Page 1
Page 1
A MEMBER OF THE UNIVERSITY OF HONG KONG GROUP
香 港 大 學 機 構 成 員
AILIE TANG LAWRENCE MUI
BIG DATA:
REVOLUTIONIZING THE RETAIL INDUSTRY
Did you ever receive a coupon offering you an extra 20% discount for the pair of shoes that
you wanted to buy while shopping in a department store? Or did you ever wonder why the
salesperson serving you could always do the right trick for cross-selling1 the complementary
items that you were most interested in? Kroger, a grocery store chain in the US, excelled in
drawing a response rate of over 70% within six weeks of mailing “snowflake” coupons to its
customers. The magic lied in the use of Big Data for Kroger and other retailers to analyze its
customers’ shopping history and personalize the offers for the members on its loyalty power
of information.
According to a McKinsey report, Big Data was defined as “datasets whose size is beyond the
ability of typical database software tools to capture, store, manage, and analyze”.2 With Big
Data, management could leverage the use of information to derive insights and make better
decisions on which direction the enterprise should sail through the waves of changes ahead.
With the rising popularity of Big Data in different industries, especially the retail, finance and
telecommunication industries, the International Data Corporation (IDC) predicted that the
global Big Data and analytics market would reach US$125 billion in hardware, software, and
services revenue in 2015.3
This case study aimed to take you on a journey to explore how Big Data was going to
revolutionize the retail industry in the coming years. Light would first be shed on the general
ideas of the Big Data concept and its differentiations from data mining. The notable 4Vs of
Big Data would then be introduced, followed by the benefits that companies would receive
when tapping into the power of information. Impacts of Big Data on customers would also be
highlighted. Challenges confronted by enterprises when making use of Big Data would then
be discussed, followed by recommendations for streamlining the implementation of Big Data.
1 Cross-selling refers to the action or practice of selling an additional product or service to an existing customer.
2 McKinsey Global Institute (2011) “Big Data: The next frontier for innovation, competition, and productivity”
3 Framingham, M. (2 December 2014) “IDC Predicts the 3rd Platform Will Bring Innovation, Growth, and Disruption Across All
Industries in 2015”, International Data Corporation, http://www.idc.com/getdoc.jsp?containerId=prUS25285614 (accessed on
30 June 2015)
To order this case, please contact Centennial College, c/o Case Research Centre, Centennial College, Wah Lam Path, Pokfulam,
Hong Kong; website: http://cases.centennialcollege.hku.hk .
This case was fully supported by a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (UGC/IDS12/14).
© 2015 by Centennial College, a member of The University of Hong Kong group. No part of this copyrighted publication may be
reproduced or transmitted, in whole or part, in any form or by any means, whether electronic, mechanical, photocopying, recording, web-based or otherwise, without the prior permission of Centennial College.
This case was prepared for class discussion purposes and is not intended to demonstrate how business decisions or other
processes are to be handled. Ref. 15/010C Published: 18 November 2015
Page 2
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 2
What was Big Data?
Big Data referred to “the dynamic, large, and disparate volumes of data being created by
people, tools, and machines”.4 Big Data of a company could be garnered from internal data
(e.g. sales transactions, customer details, inventory records), structured external data (e.g.
social profile datasets on Facebook, partner data, customer loyalty program), and unstructured
external data (e.g. climate data, news, mobile phone geo-location data) [see Exhibit 1 for
examples of exploitable structured and unstructured data]. When put together and analyzed
appropriately, the data could present the decision makers both the historical and forward-
looking views of the enterprise or the industry as a whole.
The meteoric rise of Big Data in the business world was driven by interest in finding
correlations across multiple and disparate data sources, envisaging customer behavior,
forecasting product or service sales, predicting fraud or financial risk and analyzing social
network comments for consumer sentiment [see Exhibit 2 for factors driving companies to
conduct Big Data analyses for their businesses].5 The effective use of Big Data thus became
vital to companies maintaining its competitiveness by responding to the industrial changes in
a time-efficient manner.
Differences between Data Mining and Big Data
Data mining and Big Data are different but inseparable. Gartner, an American IT research and
advisory firm, defined data mining as “the process of discovering meaningful correlations,
patterns and trends by sifting through large amounts of data stored in repositories”.6 Data
mining could be regarded as the mechanisms or algorithms for drawing useful insights from
large datasets, usually with the use of analytics software. Since decision makers only wanted
to use certain pieces of information from a vast amount of data, data miners would need to go
through the entire datasets manually (or automated with the use of software) to look for
relevant threads of information. The data mining process thus involved complicated
information search which would yield targeted and specific search results.7 After relevant data
was extracted from a large dataset, it would be analyzed from different perspectives to find
any correlations or patterns hidden among the data.8 Decision makers of a company would
then be able to identify any relationships between operational variables (such as pricing
strategy and customer demographics) and business outcomes (such as the level of customer
satisfaction and corporate profit).
