Effectiveness of Agile Implementation Methods in Business Intelligence Projects from an End-user Perspective

Abstract The global Business Intelligence (BI) market grew by 10% in 2013 according to the Gartner Report. Today organizations require better use of data and analytics to support their business decisions. Internet power and business trend changes have provided a broad term for data analytics--Big Data. To be able to handle it and leverage a value of having access to Big Data, organizations have no other choice than to get proper systems implemented and working. However traditional methods are not efficient for changing business needs. The long time between project start and go-live causes a gap between initial solution blueprint and actual user requirements in the end of the project. This article presents the latest market trends in BI systems implementation by comparing Agile with traditional methods. It presents a case study provided in a large telecommunications company (20K employees) and the results of a pilot research provided in the three large companies: telecommunications, digital, and insurance. Both studies prove that Agile methods might be more effective in BI projects from an end-user perspective and give first results and added value in a much shorter time compared to a traditional approach. Keywords: Agile methods, Business Intelligence, efficiency, end-users needs, advanced analytics, sprint, and iteration. Introduction BI complexity and changing requirements represent the most difficult challenges facing applications. During the process of BI implementation multiple components must be considered from the very start such as data integration, cleansing, modelling, warehousing, metrics creation and management, reports, dashboards, queries, alerts, and many more (Cerqueira, 2015). This requires a clear vision of future needs and a very well defined strategy from project sponsors and end-users. Projects take a long time to implement and their effects can be visible sometimes only after a few years (Kernochan, 2011). Today organizations require BI solutions more than they needed them in previous years and decades. Due to rapid market changes, organizations need to adapt to the new environment properly if they do not want to stay behind their competitors. This situation impacts users' requirements for data and reports. Thus BI projects final products are often found useless due to organizational needs that have changed during the time of project design and implementation (Eckerson, 2007a, 2007b; Marjanovic, 2011). Business cannot longer afford empty investments and needs to have quick benefits and an acceptable payback on the selected BI technology (Oxford Economics, 2015). Traditional methods of BI implementation are no longer efficient. An overly lengthy timeline, the inability to request timely changes that usually occur only at the end of the project, and overly complex approaches do not allow meeting customer targets (Vijaya, 2013). Agile methods brought a new view to a project delivery. It proves that success can be achieved more quickly by delivery of actual product in iteration. In this article effectiveness is measured from the added value brought by BI in a short time (less than 6 months), namely return on investment achieved after the first BI benefits appear and by meeting end-users' requirements. This article presents some initial research in order to answer the question, "Is Agile more efficient in BI implementation compared to traditional methods?" Agile vs. Traditional Implementation Approach For a better understanding of Agile methods for Business Intelligence (BI) system implementation projects, it is worth to compare Agile with the traditional / waterfall approach first. Agile methods of implementation require a change of thinking and a different approach compared to traditional waterfall methods. Traditional methods concentrate on project scope using them to determine cost and time schedule. …