It’s a classic business situation. IT organizations and end users face the "buy vs. build" decision when evaluating their options for information systems. Buying is usually the preference. Compared to building, buying requires fewer IT resources, gets the customer to deployment far sooner, brings with it the domain expertise automated by the application and transfers the burden of risk (especially concerning system integration) firmly onto the shoulders of the vendor.

However, finding a packaged analytic application that – out of the box – satisfies all the requirements of users and IT is particularly challenging, because the four areas of customer requirements typical of an analytic application vary from one end-user organization to another. Despite the variance, many common attributes can be anticipated by software vendors and packaged into analytic applications.

  • Internal data sources. Data structures in information systems have much in common, regardless of the format they take or the origin of the data. These systems usually contain data about products, revenue, customers, operations and so forth. Furthermore, databases store these in fairly predictable data structures. Hence, approximately 60 percent of data sources in the average company can be anticipated and addressed with prebuilt methods for data extraction, transformation, move and load (ETML).
  • Analytic data model. Many entities recur across businesses, and most businesses want to analyze these entities in the context of a multidimensional model that includes dimensions for common business entities such as time, geography, product lines, revenue and operations. As with data sources, the vendor of an analytic application can anticipate and prepackage the majority of these business entities.
  • External data sources. External data (representing geographic, demographic, consumer or business-to- business entities) has published data models that can easily be integrated with an analytic data model.
  • Reports and queries for specific users. Most analytic user bases are a mix of dedicated business analysts, regular report consumers and occasional users. Even so, users across companies share many common needs that in turn allow a vendor to supply a collection of analytic reports and queries that address most of these needs.

THE HURWITZ TAKE: These four areas of unique customer requirements – and the challenges of satisfying them – define most analytic applications. An analytic application is a package of templates and other prebuilt modules that anticipate the wide majority of customer requirements concerning ETML for internal data sources, integration with external sources, a data model for multidimensional analysis and reports and queries for specific end users.

A well-designed analytic application provides significant value out of the box, but it still requires a certain amount of customization to adapt it to the customer’s business and to fulfill the remainder of requirements. Hence, vendors of analytic applications should offer a professional services group that can perform the customization and train the customer’s IT organization to take control of the administration and future enhancements of the analytic application. Purchasers should ensure that the analytic application platform they are considering supports as much prepackaged value as possible, while involving consulting services with a well- defined end point.

This approach to analytic applications – where a single software vendor provides easily customized prebuilt components – is an excellent compromise for the "build vs. buy" argument. IT and user organizations can purchase a loosely packaged, but fully integrated analytic platform for quick deployment, then devote minimal resources to customizing it according to internal requirements.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access