Big Data, on the other hand, referred to the aforementioned large datasets containing high-
volume, high-variety and high-velocity and high-veracity information assets.9 It was from
these information assets that relevant data would be extracted for further examination and
handling during the data mining process. In other words, Big Data was characterized by 4Vs:
Volume, Variety, Velocity and Veracity.10
4 Ernst & Young (April 2014) “Big Data – Changing the way business compete and operate”
5 Information Week (2015) “2015 Analytics & BI survey”
6 Gartner (2013) “IT glossary data mining”, http://www.gartner.com/it-glossary/data-mining (accessed on 30 June 2015)
7 UCLA Anderson School of Management, “Data mining: What is data mining?”
http://www.anderson.ucla.edu/faculty/jason.frand/teacher/technologies/palace/datamining.htm (accessed on 3 August 2015) 8 Research Gate (2013) http://www.researchgate.net/post/What_is_the_difference_between_machine_learning_and_data_mining
(accessed on 30 June 2015) 9 Gartner (2013) “IT Glossary Big Data”, http://www.gartner.com/it-glossary/big-data/ (accessed on 30 June 2015)
10 Ernst & Young (April 2014) “Big Data – Changing the way business compete and operate”
Page 3
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 3
The 4Vs of Big Data
The first V, depicting the salient attributes of Big Data - Volume, referred to an extremely
large amount of data being generated that they could not be processed with the use of some
common software tools within a tolerable period of time.11 Since Big Data assets were too
large to be processed or handled by traditional tools like Microsoft Excel, they demanded
“cost-effective and innovative forms of information processing for enhanced insight and
decision making”.12 According to IBM, over 90% of data in the world was generated in the
past two years, and that 2.5 quintillion13 bytes of data were created every day.14 A forward-
looking figure released by Frost & Sullivan, a business growth consulting firm, showed that
the global data traffic would exceed 100 zettabytes15 annually by the year of 2025 [see
Exhibit 3 for the actual and projected amounts of unstructured and structured data generated
from 2006 to 2020].16 The IDC further added that the size of data on the planet doubled every
18-24 months.17 What might be considered as Big Data today might not be qualified as big
enough in the future.
Many other factors came to cause the data volume to skyrocket in recent years. For example,
more business activities and transactions were being carried out nowadays, data was
streaming in from various social media channels, and more sensors were being installed than
ever before. However, with decreasing data storage costs, the major concern now lay in how
to extract relevant data from such large datasets, and use Big Data analytics to create value
from the extracted data.18
The second V went to Variety which meant that data came in all types of format and was
being created by machines as well as human beings.19 Data was thus heterogeneous in nature,
and could largely be categorized as structured and unstructured ones. Structured data referred
to information contained within specific fields,20 such as data in spreadsheets or traditional
databases, whereas unstructured data did not have specific formats,21 and could exist as audio
messages, videos, photos, social media communications, or e-mails. Indeed, only 5% of the
world’s data was structured, whereas the remaining majority of 95% was unstructured.22 The
non-standard format of such a large volume of unstructured data thus posed difficulties to data
scientists in storing, mining, and analyzing the data.
