In this article – based on a concrete but anonymous zed case – ebs.dk presents scenarios for the current (Status) and projected future (Expectations) for an organization’ production and use of business intelligence.
The article also indicates which changes this is likely to mean, and how to get from the present to the future situation (Changes, and how we get there).
The presentation is based on analysis of System architecture, Information collection, Information production and Reporting tools.
System architecture
Status |
Expectations |
Changes, and how to get there |
Integration of information from different systems |
Compilation of information between, for example, ERP system time-registration and production system is partly based on extraction jobs encoded in SAS® and manual entry in MS Excel® spreadsheets. |
The integration of information will be based on rules and metadata (common taxonomy) described and maintained in a common repository.
Business integration between business areas is supported by a common process model.
Web services ensure access to necessary information. |
Individually developed procedures for extraction and data correlation in particular. spreadsheets will be canceled. Application of more general and uniform methods and standards-based solutions for data correlation makes this type of tasks solved more efficiently and with less need for detailed knowledge and experience of key employees than today. |
Data warehouses |
Information is stored in a variety of data warehouses based on SAS® extraction jobs, business application-included “default-data warehouses” and / or “add-on data warehouses” developed in parallel with and on top of each system. |
Data Warehouses based on consistent standard warehouses ("data mart") and standard built-in business intelligence functionality in MS SQL Server®. |
Existing data warehouses shall be shifted to standard solutions. This will probably mean phasing out of existing custom built data warehouses.
To ensure that the new data warehouses support the business need, is building with close involvement of process-, controlling- and governance-competences. |
|
|
|
Information collection
Status |
Expectations |
Changes, and how to get there |
Collection of source data |
Takes place mainly from production systems through the use of SAS® extraction jobs. |
Collection of information is conducted on the basis of reference XML information from ERM- and other production systems and from business partners via reporting systems. Data is validated against the data definitions.
Extraction is automated and supported by SQL Server Reporting Services® with XQuery. |
Improved data quality.
Unfortunately data breaks cannot be avoided as both data basis and not least the business processes and rules are changed. |
Coverage |
Core production is covered by the information generated in the production system.
Apart from that, there is no operational basis for exhaustive inventory of management information for other business areas, activities and processes. Including the production recorded in ERM solutions, etc. |
Coverage will include all of the business areas, activities and processes that are supported by ERM- and other production and knowledge management systems. |
More exhaustive source information. More dense and widespread information coverage across the organization’ processes and activities and a better basis for uniform decomposition and aggregation of information. Better information support of the organization’ governance model.
Higher complexity, underpinned by higher transparency, better data quality and better documentation. |
Organization of work |
Information is collected in many different ways by specialized unit: Processing in individual data warehouses and spreadsheets, manual enumerations in connection with quality assurance and measurement.
The information collection requires comprehensive and detailed understanding of operational procedures, rules and tools.
Not least exception handling is VERY required and depends upon deep and always up-to-date as well as historical knowledge of deviations from or not documented working practices and business rules. |
Full system support for information collection on the basis of a coherent data warehouse structure. |
More standardized organization of information collection. Wider spread of competences, less vulnerability and fewer bottlenecks. It will to a greater extent be the demand for information which will be steering for which information gets collected and to a lesser extent the possibilities. |
|
|
|
Processing of information
Status |
Expectations |
Changes, and how to get there |
Modeling and transformation of data |
Modeling of information to various data warehouses.
Further modeling based on data warehouses for specific investigations and reports.
This is today conducted in system specific data warehouses, in SAS®, Excel® spreadsheets, and in some cases directly on production data. |
Based on MS SQL Server® business intelligence framework and possible specific tools for specific needs |
Use of a relational database that supports XML and has much broader scope than the existing solutions.
Amongst other things are the current extract, transformation and data warehouse load supported in an expected future database solution "for free”, integrated, real-time and in a standardized manner.
Custom development in the existing business intelligence solutions will be abandoned. The largest portion will be eliminated; but the rest will have to be deployed in the strategic business intelligence solution.
There are expected to be significant savings by using standard and built-in functionality from the new systems for the modeling and transformation of data. In particular, the expected savings and/or strategic gains on maintenance and development. |
Analysis |
Poor system support of analysis involving business rules and data correlation.
There is only little support for analysis at the process level and even less support of process analysis across business areas. |
Common and comprehensive metadata model, process model, and descriptions of business rules support efficient process analysis — even across the organizational “borders”. |
Expected freed analysis resources.
Wider opportunities for analysis of coherent processes.
Possibilities of analysis on several administrative levels and better opportunity for decentralization of analysis, which to a larger extent will be possible to conduct locally in the relevant business areas.
Relevant parts of the analysis capabilities could be made available to the organization' business partners – this could be a competitive asset. |
Organization of work |
Processing of information is today supported predominantly in SAS® and in Excel® spreadsheets.
Processing of information requires detailed knowledge of programming, data sources as well as to business rules and practices. |
Processing of information is expected to be conducted on the basis of standardized documentation and evidence-based business rules and data sources.
Processing of information is expected predominantly to be supported of MS Office®, already used widely in the organization. |
Expected freed resources, fewer bottlenecks and wider access to the processing and analysis of information.
The use of well-known standard tools for information processing is expected to grow thus implication less dependence on specific programing competences. |
|
|
|
Reporting
Status |
Expectations |
Changes, and how to get there |
Reporting systems |
The following products are used today in the organization’ business intelligence reporting: a series of products from SAS®, Excel, and Access from Microsoft®, Crystal Reports from Business Objects®, SQL Report from Oracle® … |
General tools based on MS Office (SharePoint and Excel) with automatized publishing on intra-/extra-and internet.
Possible specific tools for specific needs. |
There will probably have to be phase-out plans for most of the existing reporting systems, which is expected to reduce barriers and bottlenecks – and imply massive cost reduction.
Standardization of competences and wider access to reporting systems. |