How Location Intelligence drives value in next-generation customer loyalty programs in Retail business



Similar documents
August 2013 Rising to the Omni-Channel Challenge

The big data revolution

SOURCE ID RFID Technologies and Data Capture Solutions

EXECUTIVE SUMMARY. Warehouse Management Systems. Technology for Global Supply Chain Performance. Authored By: Carla Reed. ChainLink Technology Series

ORACLE SALES ANALYTICS

Agile Manufacturing for ALUMINIUM SMELTERS

THE FUTURE OF RETAIL HAS ARRIVED. RAZORFISH AND ADOBE PRESENT RAZORSHOP: A SEAMLESS, DIGITALLY ENABLED IN-STORE CONTENT AND COMMERCE EXPERIENCE.

From Brand Management to Global Business Management in Market-Driven Companies *

PIVOTAL CRM RETAIL INDUSTRY

Contents WHITE PAPER. Introduction

Mind Commerce. Commerce Publishing v3122/ Publisher Sample

Data Ownership Overview: Using omni-channel data to connect one-on-one with customers

BUY BIG DATA IN RETAIL

Architectures for massive data management

Top 10 Factors That Will Increase Conversion Rates

BIG DATA & DATA SCIENCE

Maximizing Returns through Advanced Analytics in Transportation

Trackunit Telematics Solution. for OEM

Improving The Retail Experience Through Fast Data

Introduction. External Document 2015 Infosys Limited

1Current. Today distribution channels to the public have. situation and problems

SMARTPHONES & BIG DATA. Daniel Nelson Head of Enterprise Development, daniel.nelson@braintreepayments.

Enterprise Resource Planning Analysis of Business Intelligence & Emergence of Mining Objects

Product Overview.

THE INTERNET OF THINGS. HARMONIZING IoT FOR RETAIL. Kuru Subramaniam

Accenture and Oracle: Leading the IoT Revolution

Five Ways Retailers Can Profit from Customer Intelligence

FORRESTER CONSULTING INTERNET OF THINGS SURVEY - KEY FINDINGS. Building Value from Visibility: 2012 Enterprise Internet of Things Adoption Outlook

Retail Analytics The perfect business enhancement. Gain profit, control margin abrasion & grow customer loyalty

Accenture Business Intelligence for Fashion and Luxury. Creating a Differentiated Customer Experience for Long-term Brand Loyalty

In the pursuit of becoming smart

Resource Library. Consumer Location-Based Analytics Deliver Actionable Insights. From Platt Retail Institute s. Bringing Research to Retail SM

Food and Beverage. Microsoft Dynamics NAV Solutions for Food and Beverage Companies

T r a n s f o r m i ng Manufacturing w ith the I n t e r n e t o f Things

ADVANTAGE YOU. Be more. Do more. With Infosys and Microsoft on your side!

Data Mining: Benefits for business.

The retailers guide to data discovery

One Solution for all of your Global Financial Reporting Issues By Nolan Business Solutions

mobile commerce challenge or opportunity?

From Big Data to Smart Data How to improve public transport through modelling and simulation.

Converged, Real-time Analytics Enabling Faster Decision Making and New Business Opportunities

Enterprise Resource Planning

Is Your Business Suffering? The complexity of connecting disparate systems? POS breakdowns in peak times? Poor sales promotion management? Purchases b

Kingdom Big Data & Analytics Summit 28 FEB 1 March 2016 Agenda MASTERCLASS A 28 Feb 2016

Google Analytics tags migration to Google Tag Manager in a multi-site environment Monday, 28 September :55

