Despite its many benefits, business intelligence (BI) has not yet reached its full potential within many organizations. That's because many organizations have yet to find an effective way to deliver BI capabilities to more than a handful of users who have the technical and business literacy to exploit BI tools.
On the data side, organizations are moving to adopt "operational BI." They are modifying classic batch-oriented data warehouses with event-driven processing models so they can deliver real-time information to users. They also use federated query capabilities, Web services and service-oriented architectures to cull operational data in real time from disparate systems.
The promise of embedded analytics
In the end, the best way to simplify and operationalize BI is to embed it directly into operational applications and processes that drive the business. This is the definition of embedded analytics.
Embedded analytics are nothing new. Organizations have embedded BI functionality into applications and business processes for years. For example, Java and .NET developers often create reporting capabilities from scratch when building custom applications. Portals display charts and tables generated by reporting tools inside portal windows, or "portlets." Microsoft Office applications maintain live links to reports stored on BI servers. Extranet applications integrate with BI engines that let online customers view and analyze personal accounts and activity. Online applications embed predictive models that score customer transactions in real time to detect fraud, cross-sell products, evaluate risk or assess credit worthiness.
However, most embedded analytics to date have barely scratched the surface of what is possible. Usually, the analytics are fairly primitive, displaying canned views of existing reports with little ability to drill down, publish views in other formats, or compare with other data. The more compelling analytically driven applications are implemented by leading-edge companies with deep pockets and legions of skilled developers who code, debug and test monolithic applications that are time-consuming to build and costly to modify.
The future of BI
Today, however, visionary vendors and BI professionals are conjuring new ways to blur the lines between analytical and operational applications. They trumpet the benefits of composite applications, process-driven BI, business-activity monitoring, BI services, operational dashboards, integrated business application suites, Software as a Service and open source BI, among other things. With new development techniques that make embedding analytics into business processes and applications as easy as dragging and dropping objects onto a workbench, the future course of BI could change radically.
Rather than using standalone BI tool sets that require setup and training, business users will leverage embedded BI functionality that is an integral part of a larger application or package. Users will no longer shift software contexts when moving from operational processes to analytical ones. BI simply slips into the background of a primary application that users use to do their jobs. At this point, users may no longer realize that they're using distinct BI tools.
- BI as a service: To switch on a lamp, you must first plug it into an outlet that taps into the electrical grid that powers the lamp. Like electricity, BI is destined to become an enterprise service that users and applications tap into to deliver information and insights to users on demand. Embedded analytics transform BI from sets of standalone products to enterprise services that make BI easier to use and pervasive.
- BI as a container: Conversely, applications built using BI tools will serve as vehicles to launch operational processes and tasks, a kind of reverse embedding that some experts call "closed-loop BI." For example, many companies now use dashboards to monitor key business events, trigger alerts and workflows, and execute tasks within operational applications. In addition, some packaged software vendors now use a dashboard as the central metaphor for delivering role-based views of tasks and information required to manage processes in and across multiple business departments.
Developing embedded analytics
The most promising method of embedding analytics is to use emerging software development workbenches that make BI services available to developers as components that they can drag and drop onto a screen, configure and link together. These tools not only simplify and accelerate the development, testing and deployment of analytically driven applications, but they also use BI components that foster re-use and standardize the "look and feel" and delivery of BI functionality within an organization.
- IDEs: There are two types of workbenches available for the analytically minded developer. First, there are traditional interactive development environments, or IDEs, offered by commercial software vendors such as IBM (Rationale) and Microsoft (Visual Studio.NET). Many of these IDEs have added BI components to their workbench palettes so developers can embed BI functions within nonanalytical applications. These components must still be tied to analytical engines residing either within the local application server or remotely on another server.
- ADEs: Besides IDEs, some BI vendors offer developer workbenches devoted exclusively to building analytic applications (i.e., collections of interactive reports and views designed to support specific business tasks and processes.) These analytic development environments, or ADEs, work the same way as IDEs (hence the copycat acronym) but often abstract the development process to a higher level so that power users, not just developers, can rapidly prototype and build applications with the tool sets. Most of these ADEs eliminate the need for coding altogether when building BI solutions.
Embedded analytics are the next wave in business intelligence because they bring BI closer to the operations and processes that drive businesses on a daily basis. Embedded analytics won't replace standalone BI tools. Rather, they will make the functionality offered by such tool sets more readily available. By embedding BI functionality within operational applications and processes that drive the business, embedded analytics will make BI more operational, easier to use and pervasive -- key challenges facing the current generation of BI adopters.
Wayne Eckerson is director of research at The Data Warehousing Institute, a Seattle-based provider of in-depth, high-quality training and education in the data warehousing and business intelligence fields.