Have you tried bypassing development of a data warehouse by running your reporting directly against an enterprise system?
In the last few months, I have received a greater than normal amount of e-mails inquiring about the pros and cons of using transactional systems as the main information delivery data store. After all, in today's environment of large integrated source systems such as human resource and financial enterprise resource planning (ERP), customer relationship management (CRM), supply chain management (SCM) and other enterprise systems, is a data warehouse really still needed? Today's business environment continues to demand fast access to the most accurate, latest, most current information, right? Could eliminating the development of a data warehouse and reporting directly against the transactional enterprise system be the right choice? Technology today continues to provide increases in processing speed, memory and disk capacity, I/O throughput and other efficiencies, which may also allow consideration of this approach. Database management systems have provided data warehouse-focused capabilities that could be used in source systems to provide denormalized structures for business reporting needs. Is it time to revisit the need for the data warehouse in your environment?
This is the first installment of a three-part series that will examine the pros and cons of using the transactional system as the primary information delivery data store for the enterprise or business area versus a data warehouse approach. This first part will examine the arguments for use of the transactional system as the information delivery system of record. Part two will examine the shortcomings of this practice. The third part will explore the value and some shortcomings of the data warehouse method for business information delivery. This series will provide you with useful insights and questions that will allow you to examine or maybe reexamine your own information delivery ideas and approaches.
Transactional Reporting Pros
There are a number of reasons why some organizations look at using their enterprise transaction systems as the reporting system of record as opposed to using a data warehouse. The most common rationale for this approach is based on the interminable argument of time and money. Let us examine the first portion of this argument: time.
By using the enterprise transactional system for information delivery, the company can save time by not developing a data warehouse. The project time involved in the planning, analysis, design and implementation of a data warehouse data model and database can be avoided. This would seem to be a substantial amount of project time savings given that this portion of a data warehouse development cycle can easily take months or more to accomplish. The analysis time saved to develop the data warehouse data model alone would be substantial. Today's database technology should provide the required performance to run complex information requests against the enterprise transactional model. If specific performance issues arise for certain information queries, database techniques - such as materialized views or other methods - can be implemented to augment the transactional database design.
Next, there is the time saved in data integration design. Data warehouses typically follow a denormalized or dimensional approach in their database design to facilitate business understanding of the data and provide increased performance of information requests. This requires considerable analysis and development time for the complex data integration processes and programs to carry out this type of data manipulation. The project time savings here again can be significant given the typical duration of these data integration tasks.
Then, there is the time saved in refreshing the data warehouse. A data warehouse is typically updated on some sort of periodic basis such as weekly, daily, hourly or near-real time. In today's business environment where fast access to current data is demanded, running information delivery requests directly against the enterprise transactional system or real-time replicate database would seem advantageous. Issues associated with when the data warehouse was last refreshed or how accurate the data is would appear to be resolved.
Finally, there is the argument of the cost savings associated with development of the data warehouse in terms of time, people resources and technology. The cost of developing even a relatively small data warehouse from an enterprise system can typically be measured easily in hundreds of hours if not more depending on the scope. Additionally, there are cost savings that can be measured in the data storage, database management, ETL (extract, transform and load) and data processing. For these reasons and more, some companies shortcut their information delivery development by eliminating the data warehouse.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access