Advanced Analytics. The Way Forward for Businesses. Dr. Sujatha R Upadhyaya

Big data in freight transport

Shell CRM October 2014

Smarter Analytics. Barbara Cain. Driving Value from Big Data

Digital Business Services Topic Area Theaters May 17-19, 2016 Orlando, FL

WHITE PAPER Analytics for digital retail

Monitoring the Online Marketplace

An Implementation of Active Data Technology

How To Understand The Power Of The Internet Of Things

HOW TO TURN 9 RETAIL IT CHALLENGES INTO 9 BUSINESS OPPORTUNITIES

Cisco Context-Aware Mobility Solution: Put Your Assets in Motion

Click Labs Report: How Retail Leverages Mobile

Challenges for Big Data Applications in Japan: Hopes and Concerns

How To Create A Retail Analytics Platform With Tapway

Using Predictive Analytics to Increase Profitability Part III

Talousjohto muutosagenttina ja informaatiotulvan tulkkina

Differentiate Now for Retail Leadership The Omni Channel Customer Experience

Social media has changed the world as we know it by connecting people, ideas and products across the globe.

Integrated Sales and Operations Business Planning for Chemicals

Supply Chain Optimization for Logistics Service Providers. White Paper

It costs 5 to 10 more times to acquire a new customer than to retain an existing one (Inc)

Supply Chains: From Inside-Out to Outside-In

Data Analytics in the Logistics Sector In the slipstream or the fast lane?

Inventory Routing. An advanced solution for demand forecasting, stock replenishment, and route planning and execution

The Power of Relationships

COMP9321 Web Application Engineering

Unified Communications Solution for Retail Industry

Sage 300 Finance. Sage 300 Finance. Industry Solution. Generic to all Industries and Organisations. Target. Business Processes. Business Challenges

Lexmark Enterprise Software. Transforming customer engagement

Turn Your Business Vision into Reality with Microsoft Dynamics NAV. icepts Technology Group, Inc. Dynamics NAV Gold ERP Partner

White Paper. Retail Made Personal. Make the shopping experience personal, relevant, and profitable

media kit 2014 Advertise Global Mobile Ad Network

AirTight Social Wi-Fi and Analytics for the Retail Store of the Future Where Clicks Meet the Bricks

The Internet of Everything: The Next Industrial Revolution

Business Process Services. White Paper. Leveraging the Internet of Things and Analytics for Smart Energy Management

The changing role of the IT department in a cloud-based world. Vodafone Power to you

30 Ways To Do Real-Time Personalization

StratioDeep. An integration layer between Cassandra and Spark. Álvaro Agea Herradón Antonio Alcocer Falcón

I D C M a r k e t S c a p e : W o r l d w i d e B u s i n e s s A n a l y t i c s B P O S e r v i c e s V e n d o r A n a l y s i s

BUILDING OMNI-CHANNEL RETAIL FROM THE BACK END UP

Demonstration of SAP Predictive Analysis 1.0, consumption from SAP BI clients and best practices

[ know me ] A Strategic Approach to Customer Engagement Optimisation

Sales Management 101, Conducting Powerful Sales Review Meetings

Transcription:

How Location Intelligence drives value in next-generation customer loyalty programs in Retail business Andreas Nemeth a, Mikhail M. Komarov a a National Research University Higher School of Economics, Myasnitskaya street, 20, Moscow, 10100, Russia Abstract Nowadays smartphone technology offers a variety of built-in sensors and offers companies new possibilities to fetch additional and valuable data. Since companies have already begun to digitally transform their existing customer loyalty programs into mobile applications, new approaches and possibilities are resulting. We assume that that current and existing loyalty programs do not utilities their full potential and miss out the chance to apply an additional dimension to their business intelligence systems. By collecting and processing the mobility patterns of every single customer combined with the delivery of appropriate offers and incentives at the right time and the right place, may form the future s customer approach. The processing of the customers location data may reveal new information and insights from the customer buying behaviour. Hence, by the use of location prediction we are able to determine the customers next location. In this paper we propose a system to provide customers incentives in order to change their purchase behaviour. Based on location prediction methods combined with data from internal sources as well as external sources (e.g. weather), we describe three use cases of customer loyalty concept driven by context-aware location intelligence. Furthermore, a theoretical application of GraphX s property graph as a location graph model has been defined in order to provide a basis for future location intelligence. Introduction The underlying motivation for this paper is the assumption that current and existing loyalty programs do not utilities their full potential. Over the last years, mobile applications have found their ways into the customer relationship market. While smart devices provide more potential as already used, mobile applications using the full potential of smart devices sensors have not reached the retail market. Existing mobile applications only digitally transformed the well-known loyalty card format enhanced by additional data such as locations of the retail stores or information about current offers that are valid in every single retail store within the store chain. By collecting and processing the mobility patterns of every single customer combined with the delivery of appropriate offers at the right time and the right place, may form the future s customer approach. Today s retail business underlies fluctuation in the flow of customers and needs new methods in order to balance the store utilization and optimise the customers purchases. Nowadays we can already track the customers location with the aid of different techniques [5] and by predicting the customers future movement in combination with the customers data of previous purchases [3], we want to reach the customer at the right time, the right place, and with the right incentive in order to influence the customers purchasing behaviour. This paper is about the future perspective in retail business in terms of enhancement of existing or new loyalty programs with profound location intelligence focused on location prediction. With the help of the prediction of the customers next location, historic purchases and stimulation of the customers behaviour by providing attractive and appropriate incentives at the right time and at the right place, we analyse theoretical benefits of future s customer loyalty programs.

