News Stay informed about the latest enterprise technology news and product updates.

Assessing Analytic Packages for Enterprise Business Intelligence

Stuart Mullins explains that determining if an analytics package is right for a particular organization requires careful consideration and understanding of the costs and benefits.

This article originally appeared on the BeyeNETWORK.

Organizations looking to enhance their business intelligence (BI) capabilities will find the market offers more options than ever. Many, if not most, companies run significant parts of their business on packaged applications, such as those from Oracle and SAP. Today, most large enterprise software vendors offer some type of pre-built reports or dashboards centered on typical business functions such as finance, sales and marketing, and supply chain management. In addition, a number of smaller companies have developed packaged analytics built on licensed platforms and targeted to specific industries, applications, or specialized functions or to supplement other third-party services.

While a true “data warehouse in a box” has remained somewhat elusive, analytic packages bridge the gap between off-the-shelf canned reports and a ground-up custom BI solution. Such packages simplify the development process and deliver functionality with more cost certainty. Determining if a package is right for a particular organization requires careful consideration and understanding of the various costs and benefits of the solution options. Knowing these factors can help decision makers make the best choice for their organization.

Packaged analytics can take a number of forms, so let’s first establish a definition for packaged analytics in the context of this article. For my purposes, the term “packaged analytics” is used to represent pre-built reports, dashboards or other information delivery vehicle that is based on acquired data from enterprise applications and sold separately from the applications by either the same vendor or a third party. While ERPs come with some basic reporting or analytical functionality built in, we are considering packaged analytics to be an additional purchase, though certain discounts or bundling may be applied by the vendor.

The decision to build a custom solution or buy a package can be the difference in the success or failure of the business intelligence project. Before making the build or buy decision, the following factors should be considered:
 

  1. Source applications

  2. Timing of delivery

  3. Budget considerations

  4. Comprehensiveness of needs

  5. Business ability to adapt

Each factor must be considered and weighed against the overarching goal of the project. It might be tempting to make this simply an ROI case. However, other factors, such as the ability of IT to support the solution and the business’s ability or requirement to adapt can have large impacts that are often difficult to quantify. Let us consider each in a bit more detail.
 

Source Applications

The type of source applications plays a significant role in determining whether or not to custom build. If the environment is highly standardized on packaged applications from enterprise vendors, then packaged BI solutions are a natural fit. The packages’ standardized set of metrics and hierarchies can be easily mapped to the application. This exercise alone often accounts for 30-40% of the cost of a typical BI project. For many applications, this savings will pay for much of the package itself. If customized or homegrown applications are to be sourced, the mapping effort will be more significant; but there may still be a benefit to using the packaged data model and reporting application. High level requirements should be captured and compared to the packaged offering in order to determine its suitability.
 

Timing of Delivery

Packaged analytical solutions can provide a jumpstart to a BI project even if the long term plans call for customization. Utilizing a pre-built data model, ETL packages and/or reporting infrastructure can deliver value in weeks rather than months. If the business is desperate for analytics where they have none, then turning to a packaged solution may be the quickest way to get information into the hands of users. If time to delivery is not a major factor, then other considerations should probably drive the decision.
 

Budget Considerations

Packaged solutions usually incur greater costs up front than custom solutions, since more of the investment is made in software than implementation. Theoretically, packaged solutions should be less to maintain over the long term, but this may not always be the case. Demand for information is never static. Sometimes, it is difficult to know what the information demands will be 6 months, a year or more down the road. A package that allows for customizability and scalability will provide greater cost savings over the long term. Otherwise, the organization may find itself looking for a new solution sooner than expected if the analytic platform is not flexible enough.
 

Comprehensiveness of Needs

Packaged solutions tend to be rather comprehensive since they are designed to fulfill the most possible requirements or uses for a given subject area. When determining whether to purchase a package, one should be mindful of how thorough the organization’s needs are. For example, a packaged financial data model may contain as many as 40 fact tables and hundreds of dimensional structures. This provides a great deal of functionality and flexibility, but it can be more difficult to isolate problems when they occur. Some organizations may not require all the subject areas, aggregations and metrics that are included in the packaged application. A simpler custom solution may be more economical and maintainable.
 

Business Ability to Adapt

Perhaps, the single most important factor is the business’s ability to adapt to standardization that a packaged solution provides. A business community that has historically demanded and received custom applications may challenge the package environment’s suitability to their business intelligence needs. This could be due to the function of the user group. For example, marketing processes tend to be less standardized across industries than accounting or financial activities. Another reason may be the industry itself. While manufacturers or retailers may have functions that are very similar to other organizations in their line of business, many service organizations’ business processes vary substantially. It is important to understand the business requirements and include the business community in the decision-making process.

While one factor may be the primary concern that ultimately drives the decision-making process, we must be careful not to overlook other considerations. As the business will change, so will requirements. Mergers or acquisitions and new products or executives can lead to substantial changes in the business needs for information. The best solution is one that provides the greatest flexibility at the lowest total cost of ownership (TCO). No matter which direction is chosen, some compromise will be required. When all the factors are assessed, the solution that best fits the organization usually becomes apparent.
 

Dig Deeper on Business intelligence software

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.

-ADS BY GOOGLE

SearchDataManagement

SearchAWS

SearchContentManagement

SearchOracle

SearchSAP

SearchSQLServer

Close