For the System i, it's not too great an exaggeration to say that databases are at the heart of some of the greatest services the platform provides to the enterprises that use it. And if business intelligence (BI) tools are what helps the brains of various organizations make decisions crucial to the functioning of those enterprises, the combination of databases and BI tools are more than just symbolically organic to System i users.

This buyers guide takes a fresh look at some of the BI applications available today to System i users and presents a picture of the broad spectrum of functions and capabilities those products have to offer. Although we solicited responses from as large a group of vendors as we could find, only 12 were able to respond in time for publication. However, we plan to maintain this buyers guide online at SystemiNetwork.com so that additional products can be included in the future, upgraded product features can be added, and new feature alternatives can be inserted into the table as significant new BI functions become available. We hope additional vendors participate in the future by providing product information for the web version of this article. [Editor's Note: Between the time this article went to press and the time the magazine hit your desk, several vendors had requested that their information be added, and we've added their entries in the online spreadhseet.]

BI is becoming more crucial to all enterprises. Sarbanes-Oxley and the many other laws and regulations designed to protect public investors from unscrupulous business practices has put a greater emphasis on the reporting aspects of BI for accountability reasons. At the same time, the increasingly global and automated nature of business today puts more emphasis than ever on the ability of managers and others to accurately analyze how well their business is doing and figure out how to make it perform better. This pressure also puts more emphasis on a concept called "a single version of the truth," which refers to the need to verify and synchronize all sources of data in a company so that older or more inaccurate information doesn't skew a vision of the enterprise.

Paradoxically, BI projects themselves seem to be under less scrutiny, as a recent study commissioned by the Noetix Corporation showed that although 95 percent of all companies do return-on-investment research on most IT projects, a majority don't do them on BI implementations. All these factors combine to make selection of BI tools an important task to do correctly.

Hopefully, resources such as this buyers guide can help. To compile it, we began by querying not just our own technical resources but some readers and many of the vendors as well, asking them all for a list of features they considered important for a BI application. We also searched online for feature lists of BI applications on other platforms. From this wide array of sources, we built a questionnaire that tries to look at a broad assortment of features and functions available in this product type. We questioned vendors about their products' user interfaces, graphics, querying capabilities, processing and calculation abilities, potential data input and output options, report-building and distribution features, security controls, administrative tools, metadata features, and special functions. From their responses, we've summarized those features for you here. While this global view can't provide a complete picture of any product, for which we urge you to consult the individual vendor websites for more information, it does give you a general view of the geography of BI solutions available for the System i.

Among some of the details the questionnaire covered were more than 24 graphical display options, 19 query-building possibilities, 14 report-building choices, 10 report-distribution selections, 18 security alternatives, and 14 product-execution controls.

A few term definitions may help you understand feature options better.

In the data-analysis feature area, Online Analytical Processing (OLAP) is a process of gathering data to support common business functions such as budgeting and reporting. Multidimensional OLAP (MOLAP) puts data into a multidimensional array called a "cube" and outputs the result in a row and column type matrix called a "pivot." Relational OLAP stores data for analysis in a relational database and generates SQL queries to extract information. All three methods have advantages and disadvantages, discussion of which is beyond the scope of this article. However, one important aspect of choosing a BI tool is to decide which method is best for the type of information your enterprise or application is most likely to work with, then look for tools that offer that kind of processing.

In the file types area, XBRL stands for Exensible Business Reporting Language, an XML-like means of electronic communication for business and financial data.

In the special functions area, there are several terms that may be unfamiliar. Autonomic servers are database servers that are self-monitoring, automatically adjust to workload demands, automatically incorporate and self-configure new components, tune themselves, and anticipate and defend against security threats. Database parallelism is a practice borrowed from high-availability concepts that writes data to more than one copy of a database so that if the primary resource is lost, a duplicate stands ready to continue normal processing smoothly. Data cleansing is a process of checking entered information against duplicate data known to be correct, automatic elimination of typos, and other validation processes. Geospatial analytics is a process of verifying geographical metadata against known valid sources, such as U.S. Geological Survey maps and International Standards Organization benchmarks. Predictive analytics is the practice of analyzing historical data about a customer to predict the risk or opportunity of a specific future transaction.

One important aspect of any BI tool for many enterprises is the amount of time it will take to train end users and administrators in a product's use. We queried each vendor about the number of hours required for an average user to become effective with an application. Although obviously there's no hard and fast definition of an average user, the responses will give you a relative yardstick of the amount of training each vendor expects from its experience. In addition, we asked about the training options available for users, which vary from in-house instructor training to interactive and self-paced instruction via the Internet.

The questionnaire also gave vendors the option of providing pricing information. So few provided it that we eliminated that information from the table. Actual pricing is variable depending on the number of licenses or seats involved, System i models on which the software will run, and even individual negotiating skills.

Finally, we should note that while most of these solutions run directly on the System i, a few don't. However, those that list a PC-based operating system as the primary OS all have the ability to access at least DB2 databases on the System i, in addition to other possible sources.

Because of the large number of features and products included in the final table, we were unable to include it in the magazine. You can find the complete table online at SystemiNetwork.com, article ID 21092. Currently, the table includes the following products:

• Applied Logic Software's File Edit Utility
• Cognos' Cognos BI
• Datawatch Corporation's Monarch|RMS
• Help/Systems, Inc.'s SEQUEL
• IBM's DB2 Web Query
• Information Builder's WebFOCUS
• LANSA's LANSA Client
• Lawson Software's Lawson Business Intelligence
• Mathan Software Services' ARC
• mrc's m-Power
• New Generation Software, Inc.'s NGS-IQ
• Planet J Corporation's Web Object Wizard
• Silvon Software's Stratum

If you're in the market for a BI application, now or in the future, we hope this provides a resource for helping guide you to a faster decision about which alternatives to investigate first. Whether you're interested in products that provide the greatest breadth of capabilities, or need to focus on a particular feature of interest such as scripting or wizards that automate certain functions, you can find both views here.

Finally, we want to stress that the solutions covered here are full-featured, front-end BI applications. This buyers guide doesn't include products that focus on only one aspect of BI, such as report building. It also doesn't include products from closely allied product areas, such as data warehousing, archiving, or content management, although individual products may offer some features in those areas.

John Ghrist is senior products editor for System iNEWS.
Erin Bradford is an assistant editor for System iNEWS.