Location data Location data works well in combination with humans; therefore, mobile providers have collected and analysed certain information for years to improve their products and the customers satisfaction, although data are used nowadays for other purposes and are gathered by third parties. Especially, some application - installed on smart phones - are gathering location based data, despite the actual application does not need of these data. Although other applications such as Foursquare work indeed with location data and allow users to check-in at their favourite places and thus may receive incentives, recommendations and other location related services. The gathering of the single users location data is developing into a powerful resource and ability. On the one hand every single user may be targeted with appropriate advertisements, which refers to the current or predicted locations. Furthermore, data may be considered at large to reveal bigger developments. With the help of these data e.g. traffic congestion, due to the tracking of amount and speed of mobile phones on the road [4] Mining individual mobility patterns is an emergent research field. Through the wide distribution of mobile phones, a wide range of mobility-related data is constantly tracked, for example the individual s communications via phone calls and messages [3] The individuals GPS data provides the most sufficient resource for mobility traces, although a complete location management system works best by combining all different approaches and cannot perform its full potential by using one resource of data. Nevertheless, the increasing number of such devices leads also to a better understanding of mobility patterns in general. Current research has shown, that the individual mobility is highly regular in the individuals mobility, ranging from few square kilometres to thousands square kilometres. Most individuals visit a small number of locations very often. Besides that, after taking into account the individuals idiosyncrasies, the probability to forecast that an individual visits a new location or returns to an already visited location is very high. Individuals and the society benefit both from the study of mining individuals mobility patterns. A wide range of researches already cover this field, for example, computational sociology, wireless network analysis, etc. Furthermore, the understanding of the individuals mobility patterns recognizes exceptional activities and events, predict future movements and directions [3]. Digital Transformation in Retail Business As Forrester [1] stated, the retail commerce underlies a digital transformation which will drive the application of real-time analytics in physical stores as well as the customer engagement. By improving operations within the store, in-store analytics and mobile devices are enabling the digital store of the future. Major drivers are data sources such as online behaviour, in-store analytics, supply chain and labour planning. Digital Signage, the integration of third party platforms (e.g. Google and Facebook) as well as Internet of things platforms will find their way into the digital transformation of the future s retail commerce. By the usage of sensor technology such RFID, GPS, Bluetooth and low energy beacons, an entire new dimension of data will be added to the today s CRM systems. Furthermore, location intelligence is able to bring insights of the entire customer journey in the real world and may transform the future customer experience in retail [2]. Since the late 1990s, loyalty programs have drawn a huge dimension of attention within the retailing market. The common assumption is that the care of existing customers is less expensive than the acquisition of new customers. Since ever then the majority of companies and businesses implemented customer loyalty programs. Until 1999 approximately 350 million loyalty cards were given to customers within the retailing sector. Customers who are entering a loyalty program usually are expected to buy more from the respective business. Hence, loyalty programs are an important part of the customer