The third V was Velocity, stating that data was generated round-the-clock at an
unprecedented speed, and must be dealt with in (near) real time.23 Enterprises were often
overwhelmed by the waves of information that kept coming to them.24 They needed to deal
with these torrents of data immediately for the sake of getting the most updated pictures of the
11
Ibid. 12
Ibid. 13
Quintillion = 1018
14
IBM, “Bringing Big Data to the enterprise”, http://www-01.ibm.com/software/data/bigdata/what-is-big-data.html (accessed on
30 June 2015) 15
Zettabyte = 1021
bytes 16
Tare, N. (December 2014) “Outsourcing IT Resource Management to Third-Party Providers: Freeing Transportation
Organizations to Focus on Their Core Business”, Frost & Sullivan 17
Srinivasan, N. et al, (2012) “Harnessing the Power of Big Data Big Opportunities for Retailers to Win Customers”, Infosys 18
The Mckinsey Global Institute (2011) “Big Data: The next frontier for innovation, competition, and productivity” 19
Ernst & Young (April 2014) “Big Data – Changing the way business compete and operate” 20
Berkeley School of Information (26 February 2014) “Bridging the Divide between Unstructured and Structured Data”,
http://datascience.berkeley.edu/structured-unstructured-data/ (accessed on 30 June 2015) 21
Ibid. 22
ComplexWorld (5 September 2013) “What is Data Science?”,
http://www.complexworld.eu/datascienceworkshop/data-science/ (accessed on 30 June 2015) 23
Ernst & Young (April 2014) “Big Data – Changing the way business compete and operate” 24
IBM (2015) “Bringing Big Data to the enterprise”, http://www-01.ibm.com/software/sg/data/bigdata/ (accessed on 3 August
2015)
Page 4
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 4
industry and their businesses.25 Since data was perishable in nature and would lose their value
if not processed in a time-efficient manner, Big Data technologies thus came into play to
instantly process and analyze the massive volumes of heterogeneous data available in an
instant. This allowed enterprises to make timely and customized offers to individual
customers for driving sales.26 For instance, data streaming in from smartphones of in-store
customers could provide useful insights about the customers, including their demographics
and past purchasing habits, for the retailers. The instant data should then be analyzed
immediately to bring customer value,27 such as instructing the salesperson how to approach
individual customers with certain products or offers that would have a high chance of closing
the deals while the customers were in the store.
Veracity was the last V for Big Data. It referred to the bias and noise in the dataset.28
According to IBM, the IT Big Brother that first coined this fourth V for Big Data,29 veracity
also represented the unreliability inherent in the dataset.30 For instance, data of social media
mentioned, though contained invaluable information of customer sentiment or opinions about
the company, was indeed imprecise and uncertain in nature that it included many biased
comments and judgements. Since companies only wanted to store and mine the data relevant
to solving the problems, data scientists would need to be extremely careful by keeping those
biases and noise out of the dataset being analyzed.31 Leaving too much irrelevant data in the
dataset would contaminate the pictures that companies would get.
Uses of Big Data
The power of information could be tapped to serve a number of purposes for retailers. The
technological tool could help retailers customize their products and personalize their services,
cross-sell complementary products to customers, optimize business operations, and integrate
their online and offline businesses.
Product Customization and Service Personalization
One of the most common uses of Big Data in the retail industry was product customization
and service personalization. Big Data analytics could help retailers customize their offers
based on the browsing history and previous purchase records of their customers.32 A survey
conducted by Bain, a management consulting firm, showed that around 25% - 30% of the
1,000 online shopper respondents were interested in product customization that also enhanced
their loyalty and willingness to spend more money with the retailer.33
Big Data could help retailers understand more about the preferences and tastes of their
customers so that they could design products that would have a higher chance to become
25
Adshead, A. (April 2013) “Big Data storage: Defining Big Data and the type of storage it needs”, Computer Weekly,
http://www.computerweekly.com/podcast/Big-data-storage-Defining-big-data-and-the-type-of-storage-it-needs (accessed on 3
August 2015) 26
Roche, W. (2015) “Can Someone With Big Data Please Make Me A Personalized Offer?” , PointofSale.
http://pointofsale.com/201505076157/Point-of-Sale-News/Can-Someone-Please-Make-Me-A-Personalized-Offer.html (accessed on 3 August 2015) 27
FICO (2013) “Delivering Customer Value Faster With Big Data Analytics” 28
Ernst & Young (April 2014) “Big Data – Changing the way business compete and operate” 29
Hayes, B. (8 February 2013) “Big Data has Big Implications for Customer Experience Management”, IBM,
http://www.ibmbigdatahub.com/blog/big-data-has-big-implications-customer-experience-management (accessed on 30 June 2015) 30
IBM Institute for Business Value (2012) “Analytics: The real-world use of Big Data” 31
Fan, J. et al (Jun 2014) “Challenges of Big Data Analysis”, PMC, http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4236847/
(accessed on 3 August 2015) 32
Saracino, A. (23 April 2013) “Interesting Ways Businesses Use Big Data to Improve Personalization”, ClickZ,
http://www.clickz.com/clickz/column/2263262/interesting-ways-businesses-use-big-data-to-improve-personalization (accessed
on 3 August 2015) 33
Bain & Company (2013) Big Data: The Organizational Challenge
Page 5
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 5
popular among their target customers.34 Products customized in accordance with the specific
needs and taste of the users, such as styles and colors, would be far more appealing than some
generic offers. Sales, as well as customers’ loyalty to the brand, would increase with the level
of product customization and service personalization.35
Target was a successful Big Data advocate.36 The retailer first assigned each customer a
unique Guest ID, which was then linked to the customer’s credit cards, full name, and email
address(es). As data continued to accumulate, Target would then be able to create a database
with detailed profiles for each of its customers. Since the Guest ID program tracked customer
data such as purchase history, returns, website visits, and customer support, by correlating the
Guest ID database with its Target Baby Customer Registry, the retailer could then draw
insights on what kinds of products that pregnant women always bought. As a result, Target
was able to identify women who had a very high chance of being pregnant with special offers,
such as moisturizing lotion.37
Big Data analytics was of particular importance to online retailers in delivering a highly
personalized online shopping experience for their customers.38 Since in an online shopping
environment where buyers could easily compare the price and features of different offers,
product customization resulting from Big Data analytics could then serve as a tool for the
online retailer to break through the clutter and differentiate itself from the many others on the
World Wide Web. This could be done by bringing up personalized offers to website visitors
in real time by analyzing data previously gathered. For instance, Amazon claimed the throne
for customer service with Big Data-in-Personalization in the retail industry for many years.
The global online retailer had an abundance of Big Data resources: 152 million customers for
extracting useful business insights.39 Its Amazon Web Services (AWS) allowed the online
retailer to have the capability of collecting, storing, and processing data for business
analysis.40 With the AWS, should Amazon want to make purchase suggestions for its online
customers, it could simply use the relevant clickstream data and historical purchases stored in
the AWS system to arrive at appropriate product recommendations. Big Data analysis was
thus extremely crucial for Amazon to continue to take the driver’s seat of the industry.
How IBM helped SHOP.CA, Canada’s largest e-commerce marketplace, address some
common issues all online retailers face served as another example. The consulting giant
provided SHOP.CA with IBM LIVEmail, an email marketing system that could send
personalized emails with highly relevant offers and contents to SHOP.CA’s shoppers,
depending on their shopping behaviors. For instance, when a shopper left an item in their
shopping cart and left the website without checking out, the system would automatically send
the shopper an e-mail to remind them of the shopping cart’s contents. In terms of converting
browsers into purchasers, personalized emails were generated by IBM LIVEmail on the basis
of SHOP.CA shoppers’ browsing history. The emails showcasing products and offers that
could arouse the receivers’ interest were three to five times more likely to be clicked through
34
Mila (4 June 2015) “How to Use Big Data to Improve Customer Service”,
http://blog-en.mila.com/2015/06/04/how-to-use-big-data-to-improve-customer-service/ (accessed on 30 June 2015) 35
Bain Insights (11 May 2013) “Having It Their Way: The Big Opportunity In Personalized Products”, Forbes,
http://www.forbes.com/sites/baininsights/2013/11/05/having-it-their-way-the-big-opportunity-in-personalized-products/
(accessed on 3 August 2015) 36
Target Corporation is an American retailer that operates large retail stores primarily across North America that offer a
multitude of goods. It is the second-largest discount retailer in the United States, behind Walmart, and is a component of the S&P 500 Index. 37
Sidhu, I. (9 October 2014) Presentation on “Data driven business models”, UC Berkeley 38
Armonk (27 June 2013) “wehkamp.nl, Netherlands’ largest online retailer, Uses IBM Big Data Analytics Cloud Suite to
Deliver Improved Marketing and Merchandising”, IBM, https://www-03.ibm.com/press/uk/en/pressrelease/41447.wss (accessed on 3 August 2015) 39
PwC (2013) “Amazon – Translating Big Data to product opportunities”,
http://www.digitalinnovation.pwc.com.au/amazon-big-data/ (accessed on 30 June 2015) 40
Swensson, E. (December 2014) “Big Data Analytics Options on AWS”
Page 6
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 6
than standard emails, translating into a higher chance that these potential customers would
become real customers of the company.41
Vera Bradley, a women’s handbag and fashion accessory manufacturer, also used Big Data
analytics to improve the open rate, click through, and conversion of its promotional emails.
