The Significance of a Solid Data Strategy in Business
In my view, a solid reporting strategy is essential for any sizable organization, and is becoming even more important over time as data volumes and the need to integrate information from various sources grow. There are several key components that come to mind when putting together such a strategy document:
- An analysis of the various source systems for reporting, which may be enterprise business systems such as SAP or Workday, individual machines or sensors, etc. For each, attention should be paid to the nature of the data from each system. For example, does this data pertain to just one business area (FI, HR, Logistics, etc) or is it cross-functional? How is the data organized within each system? Is it a highly normalized one with many tables whose contents will need to be combined for reporting? Is the data validated upon entry (say only predetermined values can be entered in a field), possibly making reporting simpler, or is the information more unstructured? Which external data would need to be integrated into the solution, such as benchmarking data for industry salaries, current lead times for required raw materials, etc?
- Broadly, what are the business goals of manipulating or presenting different sets of data? For instance, some data may be used to support descriptive reporting, whereas some may be used as an input for further analysis before action can be taken. If data is used in a predictive or prescriptive way, such as adjusting a tool-making machine’s maintenance based on stresses measured within that machine, that may certainly impact the collection and processing of the data. While that processing may not strictly be considered “reporting,” certainly some outputs of that process (e.g., the maintenance activities, unscheduled downtime) would definitely need to be considered for reporting.
- A look at the desired reporting outputs, and the possible/chosen front-end reporting tools for various data/reporting goals – For example, is production-quality reporting (say that that would appear in an annual report) needed in some cases? How interactive will the reporting experience need to be with various sets of data? For instance, how much latitude will uses have in navigating/exploring the data? Will users enter new data to be used in reporting in what-if analyses?
A strategy document may address many of these factors, and can importantly be a starting point for conversations about the tactics needed to use the data in a way most valuable to the company. Without such thought and analysis, individual reporting solutions may be developed that meet narrow needs, while with careful attention the various reporting situations can hopefully be addressed with a more limited (and cost-effective) set of tools, and be simpler for the employees to learn and continue to use as they move from role to role in the company.
By: Jeffrey Sweet, Data Architect