Business intelligence (BI) has long been associated with tools: business user tools for viewing and interacting with reports and dashboards, and business analyst tools for querying databases and visualizing, analyzing, and modeling the results. During the past 25 years, software vendors have shipped hundreds, if not thousands, of BI tools that help businesspeople turn data into insights and action.
But the era of BI tools may soon disappear. Or more accurately, the time when organizations purchased BI licenses for every knowledge worker may be drawing to an end. Rather than distribute analytical tools to employees, organizations will increasingly embed analytical output—reports, dashboards, insights, visualizations, and recommendations—into business applications that businesspeople both inside and outside the organization rely on to do their jobs every day. In short, the future of BI is embedded analytics. (See figure 1.)
Figure 1. The Evolution of BI — and Its Future
With embedded analytics, business users no longer have to exit a business application to view results, analyze performance, and view recommended actions; they can do this inside the application itself. Organizations can not only embed reports and dashboards into applications and portals; they can also embed syndicated data (e.g. demographics) and the output of predictive models, turning customer-facing applications into powerful tools for reducing costs, growing revenues, and increasing customer satisfaction.
Critics of BI have said that by the time business users look at reports and dashboards, it’s too late to take action and change outcomes. Embedded analytics addresses that challenge, turning BI from a reactive activity into a proactive one. It helps close the proverbial “last mile” of BI—turning insights into action that help an organization achieve its strategic goals and financial objectives. (For an in-depth discussion of embedded analytics, see Embedded Analytics: The Future of BI and a companion report Which Embedded Analytics Product is Right for You?)
The embedded analytics market is not new; it’s been around as long as there has been BI software. But the way BI has been embedded into applications has changed immensely since the mid-1990s. (See Table 1.)
For instance, the type of embedded analytics functionality has evolved from static reports in the 1990s to interactive reports and dashboards in the 2000s to self-service, predictive, and blended analytics today. In addition, embedded BI tools can now access a much broader range of data than relational databases, which were the predominant data source in the 1990s. In the 2000s, BI tools could routinely query OLAP and XML sources, and today, many support a bevy of cloud applications and big data sources.
Moreover, embedded analytics software moved from the desktop to the Web 15 years ago and now runs in the cloud, where organizations can rent the software on a monthly (or sometimes hourly) basis.