relationship management that is utilized by many marketers in order to recognize, reward and retain their customer base. The benefits of loyalty programs have not changed over the last decades and are still irrevocable means to approach customers. Loyalty programs represent a critical component of customer relationship strategies in companies. While the concept of loyalty programs is easy to understand, companies are facing challenges according to their implementations. Many companies experience that their programs do not create the loyalty as anticipated and have to address the critical areas in order to match the desired objectives with the current reality [6]. Big data analytics offers businesses new perspectives towards a 360- degree view of the customer. Already existing loyalty programs with mobile application support can be extended by tracking the customers movement and offer him/her certain incentives and offers for the products they have bought in the past, if they will go a different store within a chain. These analytics have to be done in real time to achieve a real impact. Therefore, we need to identify internal and external data sources for this application: The current amount of customers within the stores and the optimal amount of customers according to analytics of past turnover and off peak times. Previous purchases of customers extracted from the ERP system in order to offer the customer an appropriate discount or a special price for previously bought products. The current and predicted location of the customers provided by the mobile loyalty program application and processed by a graph analytical system. Access to additional external context data e.g. weather information The data has to be processed in near real time and can also be considered as fast changing if we take into account the current amount of customers in a store as well as the movement of the customers. One of the main benefits behind this approach is the utilization of different stores within the chain in order to balance the flow of customers in each store. Location Intelligence as a Graph representation in GraphX Systems that deliver insights-driven action at scale must support data and analytics delivered as batch processes and real-time insights. Spatial data and analytics provide unique customer insights in every phase of the engagement life cycle. Based on a survey conducted by Forrester [1], 1805 global data and technology decision-makers have been asked, if they plan to implement and expand their use of location analytics. 46% responded, that they are already expanding and implementing, while 22% answered, that they are planning to implement location analytics within the next 12 months. GraphX represents the Spark API for graphs and graph-parallel computation. By extending Spark RDD abstraction and introducing the property graph, allows to create directed multigraphs, where vertices and edges are stored in 2 Resilient Distributed Dataset (RDD). RDD is the basic abstraction in Spark. RDDs represent an immutable, partitioned collection of elements that are operable in parallel. Hence, each vertex and edge can consist of properties. Furthermore, GraphX supports a variety of graph computation algorithms as well as operators and by that, it helps to simplify the analytical tasks. The property graph is defined as directed multigraph due to the fact, that multiple edges can share the same vertex as source and destination. Each vertex can be uniquely identified by a 64-bit key, which is further known as VertexID, while attached properties are stored as Scala or Java objects in each vertex or edge. Counter wise, each edge consists of the corresponding source and destination identifiers [7,8].

Figure 1. Graph model The proposed location graph data model as theoretical application of GraphX s property graph satisfies the needs for common analytics in location intelligence as well as the application of location prediction methods. Our simple location graph model consists of stay points and stay links: Stay points represent locations (e.g. super market, home, work place, etc.) where the user is located over stay time defined by a threshold value. Furthermore, a change in the transportation type (e.g. from metro to walking) also triggers the creation of a new stay point in the Simple Location Graph Model. A stay point stores at least a unique identifier key, latitude, longitude and radius value. Optionally, the selected context attributes are stored in connected visit vertices as an addition. Stay links represent the edge between two stay points. A stay link stores at least the id of the source stay point, the id of the destination stay point and the average travel time. Optionally, the selected attributes (e.g. movement data such as acceleration, therefore, the type of transportation; utilization of the respective stay link) can be stored additionally. Furthermore, way points and way links are additional entities used by our Advanced Location Graph Model and are embedded in the stay links between two stay points. This approach allows a higher granularity and analysability of a given customer journey between two stay points on demand. Benefits of location prediction in customer loyalty programs The system in a nutshell may be described as followed: The system shall influence the customers behaviour with the help of the prediction of the customers next location, their historic purchases, in-store utilization and purchase stimulation by providing attractive and appropriate incentives at the right time and at the right place. Current approaches in location intelligence focus mainly on analysing location data without the application of location prediction methods. As followed we will discuss the use cases of using location prediction methods in combination of internal (e.g. CRM) and external context data:

