KDnuggets Home » News » 2016 » Sep » Opinions, Interviews » Embedded Analytics: The Future of Business Intelligence ( 16:n35 )

Embedded Analytics: The Future of Business Intelligence


 
  http likes 286

An overview of the evolution of Business Intelligence, and some insight into where its future lie: embedded analytics.



By Wayne Eckerson.

An Era of BI Tools

 
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.)

Evolution of BI

Figure 1. The Evolution of BI — and Its Future

Closing the Last Mile of BI

 
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?)

Evolution of Embedded Analytics

 
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.)

Decade BI User Features Data Sources Platform BI Software Client Code APIs Developers Pricing
1990s Static
Reports
RDBMS, files Desktop Desktop BI tools Windows, Unix Code-specific libraries Independent software vendors (ISVs) User-
based
2000s Interactive reports, OLAP, dashboards OLAP and XML Web Web BI tools ActiveX, JVM, Flash, Silverlight SOAP, iFrames ISVs and internal developers Server-based
2010s Self-service, predictive, and blended analytics Cloud apps, big data, NoSQL, streams, search Cloud BI platforms AngularJS, Ember.js jQuery, AJAX, etc. REST, JavaScript, ISVs and internal developers Value-based

Table 1. Evolution of Embedded BI

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. Web and cloud-enabled BI applications run on separate servers and thus share no code or libraries with host applications, making it easier to embed analytics into those applications.

APIs and Developers

 
APIs. The embedded analytics market revolves around application programming interfaces or APIs, which developers use to either customize or extend an analytical tool or call its functions from within a host application.

Many BI vendors today ship platforms, which come with a rich set of APIs and cloud-based deployment features that make them ideal for embedding into other applications. With traditional BI tools, what you see is what you get—you must embed the entire tool, including the graphical user interface (GUI). But analytic platforms are infinitely customizable and extensible and can be “beheaded.” That is, customers can ditch the product’s packaged GUI and write their own that exploit’s the analytic platform’s core functionality.

APIs have evolved from client/server programming libraries (e.g., COM) to Web-based interfaces such as SOAP, iFrames, and client-side plug-ins, to modern standards such as JavaScript and REST. Many BI vendors, especially the sponsors of this report, supply JavaScript APIs that make it easy to create new charts and graphics, modify existing ones, and embed BI visualizations into other Web applications. In contrast, REST APIs are generally used to access back-end administrative functions, such as publishing, provisioning, scheduling, and user administration.

Developers. The community of developers has expanded from coders working at software vendors who used tools like Microsoft Visual Basic and Visual Studio to embed static reports into commercial software products. Today, developers are found within information technology (IT) departments and application development shops in user organizations. In other words, embedded BI is becoming pervasive.

Although software vendors still comprise the lion’s share of organizations implementing embedded BI, corporate developers are catching up. As executives recognize the value of their data assets, they now want to monetize it by developing data-rich applications for customers and suppliers. This typically involves embedding reports, dashboards, self-service analytics, and predictive models into tiers of outward-facing applications.

Brave New Market

 
Embedded analytics is not new, but the technology for integrating charts, reports, dashboards, and self-service tools has evolved considerably in the past 30 years. Formerly, only software vendors embedded analytical tools into applications, but now organizations in every industry are doing so. Consequently, many BI vendors now target the embedded analytics market as a high-growth area and have architected their products to make them easier to embed and manipulate from within other applications.

Wayne EckersonBio: Wayne W. Eckerson is an internationally recognized thought leader in business intelligence and analytics who thinks critically, writes clearly, and presents persuasively about complex topics. He is a sought-after consultant, noted speaker and bestselling author. Eckerson is founder and principal consultant at Eckerson Group, a research and consulting firm that helps business leaders use data and technology to drive better insights and actions.  He can be reached at wayne@eckerson.com.

Related: