RoadMap: Taming Data with a Data Mart
Data that is stored in your computer systems are valuable business assets. However, a piece of raw data, by itself, does not tell you anything. Until it is combined with other pieces of raw data and analyzed, its value is hidden.
Every organization has a wealth of raw data waiting to be analyzed to be used to show what your organization has accomplished and to help decide what your next steps should be. Human beings do not easily see the relationships among raw data points without the help of tool to help see what relationships exist.
If you search the Internet, looking for ways to manage your data, you will find lots of references to data warehouse solutions. A data warehouse is usually a much more complex undertaking than is necessary for a small business or non-profit. A better solution is a data mart. While a data warehouse tries to organize and analyze all of your data, a data mart uses a subset of your organizational data to create a data store oriented to a specific purpose.
A data mart is tactical and meets immediate data needs. It is smaller and less complex and is easier to create than a data warehouse. In addition, it is much less costly to develop and maintain. Potential users can be more clearly identified.
A data mart provides easy access to frequently needed data and allows a group of users to look at the same data to ensure that they all are 'on the same page'. This greatly improves response time, because reports can be created more quickly and the information can be given to end-users in a more timely manner.
What steps do you need to take to create a successful data mart?
1. Create a cross-functional project team.
A data mart team must be cross-functional. Neither the technologist nor the business subject matter expert (SME) can create the data mart alone. The SME understands what the data means to the organization...and this meaning is not always intuitive. The SME knows how and why a data field is being used for a certain piece of information. The SME is also acutely aware of what information is being requested by others within and outside the organization and what process is needed to get the correct data out of the system to create an accurate report. The technologist knows how to manipulate data, what types of data mining techniques can be employed, and can suggest the most cost effective data mart solution once he understands the true needs of the business.
2. Determine the purpose of your data mart.
Define why you are creating a data mart. Decide who will be using it and who will receive the output from the data mart.
3. Analyze your data.
Make sure that your data is as clean and accurate as possible. You know the old saying 'garbage in, garbage out'. If your data is not scrubbed to detect and correct inaccuracies, you will not be able to produce accurate reports. At the same time, you also need to analyze your data to ensure that it is complete and has internal integrity...for example, you don't want a client's age and birth date to conflict within the data mart.
4. Select appropriate technology.
It is important to ensure that the technology you select for your data mart works with the other technology that your organization uses. The cost of the technology you select and its ongoing maintenance cost is also a big consideration.
5. Use an iterative approach.
Don't try to do everything at once. Pick the low hanging fruit first. Experience success in creating and refining little pieces of the data mart, one at a time. This will keep everyone energized and working hard, and will confirm that you are on the right track and are creating a system that will bring value to the organization.
6. Get buy-in from users at all levels.
Show off what you have done to the people who will use the data mart and to the people who pay the bills. Have them confirm that the data mart meets their needs and that what you are doing is valuable to the organization as a whole.
7. Deploy and fine-tune.
Put the data mart into production, but expect a few small changes. No matter how many times the data mart users have looked at the system, they will realize that they need small changes once they begin doing their work using the data mart. If you have faithfully followed the above steps, you will find that these changes are truly minor and will make the data mart that much more successful.
Hopefully, once your data mart is complete, you will be able to say, as one of our clients did about her employee who is responsible for reportin..."now she can drive the reports instead of the reports driving her crazy." Good luck!!
You can view a SmartSite DataMart here. Although they are not incorporated into the website, this data mart also includes dynamic Excel reports that can be refreshed with new data every month. Dynamic Excel reports allow users to create reports month after month with just a touch of a button.
Let us know if you would like more information about creating a data mart for your organization.
|
