Data Warehouse Modernization: Developing the Implementation Road Map
Drastic transitions in your environment can't happen overnight. To ensure success, develop a phased approach to migrate to your new environment.
- By David Loshin
- February 26, 2016
In past articles I have examined the motivating factors for data warehouse modernization, considered the business, technical, and strategic perspectives for assessment and identification of gaps in the environment, and explored agile technologies that are alternatives to be adopted as part of a hybrid data warehousing and analytics environment. Even if you have chosen one or more alternatives, remember that your environment still needs to continue operating without significant interruptions as you begin to acquire new technologies, deploy them, and move them into production.
It is unrealistic to presume that any drastic transitions in your environment can be effected immediately. At the same time, you cannot embark on a multi-year development project without having any interim deliverables and demonstrations of value. Therefore, to mitigate risks of a "big bang" approach to development, lay out a phased transition plan that will incrementally migrate data and/or functionality to the component technologies that will comprise what will eventually be your future environment.
You will need to develop one or more implementation road maps. Each identifies a sequence of discrete projects intended to steer the environment in the direction of your future vision. If the changes to your environment are not complex, one road map will do, but broader environmental changes may require multiple implementation stages, each with its own road map. The goal of each road map is to demonstrate that a limited investment in staged development will still deliver concrete results that can either be put to productive use immediately (such as adding a new predictive analytics capability) or contribute to the upgrade, replacement, or renovation of existing capabilities.
For each implementation work stream (such as platform renovation, visualization tools, end-user self-service data access for BI, reporting, or analytics), developing a road map begins by contrasting the current environment with your vision for the future. The road map embodies a sequence of changes to your current environment that share these characteristics:
- Each high-level project, consisting of a sequence of tasks, will have a clearly specified desired goal
- Each activity associated with designing or modernizing your data warehouse environment can be isolated as a set of complete project task plans that, when combined, will achieve the desired goal
- The completion of each project will demonstrate recognizable value to your organization
- The completion of the set of projects will contribute to the development of your future environment
To ensure alignment with the business, technical, and strategic perspectives, make sure that your road map shows how each project addresses specific business goals or improves the technology landscape. In addition, provide metrics that show how the successful inclusion of new capabilities addresses one or more of the identified systemic gaps. Finally, to address any concerns of "big bang" implementations, suggest how the successful completion of one project implies the creation of a new detailed sequence of project tasks deploying the next part of the transition.
In this way, successfully completing each project results in incremental improvement in the data warehouse environment as well as contributes toward incremental actualization of the strategic vision. At the same time, any newly identified risks can be considered without affecting the entire modernization plan. Such risks will trigger adjustments to your longer-term strategy, allowing you to make progress amid uncontrollable events.
David Loshin is a recognized thought leader in the areas of data quality and governance, master data management, and business intelligence. David is a prolific author regarding BI best practices via the expert channel at BeyeNETWORK and numerous books on BI and data quality. His valuable MDM insights can be found in his book, Master Data Management, which has been endorsed by data management industry leaders.