Time marches on and soon the collective retirement of the Kimball Group will be upon us. At the end of 2015 we will all retire. In my final Design Tip, I would like to share the perspective for DW/BI success I’ve gained from my 26 years in the data warehouse/business intelligence industry. By Bob Becker – © Kimball Group. All rights reserved.
While data warehousing has been around now for a long while, there continue to be significant opportunities for organizations to leverage information for more effective decision-making. New technologies consistently emerge that create and allow us to gather increasingly more granular data from the world that surrounds us. The data and the analytics surrounding it are as interesting and important as ever – perhaps more so. New and improved DW/BI solutions will continue to be deployed. There will be no future shortage of opportunities to apply the lessons learned about creating a successful DW/BI environment.
By now, we all recognize that data warehousing is a mature industry. We are long past the pioneering stage of DW/BI evolution. We are working with proven, capable hardware and software technologies with new and more capable technologies evolving all the time. As an industry, we have been implementing data warehouse and business intelligence capabilities for a long time – decades in fact. Lessons have been learned, technologies improved, techniques honed and methodologies matured. Countless individuals have been trained, baptized by the fire of experience, and have proven capable of building successful DW/BI solutions. Clearly, there will always be newbies to be trained, mentored, and seasoned. Yet, as an industry, we know what it takes to be successful. There are no excuses for failure.
My suggestion for ongoing success is to keep your eyes wide open and constantly focus on the basics – the fundamental blocking and tackling of data warehousing. Embrace these tried and true concepts that years of experience have revealed to be true:
- Focus on the business and business requirements– Never forget to maintain a laser-like focus on achieving business value. The success of a DW/BI effort is totally dependent on business user engagement; keeping the business users involved and meeting their requirements dramatically increases the probability of success. In fact, it ensures it.
- Obtain strong senior management sponsorship– Lack of organizational support undermines the success of a DW/BI effort. Senior management must accept, support, manage and fund these efforts as long-term programs.
- Organize for success– Successful DW/BI initiatives are undertaken as a partnership between the IT team and the business units. The DW/BI effort cannot be identified as just an IT effort. The business community (not IT) needs to take ownership of the vision, strategy, roadmap (priorities), scope, timelines, governance, and quality of the DW/BI environment.
- Integration is critical– Leverage conformed dimensions as the basis for integration. Understand and embrace the discipline of dimensional design as the organizing theme for the data within the DW/BI environment.
- Establish and enforce common vocabulary– The first step to eliminating data inconsistency issues is to establish and enforce common names, definitions and descriptors across the DW/BI environment. Again, embracing the discipline of dimensional modeling is the secret to victory.
- Focus on data quality– At its foundation, the DW/BI environment must be a repository of high quality data. The biggest challenges for most DW/BI efforts are data-related. In general, the effort to develop the ETL processes required to provide high quality data is typically more demanding than anticipated. Adequate time and resources, including key subject area experts, must be allocated to these tasks.
- Establish a phased design, development and deployment plan– Constructing a DW/BI environment is a significant effort. It is nearly impossible to tackle everything at once. Embrace an iterative development plan that avoids an overly ambitious scope. Project iterations need to be identified to phase-in the overall design, development and deployment effort. The environment should grow by tackling new and additional business processes (i.e., fact tables).
Exhibit patience – The initial phase of a DW/BI initiative usually takes a disproportionately long time; I call this the “big gulp.” This is necessarily true. While the initial phase corresponds to deploying a single business process, the bulk of the time in this phase is committed to designing, developing and deploying the critical, core set of conformed dimensions that will be leveraged throughout the DW/BI environment in later phases. This is normal and to be expected.