Almost everything we do generates date including surfing the web; buying groceries from the supermarket; and sending text messages. In fact, with the right mobile technology, simply walking into our local malls generates data.
With this explosion of data, developing a top-notch data warehousing system is paramount to the success of any company. Processing data in a well-developed data warehousing system provides the competitive edge that companies are striving for. The question then becomes: What steps should you take to ensure you build a data warehouse that enhances and supports the decision-making process?
Mike Ferguson makes the point that you and your data warehouse development team have to understand certain things including:
- The corporate business strategy: For your decision makers to make the best use of the data, your data warehouse development team must understand the corporate business strategy. Once they do, they can work with the decision makers to determine which objectives are priorities. They can also figure out how the data can lead to a higher return on investment. They must continue to work with the decision makers to join the data elements to the objectives then determine how to capture the appropriate data and build the necessary dependencies to make the data meaningful.
- The data requirements: Your development team must work with the corporate data scientists and analysts to define data requirements, data sets and desired data visualizations to ensure that the data warehouse is highly interactive and that it allows users to customize the data sets, charts, and graphs. And they must make that information available in a variety of formats including dashboards, scorecards, and reports.
- The technical environment: You must learn as much as you possibly can about the proposed technology and use that knowledge to draft the data warehouse technical architecture. Then you should participate in all facets of the technology selection process and work with the team to develop an implementation plan, which should include:
- A metadata repository – to track information about both the data and the system including processes. Define the business vocabulary, store it within the repository and share it across the organization.
- An ETL process – to extract data from transaction systems, transform the data into something suitable for the data warehouse and then load the data into the target system; i.e. the data warehouse. Pay close attention to how the data is manipulated, how long the process takes to completely execute, and the accuracy of the data. Be prepared for data that may be incomplete, incomprehensible, and inconsistent and develop processes to handle the issues.
- Security – what level of security will be required? How will the data warehousing system be maintained to the level required by the organization, internal compliance, and external laws and regulations?
- Data integration – combine data from several disparate sources into a unified data warehousing system. Then attempt to embed the system within existing corporate software for quicker adoption as the user may feel familiarity with the look and feel, leading to less end-user training.
Next Steps:
- Subscribe to our blog to stay up to date on the insights and trends in big data, data analytics and data warehousing.
- To learn more about how analytics can improve your business and increase your bottom line check out these complimentary guides:
- “5-Minute Guide to Business Analytics,” to find out how user-driven “analytic” or “data discovery” technologies help business and technology users more quickly uncover insights and speed action.
- “5-Minute Guide to HR Analytics,” to discover the three critical capabilities a modern analytic environment must provide to the entire spectrum of HR staff so they can adequately support the enterprise.
- “5-Minute Guide to CRM Analytics,” to learn how agile analytics technology can help you deliver critical value to executives and front-line marketers.
Dennis Hardy
Spotfire Blogging Team
