The Conclusion

With the increasing need to maximise profit in organisations, managers and executives will look for ways to improve their bottom line. One method is to engage Business Intelligence tools that will allow them to analyse their data for decision support. Tools like Decision Support Systems (DSS), Executive Information Systems (EIS) and Business Intelligence (BI) are becoming more common and will most likely require a Data Warehouse to host and support these tools.

Success of a data warehouse implementation can be measured when the consulting firm and the client have executed the project according to the agreed requirements and project scope. The implementation of the technology along with “Subject Matter Experts” (SME) who can provide and support the organisation with decision support tools are not the only elements for a successful data warehouse implementation.

As discussed in the earlier articles, data for decision support is backed up by factual evidence from transactional systems. The use of a Data warehouse is used to address problems relating to limited decision support tools (operational reports) in a transactional system and will prevent the 're-inventing the wheel' syndrome for each new decision support tool (Meredith, O'Donnell et al. 2008).

Good project management by itself will not determine the success of a data warehouse implementation. There are other pitfalls that the project manager should be aware of and therefore need to take steps to avoid them. As discussed in the earlier articles, the main issues in a Data Warehouse implementation are not merely technology or system architecture, the majority of the issues are related to:

• Stakeholder involvement and sponsorship
• Sufficient Funding
• Cultural awareness
• Agreed understanding of project scope and Scope management
• Organisational Politics
• Stable workforce
• Good project management

The Extraction, Transformation and Load (ETL) process is perhaps the most critical component of a data warehouse, however getting this right does not conclude that the implementation is a success. ETL is a continuous process as the source database will continue to change and when this occurs, the business rules or transformation logic will also need to be changed and thus, code changes may be required. Data Warehouse teams cannot “set and forget” ETL routines after a successful iteration of the ETL routine. This routine needs to be fine-tuned or refined regularly to protect the integrity of the data warehouse.

Even though reports are being produced by the BI application, care is required to maintain the integrity and validity of the data in the data warehouse due to changes that could occur in the transactional systems. Transactional and data warehouse teams need to standardise a process of communicating changes to the source system (Westerman 2001).

In a data warehouse environment, it is unlikely that stakeholders are able to specify all their reporting and analysis needs upfront. Therefore, getting out some sample data from the data warehouse early on in the project will give the users a taste of the sort of information they can get out of the data warehouse. This iterative manner of “show and tell” will generate far more accurate understanding of the requirements thereby ensuring that the data warehouse implementation is tracking accordingly.

Some projects will fail regardless and others will succeed but the key is to understand the issues of the organisation and their stakeholders so as to better the success outcomes of the project. Failed projects will make news whereas successful projects tend not to be heard. An example is the NASA’s Mars Climate Orbiter loss (Sauser, Reilly et al. 2009) millions of dollars were spent where the return of investment was negative. They believed that the project failed due to technical issues but the real cause of the failure was due to bad management.

REFERENCES
Meredith, R., P. A. a. O'Donnell, et al. (2008). Databases and Data Warehouses for Decision Support. Handbook on Decision Support Systems. F. V. a. Burstein and C. W. Holsapple. Berlin, Germany, Springer-Verlag. 1: 207-230.

Sauser, B. J., R. R. Reilly, &, et al. (2009). "Why projects fail? How contingency theory can provide new insights – A comparative analysis of NASA’s Mars Climate Orbiter loss." International Journal of Project Management Available online 14 February 2009: 15.

Westerman, P. (2001). Data Warehousing: using the Wal-Mart model, Academic Press.