Big Data in Retail Big Data Analytics Central to Customer Acquisition and Retention Strategies in Retail MAA7-67 September 2014
Contents Section Slide Numbers Executive Summary 4 Methodology 6 Fundamentals 8 Big Data Analytics Value Chain 13 State of the European Market for Big Data in Retail 19 Applying Big Data in Retail 28 Trends & Opportunities 33 Challenges 43 Outlook 51 The Frost & Sullivan Story 54 MAA7-67 3
Key Findings Big Data Opportunity Retail Need Retail Application Growth Potential Big Data Challenge Large volumes of unstructured and structured data originating from a variety of sources contains valuable insight about customer behaviour, which has the ability to contribute to the growth of retail businesses. The presence of numerous retailers provides endless choices for customers, making it difficult to retain customers. Big Data analytics will help retailers understand what customers want and what needs to be offered to remain competitive in the market. Analytics will be applied for both front-end as well as back-end applications. While online and offline retail are two broad segments, the underlying aim of every retailer is to integrate data from multiple channels to provide an omni-channel experience for the customer. There is tremendous growth potential in European markets which is a mix of growing and nascent markets. While the United Kingdom, Germany and France are faster adopters, retailers in Southern and Eastern European countries are also realising the value of analytics. Apart from the technological and organisational challenges facing retailers, privacy and data regulations issues are particularly important in Europe. The European Commission stipulates adherence to specific regulations to avoid hurting customer sentiments. Source: Frost & Sullivan MAA7-67 5
Objective, Scope, and Methodology Objective This report will introduce the Big Data opportunity in the retail industry in Europe, throwing light on the opportunities for applications in this ever-expanding industry. It will provide an overview of the Big Data value chain, the competitive landscape, the trends driving the uptake of Big Data analytics among retailers, and the growth challenges in this market. Scope Big Data in the retail market includes a wide range of services such as hardware (server storage) and software providers (processing and analytics). This research service will focus on the market for analytics services. Though this market is global in nature, rates of adoption differ from one region to another. Additionally, some of the opportunities and challenges may have more impact in one region than another. This report will highlight these differentiating factors, providing an overview of the state of the market in Europe. Methodology This report will draw upon secondary research, which includes internal Frost & Sullivan databases and external insights on the topic. Primary research is composed of discussions with key global and European providers in the market and will be used to assess the state of the Big Data analytics market in the retail industry. Source: Frost & Sullivan MAA7-67 7
Big Data Analytics Going Beyond Business Intelligence The term Big Data analytics needs to be differentiated from Business Intelligence (BI). BI solutions have been in existence for many years and are focused on transforming historical data into patterns and derive insights for analysis and reporting purposes. Cognos* and Business Objects** are key examples of vendors of BI solutions that are commonly used by organisations today. Key examples of BI solutions are traditionally designed to rely on relational data, which are mainly structured. Such data are aggregated in different ways based on predesigned aggregation and statistical programs. BI solutions are unable to process high-speed, unstructured data from online sources. BI lacks the ability to forecast and deliver predictive insights from large, varied and rapidly changing data sets. This is where Big Data analytics solutions come into the picture. *Acquired by IBM in 2007. **Acquired by SAP AG in 2007. Source: Frost & Sullivan MAA7-67 10
Volume, Variety, and Velocity Marking The Transition To Big Data-Driven Business Intelligence Key Takeaway: Big Data is most commonly seen as defined by large volume, high variety, and high velocity. Big Data in Retail Market: Defining Big Data, Global, 2013 Real time Velocity Batch POS data Data Warehouse Structured Location data In-memory analytics Sensor data Payment data Shipment data Transaction history data Weather Customer profiles HR records Financial data Volume & Variety Google+ Twitter Call centre Online forums SharePoint Text documents Facebook Text messages Big Data Clicks Hadoop/MapReduce Video Environmental data Unstructured Note: Booz & Company was renamed as Strategy& after its acquisition by PwC in 2014. Source: Booz & Company; Frost & Sullivan MAA7-67 11
Generic Approaches To Big Data Projects Top-down approach o The most common approach to planning Big Data projects starts from the definition of a business problem, followed by the identification of data required to solve it. o Following this traditional business intelligence approach, benefits of Big Data projects are often only incremental. Bottom-up approach o Under this approach, the starting point is available (internal and external) data, which is searched for patterns to allow insights to emerge. Source: Booz & Company; Frost & Sullivan MAA7-67 12