For starters, it might be useful to better define what we mean by business intelligence (BI) tools, since there is often market confusion. BI applications are tools built specifically to mine and manipulate the massive amounts of business data from your company’s existing data stores. The ability to spot changing revenue and payout trends to manage risk appropriately is essential in the competitive insurance carrier market segment. Easy to understand and use tools that can quickly extract data from large data warehouses and turn it into useful actionable information for busy executives are highly desirable. Often BI applications are extensions to and tightly integrated into the underlying database tools allowing for real-time views into company operations. Springfield has a considerable advantage here, given that it has already invested heavily in an HP/UX data warehouse and has the infrastructure and staff to support it in place.

Business intelligence tools can definitely spot business trends and guide Springfield’s management team with developing its strategy, but the tools are complex and can be deceptive if not used properly. Sometimes, what might look like a trend is really an artifact of the data or the system itself. Management must work closely with the development team to make sure they create or build the right analysis tools for the job. In the insurance industry, where everything is so dependent on risk management and close analysis of cash flows, this is a particularly important consideration.

So the main question remains, how easy will it be to port the existing HP/UX infrastructure to a Linux platform and what business intelligence tools can be added to the existing applications. The answer to the first question is relatively straight forward; since Linux is similar enough to HP/UX, porting the existing data structures and applications is more a mechanical exercise than anything technically challenging. Once the core applications have been ported, the next step would be to tackle adding business intelligence tools to the existing system. In the past, business intelligence software had a well deserved reputation for being expensive, difficult to develop, and hard to use because of the massive amounts of data that it needs to sift through and the proprietary nature of most of the available systems. In recent years, the software has gotten easier to use, but most of what is available is still expensive, complex and tightly coupled to the underlying system.

Given the current state of BI software, Springfield has three choices: building proprietary tools in-house, purchasing commercially available BI tools, or using open source tools. Each approach has advantages and disadvantages. The tradeoffs, as is frequently the case with these things, are development time and customization vs. money and standardized products. Since Springfield has a strong development group in-house and the decision has been made to use them to port the existing HP/UX applications to Linux, you will have a well-trained team that is both familiar with the company’s requirements and the particulars of the system in place when you are ready to implement the BI tools. There is no doubt that they will be able to build a highly customized tool that will meet your exact needs – for now at least. In-house development can be expensive if you don’t watch out for scope creep. The disadvantage of developing such a tool in-house is that you will not be able to take advantage of the expertise and “best practices” that the commercial vendors offer.

Looking at available commercial BI applications, the picture looks pretty good. The leading business intelligence vendors are waking up to the potential Linux market and are increasingly supporting the platform. IBM and Hyperion both have Linux-compatible BI products, while Business Objects recently announced that Crystal Enterprise 10 is fully certified and compatible with Red Hat Enterprise Linux. In addition to Linux compliant commercial tools, the market continues to pressure vendors to provide more open source or free options as well. IBM recently released a Java-based database, Cloudscape, into the open source community via the Apache Software Foundation. This may force Oracle and other major database vendors to follow suit. What this means for Springfield, is that the good news is that there is good commercial software available. But as you well know, these are expensive products with long deployment cycles and steep learning curves.

Finally, looking at what is available from the open source community, clearly the open source software development community has a long way to go. There are many Linux open source database projects, only a few have any analytical tools or report generation capabilities. Mondrian is often mentioned as an application with enough OLAP (online analytical processing) utilities to become a springboard for some serious business intelligence tools – if developers are willing to take the time to build them. It implements the Multi-Dimensional Expressions (MDX) language, and the XML for Analysis and Java OLAP specifications. It reads from SQL and other data sources, and aggregates data in a memory cache. For the immediate future, the roll your own and open source business intelligence tools are going to be rudimentary at best, but as more companies look for increased value for their BI spending and the open source community increasingly devotes energy to projects beyond the systems and back office markets, we will see more of these toolkits available to help Springfield build the business intelligence applications it needs. The option of using some of the open source tools as a platform or springboard for further in-house development could be an attractive option to consider. This has the advantage of using in-house talent to create a customized set of applications using at least some existing tools.

If Springfield’s management team is comfortable with sophisticated IT tools, then they will definitely gain competitive advantage by investing in and using an executive dashboard analysis tool. The approach that will give Springfield to get the most out of its investment in business intelligence tools in the short term would be to invest in a commercial product rather than building tools from scratch. Another strong option would be for the in-house team to take advantage of the existing open source applications and build a platform from there. This approach would be slower to implement, but more tightly aligned with Springfield’s specific business needs.