no longer a challenge. My fitness and technique have improved, allowing me to progress through increasingly more challenging rides with the confidence that I'll succeed. T

Planning your training rides What can you do to ensure the first couple of projects or phases are productive without being overly complex or ambitious? > Plan these early phases to be about three months in duration. With experience, you will get much more aggressive and be able to handle multiple projects in parallel. > Port an existing application. This can be simple or difficult depending on the application code. For example, moving the Data Dictionary Language (DDL) and data is extremely simple. Keep in mind that fully exploiting the technology and gaining the most value from the data warehouse will likely require eventual reengineering of your data model, your extract, transform and load processes, and your applications. > Control the number of data sources and sourced data elements. This may be as few as one to two data sources, or as many as 150. It depends on your source data and your experience working with that data. Data sourcing is often the most underestimated activity on a new project. > Focus on a single data subject area and application space to reduce complexity. But be sure to do these implementations in the context of longer-term architectural goals. > Target a small number of users. This might be five to 10 for the first phase and no more than 50 for the second phase. You will want to architect the applications so they can scale to the number of users you will need to support in the long run, but you will want to limit the exposure during these early learning phases. > Deliver business value. Even though it is good to keep the scope and complexity of the initial phases small, you always want to deliver something of value to the business. Improved performance is nice. Improving performance and adding capability is even better. —D.H.

Why Teradata Teradata technology has been used by hundreds of industry-leading companies to achieve their goals and to push the envelope of what can be accomplished with data warehousing. But most of these customers began using Teradata on relatively small, simple decision support projects. Still, they saw the benefits of having this best-of-breed solution from the outset because: > Teradata scales with you as your requirements and abilities grow. > Teradata's extensive functionality and capabilities enable you to take on the most complex data warehousing challenges. > Teradata's "purpose-built," integrated data warehousing platform makes it relatively easy to exploit those capabilities when you need them. But as with biking, you can't stop at acquiring the technology. Learning from and working with experienced professionals can help you reach your destination faster: > Teradata's professional services consultants can help you achieve your goals by coming alongside, helping you get the work done and coaching you as you implement your data warehouse. > As a Teradata customer you ride with the best companies that have been where you are. You can benefit from their experiences as you develop a network of colleagues and participate in regional Teradata User Groups and the annual Teradata PARTNERS User Group Conference & Expo. > Resources such as the Teradata Education Network and Teradata.com can also help you understand the best ways to exploit the technology as well as best practices for implementing your data warehouse. Find out more about all of these opportunities by visiting Teradata.com or by calling your Teradata representative. —D.H.

Photograph (1) by Phillip & Karen Smith/Getty, photograph (2) by Sean Justice/Getty.

Teradata Magazine-June 2007