What happens to your data warehouse with the new system
Most companies use data warehouses to store, analyze and report data about their business operations. They are an important part of the information systems landscape. They collect data from multiple operations systems. They help users make sense of the business and drive their key decisions. Large operational system replacements such as a new ERP system trigger the question of what to do with the existing data warehouse. Let’s try to shed some light on this important topic.
First, we need to assess how much the business trusts the data in the current data warehouse. If there is no trust, there is no legitimate reason for its survival. You will be surprised how many data warehouses exist out there just because someone has created them at one point in time. Most are not kept updated with the changes in the business. They do not serve their intended purposes. Thus, it is important to verify how businesses use them before we make a call on whether we keep or replace them.
Second, we need to understand how much of the data within the data warehouse is fed by the system being replaced. The lower the portion, the higher the chances that we will keep it and feed it. Imagine a data warehouse collecting data from multiple systems such as Forecasting, PLM, ERP, Retail, e-Commerce, etc. systems. Replacement of any of these individual systems would not trigger a replacement of the data warehouse. It is most likely that the new system will end up feeding the existing data warehouse. In addition, the new system may have its own out-of-the-box reporting capabilities on the side. On the other hand, imagine a data warehouse being fed mainly from the ERP system. Replacement of ERP will trigger an assessment of the existing data warehouse solution. In such cases, there are three options - append, co-exist, or full replacement.
If the new system is an upgrade where the data schema of the old and the new system is similar, you may decide to append your data warehouse. This will help you keep your old data while feeding the new data with additional elements. You can repurpose your reports, add new ones as needed, and report across your old and new data.
If the new system is a full replacement where the data schema of the old and the new systems are different, you made decide to co-exist with your data warehouse. You cannot afford to lose your old data. You may create a new data warehouse leveraging the latest technology and the new data schema aligned with the new system. Upon go-live, you’ll stop feeding your legacy data warehouse and keep it on read-only mode. Your reporting tools will tap into both legacy and new data warehouses to give you historical analysis.
The full replacement option is the most expensive proposition. You need to extract, transform and load the data from the old data warehouse to the new one. Your reporting tools will have one resource to tap into yet these are difficult and costly projects that are rarely undertaken. If the legacy data warehouse was built on an unsupported technology, not trusted by the business, or no longer addresses the ever-changing business needs, you may justify the cost.
If you are interested to learn more, please connect with me on LinkedIn, follow me on Twitter, or watch me on YouTube.
My name is Cem and this has been another gem.