By sending more focused emails with offers chosen in accordance with individual consumers’
previous purchases, Vera Bradley found itself needed to send 63% less emails, while
generated a double email click-through rate and a 275% increase in conversion rate from
online shoppers to real customers.42
Cross-selling and Customer Satisfaction
Customer satisfaction was of critical significance to retaining existing customers and
developing a decent reputation for a retailer by word of mouth. Some marketing tactics, such
as cross-selling, could be used with Big Data to improve the shopping experience and
satisfaction level of a retailer’s customers. Big Data analytics could play a vital role in the
implementation of cross-selling that referred to the practice of recommending related products
or services to a customer considering buying something. 43 Put differently, cross-selling
encouraged existing customers to purchase additional or complementary products from the
seller’s assortment.44
For online retailers, Big Data could also help identify profitable cross-selling opportunities.
Since most e-commerce conversion rates fell between 2% and 3% (9.6% for the case of
Amazon),45 cross-selling seemed to be quite useful for bringing up the average order value of
a company, showing high-margin products to customers, or enhancing customer satisfaction.
With the “Frequently Bought Together” and “Customers Who Bought This Item Also Bought”
sections on its online selling platform promoting related products to customers, Amazon
attributed up to 35% of its revenue to cross-selling.46
With the art of cross-selling polished by Big Data analytics, customers would be more
satisfied with the retailer’s offerings.47 Big Data analytics could then be used to monitor
consumer sentiment for identifying any adjustments needed for the products and services
offered by a retailer.48 The data analytic tool gave retailers an across-the-board view of their
customers, so that they could better understand the customers’ behavior and sentiment. This
could be done by correlating such architectural elements as previous business transaction data,
customers’ online browsing history, customers’ in-store traffic flow, product and brand
preferences, and other relevant data streams.49
41
IBM Software (2013) “SHOP.CA optimizes and analyzes customers’ searches and shopping experiences” 42
Scott Steever (2014 ) Presentation on “How do you apply Big Data and analytics to fashion?” 43
Rouse, M. “cross-sell”, TechTarget, http://searchcrm.techtarget.com/definition/cross-sell (accessed on 30 June 2015) 44
SmallBizConnet “Business to Business Selling”, http://toolkit.smallbiz.nsw.gov.au/part/19/96/437 (accessed on 30 June 2015) 45
Subramanian, R. (8 March 2013) “Tips for Using Big Data to Optimize Upsell and Cross-Sell Strategies”, TotalRetail,
http://www.mytotalretail.com/article/big-data-can-grow-your-retail-business-upsell-cross-sell-strategies/ (accessed on 30 June
2015) 46
Perla, M. (9 May 2013) “The right way to cross-sell products”,
http://www.imediaconnection.com/content/34172.asp (accessed on 30 June 2015) 47
Spaulding, E. et al (16 September 2013) “Making it personal: Rules for success in product customization”, Bain,
http://www.bain.com/publications/articles/making-it-personal-rules-for-success-in-product-customization.aspx (accessed on 3
August 2015) 48
Pannaman, E. (27 May 2014) “Using Big Data To Improve Consumer Sentiment”, innovation enterprise,
https://channels.theinnovationenterprise.com/articles/using-big-data-to-improve-consumer-sentiment
(accessed on 30 June 2015) 49
Hinnsen, P. (2015) “Analysing customer behavior”, Greenplum
Page 7
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 7
With a more thorough understanding of their target customers, retailers could then better
communicate with their customers50 in the store, on social media, or via other channels. For
instance, retailers would know where to locate their target customers and what kinds of
marketing appeals would best arouse their interests in the products. The noise-free results
generated by Big Data analysis also had the advantage of being less biased. This allowed
retailers to develop products, including cross-selling items, that would better meet the needs
of their customers, determine the inventory assortments and pricing strategies aligned with the
purchasing habit of customers in different locations, and find new ways of boosting customer
satisfaction and loyalty.51
Optimizing Operations As put forward by Accenture, “Digital technology is disrupting traditional operations and now
every business is a digital business. The impact on supply chain management is particularly
great. The digital supply network is more connected, intelligent, scalable and rapid.”52 Big
Data was also in place to streamline the supply chain of a retailer by optimizing its inventory
level and preventing stock-out situations.53 This was of paramount significance to companies
since out-of-stock incidences would be quite costly as stockouts directly affected the level of
customer satisfaction, sales volume, and thus profit. If they happened too often, customers
would simply shift to the companies’ competitors for service.54
Not only could Big Data analytics be used to prevent ordinary stock-out situations, it could
also help companies adjust its stocking level by responding to some uncertain real time
demand. Tesco developed a Big Data analytics model that used weather patterns to properly
stock its store shelves. By focusing on the historical buying data of its customers and the
temperature forecasts, the British grocery and merchandise retailer could then precisely
predict what customers in a specific location would like to buy as the weather changed. Tesco
could then deliver the right merchandises to the right locations at the right time. For instance,
if an area would have a few warm days after some days of super cold weather, Tesco would
ensure that the stores nearby had ample barbecue meats for meeting consumer demands.55
One example for Big Data to optimize business operation was that it could improve store
layouts by analyzing the traffic flows within a store.56 Retailers could then re-design their
stores to facilitate product browsing for customers and fully utilizing the limited store space.
