The combination of effort, skill and industry data knowledge required to extract and load large amounts of data "accurately" into a data warehouse is too often underestimated and discounted. Professionals (DBAs and BAs) without the appropriate skillset and industry knowledge are often assigned to these mission critical projects while also being held to other responsibilities. Sometimes inappropriate skillsets are recruited or team members are assigned tasks for which they are poorly suited.
Often the magnitude of the server resources needed to process the new data volumes is underestimated. Existing equipment is sometimes assigned without a full analysis of a fast growing data environment's real capacity, budget, and scale requirements.
Successful execution requires commitments from the CEO, CFO, CIO and CMO. If the user community does not buy into the potential business benefits of the new analytical capabilities, and if data analysis is not made an inherent part of corporate culture, then these projects tend to fail. Success in Business Intelligence radiates from a steadfast commitment from the top.
Is BI in the Cloud a better Solution?
Using the cloud, the issue of equipment scalability is essentially eliminated. Resources can be added or removed easily as needed.
A comprehensive cloud-based Media data warehouse design equipped with working ETL, star schema, Media cubes and a presentation portal will accelerate execution and reduce risk. Of course, the organization must also recognize that selecting the BI toolset is not the solution itself, but rather the beginning of the project.
The data warehouse infrastructure can be managed by internal DBA and BA resources or outsourced in a SaaS (Software as a Service) model.
A corporation can decide which components of the BI architecture reside in the cloud and which reside inside the corporation. This process mitigates concerns regarding corporate data security.
With many media source systems now utilizing cloud environments - either private or public - the transfer of data from these systems becomes simpler and more transparent.