As you set out on a new Microsoft Dynamics 365 ERP project, it is very important to plan for and consider the role business intelligence (BI) should play. To start, gather an inventory of all the data sources that will be used for your BI or analytics. One mistake often made during analytics projects is to ignore the ‘shadow IT’ systems and data sources. These can range from Excel spreadsheets and Access databases, to accountancy systems running on a local PC or unconnected cloud based CRM services.
The inventory of data sources created here will be useful not only as we plan the deployment; it will be useful when the project goes into production. Having an enterprise-wide list of all the data sources used for reporting and analysis and making that list available to everyone, will help people discover, gain access to, and understand data sources they may not have known existed.
We also want to consider the types of data sources we have. Below are some key points to think about in regard to data source types:
Another major topic to think about is governance. We want to make sure that the right people have access to the right data. Like taking inventory of your data sources, you want to take an inventory of the data consumers. Make sure you understand how they will access the data as well as what pieces of the data they need access too (certain companies, departments, etc.).
Other key areas of data governance, but no less import are availability, usability, consistency & data integrity.
As with the importance of understanding your data, you must also understand and plan out what dimensions and measures will be needed for your enterprise analytics needs. This can be accomplished by speaking with Subject Matter Experts (SMEs) early in the project and creating an inventory for each area. This inventory should include all the ways that the data needs to be sliced/filtered on (Customer attributes, Date, etc.), as well as what measures (calculated values, %’s, counts, etc.) will be used in reporting.
Once we understand our data, the next step is to build the semantic model. This step is important as this is how your report developers and consumers will interact with the data. If the data is not easy to browse or use, there is risk of losing buy in or creating problems with user adoption.
Note: We do not provide specific information around choosing the architecture/technologies in this document. There are too many variations from client to client to list those here. If you need assistance deciding what physical system architecture you need for your analytics, we can provide an Enterprise Analytics Assessment and guide you to a solution.
©2023 All Rights Reserved. Privacy Policy Disclaimer