Figure 2. Use case 1 As shown in the Figure 2, it states a possible scenario of a sequence of visited locations enhanced by context information. Moreover, the location data is connected with the contextawareness of a time dimension (day of week) and context information of the location (e.g. sunny weather.). After the user has visited the third stay point, the system already compares the sequence with the historic location graph and predicts the visit of Retail store Beta. The system predicts a store utilization of 107 percent and sends an offer to the user which is only valid in Retail store Alpha with a store utilization of 75 percent. The Retail store Beta is in an acceptable range of the predicted store, hence a significant increase of effort on the customer side may be prevented and the retail chain is able to outbalance their stores utilization.

Figure 3. Use case 2 The Figure 3 shows an example of a different application of the proposed system. While the scenario takes the store utilization under consideration, scenario 2 focuses on the inventory of the user s usually-bought product inventory, hence the temporary unavailability of perspective products. As use case 2 shows, the users most probable sequence of stay points is predicted upon the historic location graph. The user receives an offer for Retail store Alpha due to the fact that a usually-bought product is out of stock in Retail store Beta. In this case the movement to the store of a competitor may be prevented, as the customer remains satisfied.

Figure 4. Use case 3 In Figure 4, another scenario is shown, in this scenario the customer is intended to visit a competitor s store that is predicted by the application of the user s historic location graph. In this scenario the customer visits the locations, as a sequence of specific stay points in the mentioned order. Based on his/her predicted sequence of stay points, a visit of a competitor as the customer s next location is very probable. On the way to the competitor s store the customer receives an offer valid in Retail store Alpha and decides to claim the received offer. Instead of going to the competitor s store, the customer goes to Retail store Alpha and goes afterwards home. Conclusion The current technologies offer a wide variety of possibilities to track the individuals mobility, although the current battery technology does not provide the needed performance and endurance. Existing loyalty programs may be enhanced by location data provided by the customers smart devices. However, the proposed concept is a combination of location data processing, location prediction, enhanced by data deriving from a system such as ERP and CRM systems that are already used by companies. In order to implement the data of several systems has to be orchestrated and processed in near real-time. By implementing this concept, new and existing loyalty program applications can leverage the customers location data throughout the entire customer journey. Moreover, the analysis customers location data adds a new dimension to the existing customer profiles in CRM systems, while the customers themselves benefit from the proposed system as well. Although we have to consider the impacts and consequences of abuse of gathered data and therefore knowledge that may be the result. Companies and Governments have to handle

these data carefully, additionally regulations shall also be considered since data security and data ownership are important topics nowadays. References 1.Forrester. (2016). Drive Intelligent Customer Interactions With Spatial Analytics. Forrester. 2.Forrester. (2015). Predictions 2016: The Digital Store Engagement Surprise. Forrester. 3.Lin, M., & Hsu, W.-J. (2013). Mining GPS data for mobility patterns: A survey. Singapore: Elsevier. 4.Mayer-Schönerbeger, V., & Kenneth, C. (2013). Big Data - Die Revolution, die unser Leben verändern wird. Munich: Redline Verlag. 5.Pirozmand, P., Guowei, W., Behrouz, J., & Feng, X. (2014). Human mobility in opportunistic networks: Characteristics, models and prediction methods. China: Elsevier. 6.Reinartz, W. J. (2010). Understanding Customer Loyalty Programs. Retailing in the 21st Century: Current and Future Trends. 7.UC Berkeley AMPLab. (2014). Graph Analytics with GraphX. Retrieved 03 02, 2016, from Graph Analytics with GraphX: http://ampcamp.berkeley.edu/big-data-mini-course/graphanalytics-with-graphx.html 8.Xin, R. S., Crankshaw, D., Dave, A., Gonzalez, J. E., Franklin, M. J., & Stoica, I. (2014). GraphX: Unifying Data-Parallel and Graph-Parallel Analytics. Berkeley: UC Berkeley AMPLab.