Big Data Retail Analysis and Product Distribution (BREAD) Model for Sales Prediction

UAE retail sector is expected to grow by 5% every year and an estimated retail market to be AED 200 billion (Emirates247, 2016), which makes UAE ranked 7th in Global Retail Development Index. Euromonitor and ATKearney (Hana Ben-Shabat, 2015) study about UAE retail market, opinions that consumer confidence in UAE has not negatively wedged despite slow economic growth, rather, unexpectedly, resulted in pierce retail competition. In a state of such penetrating race, analytics can be a foremost differentiator for the companies. (Steve Lavalle, 2010) found that the top performing companies are three times more effective than those without analytics, making analytics as a sole competitive differentiator. Analytics lashes the passage from merchant-driven business models to digital models, where every decision is cognizant by data. Another study conducted by (Brynjolfsson, 2011) reveals that data-driven businesses have an output and productivity of 5-6% higher than similar companies who do not use data-driven decisions. (Luckie, 2012) state that poor data management decisions can cost up to 35% of a businesses operating revenue. In this study, a new technique called BREAD (Big Data Retail Analytics and Product Distribution) is developed for sales prediction for retail scenario. As an experiment, the model was used for sales prediction of ABC Stores (name changed, as requested), based on their 2015 sales data for 16 Item types, divided into 1559 items across 10 stores (a total of 8523 records). Based on the study, the paper gives recommendations to ABC Stores to embed analytics in its predictions.