In this age of rapidly expanding electronic data, enterprises have found it increasingly difficult to efficiently or effectively leverage their data to meet business goals. Important information may be found in every operational area of a company: sales, finance, customer service, engineering, marketing and more. Companies have traditionally tried to use their operational software programs to both run operations and analyze the resulting data. Enterprises have tried to use operational software programs designed for a specific functional area (e.g., sales automation, marketing automation, inventory management, ERP, or financial services) or export the data to Excel or a database for manual analysis.
Operational programs, however, were not designed to integrate, analyze or present data in large volumes or from multiple operational sources. Each of these approaches has drawbacks, summarized in the following table:
| Approach |
Drawbacks for business analysis |
| Reporting available within operational software programs (CRM, ERP, HR, accounting, etc.) |
- Primary purpose of operational software is for transactions and conducting operations, not analysis of the resulting data. Reporting capability tends to be very limited in terms of the variables and dimensions of analysis, requiring significant manual work (such as exporting data to a spreadsheet) to do analysis that business users really want.
- Reporting functionality within the individual program is only designed for the data captured by that program, so the user gets just a partial picture of:
- How this business function impacts the rest of the business; and
- How what is happening in other operational areas, in turn, impacts them.
- Analyzing data in an operational system can slow down the capability for which the system was originally designed – running operations. This problem grows worse as data volume increases.
- Operational applications are designed for simply queries and to make insert/update operations quickly;
- Business Intelligence requires complex queries that span many tables and operate on large row sets; and
- Making large-scale analysis demands on an operating system can crash the system or slow it significantly.
- May require expensive and time-consuming integration with other operational groups’ programs in order to achieve desired cross-functional analysis.
|
| Excel |
- Highly manual. Extremely cumbersome for data analysis, report presentation, and for ongoing data updating and maintenance. . Updating an analysis with new data often requires re-running the analysis from scratch. The spreadsheet is often only easily understood by its original owner.
- Smaller scale data sets only. Does not handle the larger data volumes which BI solutions can easily handle.
- Insufficient analytic capability:
- Does not provide typical BI capabilities like slice & dice, drill down capabilities, pixel-perfect report layout;
- Requires an analyst mindset and advanced Excel proficiency; and
- Not usable for advanced analysis by the typical business user.
- Integrating data from multiple operational sources requires significant time and skill.
- Prone to error when updated.
|
| Databases |
- Databases such as Oracle or SQL are good places for storing data, but do not provide the breadth of analysis and presentation tools required for business analytics.
- Presentation tools are lacking:
- Lack charting, pivot table, dashboard functionality;
- Requires technical knowledge to craft SQL statements and present the results as a plain table; and
- Doesn't address how insights are shared in the organization.
- Integration of data from multiple systems may require custom work.
- Significantly slowed down by Business Intelligence software's large cross-functional data sets:
- Can crash the system;
- As with operational applications, OLTP-type databases are designed for simply queries and to make insert/update operations quickly; and
- Not well suited for BI type queries.
|
Continue to Examples Benefits of Shifting from Operational Application Reporting to BI software
