Business Intelligence Success, Lessons Learned

A recent, extensive 269 pages industry report, OLAP 3 (published November 2003) by Nigel Pendse explains that BI benefits are very real. While the report covers many aspects of OLAP and BI, we will focus on business benefits and overcoming the obstacles of achieving those benefits.

According to the report, approximately 19 percent of companies implementing BI claim they have met or exceeded their business goals. Over 60 percent state they have at least largely met their goals. As always, soft benefits were more easily obtainable than hard benefits. A detailed look at the types of benefits reveals the following:

Benefit	% Companies Realizing Benefit
Faster, more accurate reporting	81
Improved decision making	78
Improved customer service	56
Increased revenue	49
Savings in non-IT costs	50
IT savings	40

When the statistics on benefits from a BI investment are compared to those of ERP or SCM investments, we see that BI appears significantly more beneficial. However, it is likely the combination of the ERP or SCM investment and BI generate the benefits because benefits imbedded in the ERP or SCM system cannot be unearthed without the BI tools to dig them out. Moreover, a common opinion on ERP and SCM software is that they generate too much data. BI allows this data to be analyzed to bring forward the most important aspects of the data.

Certain application areas are more widespread than others. The most common applications areas for BI are

1. General data warehouse reporting
2. Sales and marketing analysis
3. Planning and forecasting
4. Financial consolidation
5. Statutory reporting
6. Budgeting
7. Profitability analysis

The implementation of a BI project is not without challenges. Although the selection of the BI product is important, many of the major challenges remain internal. The number one major non-product challenge includes company politics. Experience has shown that single department implementations are less challenging but often result in fragmented, non-integrated approaches that, in the long run, increase interdepartmental issues instead of simplifying them. Experience also tells us that, as with many enterprise projects, those projects crossing department boundaries typically yield a greater value than departmental specific projects.

As with any project, the user is very important and the interest of the user must be considered key to project success. The factor of ease-of-use for users is critical in both product selection and overall success. The OLAP 3 report indicates that the inability to get users to agree on requirements is a common problem with BI implementations and if the requirements are agreed upon, staying the course without changing the requirements proves difficult.

In the report, the most common product issue was response time. This includes the inability of the product to function effectively with large databases. Another limitation that is both a product and architectural issue is the ability to include various data sources in the overall project.
When companies were asked what they considered important in product selection, they answered with the following priorities:

1. Functionality
2. End-user ease-of-use
3. Integration to existing applications
4. Price
5. Performance
6. Ease-of-use for application builders

However, when correlated with success factors, a proof-of-concept completed rapidly proves second only to performance as a key criterion for product selection.

Additionally, implementation leadership has a direct impact on project success. Projects led by BI specialist consulting firms are the most successful and have fewer problems. However, the most widely used implementation resource is in-house. Projects led by in-house business users are more successful than those led by in-house IT specialists. Projects led by large, general-purpose consulting firms cost more, are the least successful in business terms and experience the most problems. There is little correlation between project success and consulting dollars.

Strong evidence supports that the more quickly projects are rolled out are more successful. Experience also shows that integration into existing applications, prepackaged analytics and ease-of-use are key to compressing the implementation cycle.