Understanding Quality Management For Data Warehouses
Quality is an important concept when it comes to data warehouses, as well as their environment. Quality should not be defined in terms of data, even though having quality data is important. When I talk about quality in this article, I’m talking about the big picture.
I’am referring to the success rate of the data warehouse in conjunction with its ability to help the company achieve its goals. In addition to this, it is also important for companies to learn when quality needs to be emphasized before the actual data warehouse is built. A company that wants to succeed must measure what they’ve already done, along with making the necessary adjustments for the actual measurement of the data.
Quality can simply be defined as reaching the expectations of your customers without going above them. The reason for this is because getting higher levels of quality is costly, and even if you surpass the expectations of your customers, there is no guarantee that you will have a higher rate of return. Some business executive may want to know why taking measurements is so important. If a company doesn’t take measurements, everything they perceive will be highly subjective. In other words, a company won’t know if they are continuing to improve over time. This is why data warehouses are referred to as a process rather than just a technology or a product.
Because data warehouses are a process, it is process based measurments that should be used. Companies will want to measure things such as "activities" and "lengths." Measuring a process is much different than measuring a product, and a data warehouse much be approached from a process oriented perspective. With a product measure, you will measure things such as the volume of the data you have. While getting quality in your data warehouse will not be free, the costs will be much lower than having a data warehouse with poor quality. The costs that you will have to pay for quality will come in the form of re-planning, implementation, and measurments.
It could be argued that re-planning is the most important factor in data warehouse quality. Once the problems of today are solved, and company must be prepared to deal with the problems that will occur tomorrow. It is also important to analyze the value of data warehousing from the business perspective. For business people, the purpose of using a data warehouse is clear: to gain a powerful insight into decisions they can make to help their company become more productive. Based on this, the true measurement of a data warehouse is whether or not the data warehouse can help the business succeed. Over time, the upper management in the company must be able to see progress. If they cannot, the data warehouse project will be considered a failure.
Many companies make the mistake of believing that a data warehouse is silver bullet. They think that by simply using the most cutting edge technology, they will automatically be given an edge in the marketplace. It is attitudes that like that often cause data warehouse projects to become failures. A data warehouse is not one technology. It is multiple technologies combined, and once a company purchases it, it will need to be customized. Most importantly, the company will need to establish guidelines for operating the data warehouse if they wish to run the program efficiently.
It is also important to realize that data warehouses are tools that must evolve. This is precisely why they are often built in an incremental format. Some experts feel that data warehouses is a process of evolution, and they also feel that companies need large scale projects that can be built in three months rather than three years.
Companies that want to produce quality management for their data warehouses must know what they have done right, as well as what they have done wrong. This is where metadata becomes so useful. Metadata can play an important role in the measurement and quality of your data warehouse.
There are three types of success that companies must aim for, and this is political, economic, and technical success. When the data warehouse increases the bottom line, a company has succeeded economically. When the company is using the data warehouse daily, it has succeeded politically. When the right tools have been chosen for the right tasks, the company has succeeded technologically.