The transformation of Business Intelligence (BI) from an IT-centric, centralized process to a self-service, decentralized process is clear:
What is surprising is the form that self-service BI in the hands of end users will ultimately take. Traditionally, self-service analytics options were based on Data Discovery tools: search, visualizations, dashboards, data mashups, etc.
Advancements in vendors and analytics offerings, however (spanning everything from Machine Learning predictive analytics to conventional BI) are significantly broadening the scope of self-service BI to include structured and unstructured data, data preparation (including ETL), data-centric storytelling, and a host of other capabilities.
Consequently, self-service BI will account for seamless and expedient data integration between any number of data assets within and outside of the enterprise, granting the business user more control over BI than ever before.
Updating ETL
Traditionally, the ETL process of transforming and loading data into a warehouse for Business Intelligence was one of the most time consuming aspects of issuing reports. These complications are exacerbated by the size and velocity of Big Data; although it is possible to stream data into platforms with such as Hadoop for BI purposes, such options still do not account for the agility and variation of data types a business user requires for ad-hoc exploration for real-time decision making.
Enhanced Data Discovery
Vendors have equipped Data Discovery tools with a host of new functionality that transcends the traditional capabilities associated with these options, which are available either as components in BI platforms or as standalone alone solutions. In addition to assisting with ETL processes, such tools also come with data lineage and hierarchy generation, simplified user interfaces, automated visual data flow building, data blending, textual representations, semantic auto discovery features, and more. This added functionality helps laymen users successfully prepare data for integration with disparate sources (including those outside the enterprise) and types (including unstructured data).
Storylines
The efficacy of discovery tools such as interactive visualizations and dashboards has helped to shift the focus of self-service BI from numeric representations to graphic images. 2015 will see the natural progression of this shift from images to data-driven narratives that explicate analytics results.
Updating ETL
Traditionally, the ETL process of transforming and loading data into a warehouse for Business Intelligence was one of the most time consuming aspects of issuing reports. These complications are exacerbated by the size and velocity of Big Data; although it is possible to stream data into platforms with such as Hadoop for BI purposes, such options still do not account for the agility and variation of data types a business user requires for ad-hoc exploration for real-time decision making.
Automated Event Triggers
The semi-automated aspects of the data preparation of certain Data Discovery platforms is indicative of the growing trend towards automation in which analytics is utilized to effect action. In these cases—the most powerful of which is evinced through the numerous applications found in the Internet of Things (IoT)—the results of analytics triggers an event that can aid in the business process. Such automation can be as simple as issuing an alert via text or email that requires an employee to do something, or actually implementing that action electronically such as ceasing production on a cell phone tower when there is a default detected with it. In these cases, analytics takes a proactive approach that will increasingly bring value to the enterprise in the coming years.
The IoT
The IoT will influence the analytics landscape in several ways. Although its presence is currently most prominent in the management of equipment assets via the Industrial Internet, the aforementioned automation capabilities it facilitates place special demand on analytics to account for its enormous quantities of data via real time streaming. Architectural concerns include performing analytics near the source of the data, which enables the results of those analytics—and not the huge quantities of data—to go back to the enterprise or to a Cloud platform, as well as utilizing solutions that prioritize speed. The number of analytics solutions created expressly for the IoT and real-time analytics will continue to grow in 2015; a number of them will utilize the scalability of the Cloud.
Cloud Integration
The newfound integration prowess of self-service BI options will likely hinge upon the Cloud in several ways. Analytics pertaining to Big Data and the IoT will typically revolve around Cloud access in the coming year. One of the most discernible trends relating to BI and the Cloud is the increasing frequency at which the enterprise will utilize the Cloud for the analysis of data residing on premise. According to Tableau, “In 2015 companies will begin to choose the cloud when it makes sense for their business case, not only because the data is there.” One of the most salient reasons for doing so is to integrate on-premise data with those in the Cloud. Additionally, the Cloud provides an accessible, expedient way to aggregate data from external sources (such as news, competitor announcements, etc.) and to integrate them with that from internal sources. The transition to self-service BI will help the business user to do so with relative ease, granting him or her greater autonomy and utility from a widening body of data.
