An InformationWeek survey found that nearly 40% of IT executives cited expensive software licensing costs as the number one reason they are unable to roll out all the BI initiatives they require, and a study performed by TDWI found that almost 60% of IT managers have cited recent economic conditions as the reason for conducting a re-evaluation of their software vendors in an effort to reduce costs and implement the business intelligence systems their business needs.
In addition, the recent global recession has caused what many believe to be a permanent change in mindset away from high-priced proprietary hardware and software to commodity technologies that are oftentimes found to be nearly technically equal, and sometimes actually superior in terms of overall efficiency and performance. TDWI found that nearly 60% of its respondents are now considering lower cost options for their BI database. Reasons for this included reduced budgets, frozen hiring, projects being put on hold, software acquisition denials, and staff reductions.
General Ways in which a Column Database Help Save Money for Business Intelligence
Column-oriented databases and the technology that is beneficially utilized by them may certainly sound enticing, but the question needs to be asked: how can column databases help contain IT costs? Here are a few general ways column databases help defray expenses:
- Most all column based databases exploit the advanced in commodity hardware (e.g. 64-bit, multi-CPU/core) and do not require expensive machines be used to achieve exceptional performance. For a column database that supports MPP architecture, the price-performance metric (taking commodity hardware into account) can become quite impressive
- The majority of column databases offer some form of data compression – either physical or logical – that helps save storage costs
- In addition to general data compression, many column based databases do not require the use of indexes, materialized views, or other supplementary storage structures that are used to help the performance of traditional RDBMS’s. Because these structures are not used, additional storage cost savings are the end result
- Because column based databases do not require indexes or complicated database design schemes (e.g. data partitioning, etc.) and do much of the performance design work automatically for the end user, oftentimes the need for highly experienced database staff is negated. This means that inexperienced IT staff can create business intelligence databases that perform extremely well without having to possess sophisticated knowledge of data warehouse designs and performance tuning techniques
- The increased speed at which column-oriented databases operate typically means that more business intelligence reports/requests can be delivered and decision makers can more quickly perform their job, which lessens staff idle time
Commenting on column databases and cost savings, Gartner’s Feinberg says, “In the two specific areas where column-store DBMS excel (analytics and archiving solutions), there can be a large cost savings and, with analytic applications, a large performance gain”.