Sensor data generated by RFIDs or QR codes could be used to keep track of in-store traffic
and customers’ purchasing habits. 57 Memove, a Brazilian retailer, drew upon RFID
technology to monitor the clothing from manufacturer to customer, hereby tracking the
movement of in-store inventory and the popularity of each product.58
50
McKinsey & Company (22 July 2013) “Big Data, Analytics And The Future Of Marketing And Sales”, Forbes,
http://www.forbes.com/sites/mckinsey/2013/07/22/big-data-analytics-and-the-future-of-marketing-sales/ (accessed on 3 August 2015) 51
Oracle (2013) “Utilities and Big Data: Using Analytics for Increased Customer Satisfaction” 52
Hanifan, G. et al (2015) “The Digital Supply Network A New Paradigm for Supply Chain Management”, Accenture 53
SAGE (2013) “Better Inventory Management – Big Challenges, Big Data, Emerging Solutions” 54
Ibid. 55
Swabey, P. (16 April 2013) “Tesco saves millions with supply chain analytics”, Information Age
http://www.information-age.com/technology/information-management/123456972/tesco-saves-millions-with-supply-chain-
analytics (accessed on 30 June 2015) 56
Ingram Micro Advisor (2015) “6 Big Data Use Cases in Retail”,
http://www.ingrammicroadvisor.com/big-data/6-big-data-use-cases-in-retail (accessed on 3 August 2015) 57
Ibid. 58
McGraw Hill (2013) “Retailing Management – Newsletter for Instructors”
Page 8
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 8
Macy’s, a department store in the US, used Big Data analytics to figure out the number of
people in the store by checking the parking lot usage level. 59 This would allow data scientists
to estimate the sales volume of the day even before Macy’s itself would know the numbers.
The data could also help Macy’s better manage its stock levels and facilitate its daily
operations. Analyzing these real or near real-time data could make the retailer more agile than
its competitors.
Connecting Clicks and Bricks Big Data could be used to integrate the online and offline businesses of a company. Kohl’s,
another US department store, tested real-time offers in five of its stores for shoppers who
were willing to receive offers via their smartphones. For example, if its shoppers visited the
shoe department, the Kohl’s app could retrieve their past shopping and browsing history and
deliver an electronic coupon to their smartphones instantly for the shoes they checked online
earlier. With the timely e-coupon, Kohl’s would stand a high chance of completing that deal
since customers were more likely to respond to an offer at the moment of purchase when they
were shopping in the store [see Exhibit 4 for more findings on customers’ preferences among
in-store mobile offerings].60
In addition, marketers could leverage the use of Big Data analytics to adjust its integrated
marketing communications (IMC) so that the marketing message, content, and appeals used in
the online and offline marketing channels would be more effective in moving the target
audience.61 Marketers could make use of data like point-of-sale transaction logs, customers’
online activities, demographics and shopping patterns to optimize marketing effort [see
Exhibit 5 for a wide variety of data retailers could draw on to enhance marketing
effectiveness]. New marketing opportunities could also be identified when such information
was analyzed. Since customers were indeed a retailer’s most important data domain,
generating 360-degree customer profiles would enable the retailer to draw insights on the
latest market trends that it should capture when tailoring its marketing campaigns and
launching new products.62
More effective marketing campaigns that would enhance the viewers’ affection towards the
product were now possible with Big Data analytics. Retailers could optimize their marketing
expenditures as they would know which marketing channels (e.g. TV advertisements, social
media communication, in-store promotion) would best reach their target consumer segments,
and what kinds of marketing messages and appeals would best suit their products.63
Impacts of Big Data
The incorporation of Big Data into business operations changed the ways how marketing
worked. With the power of information, marketers now used a different approach to raise
customers’ awareness, arouse their interests, boost their desires, and eventually lead to
59
Brynjolfsson, E. et al.(October 2012) “Big Data: The Management Revolution” Harvard Business Review,
https://hbr.org/2012/10/big-data-the-management-revolution/ar (accessed on 30 June 2015) 60
Thau, B. (24 January 2014) “How Big Data Helps Stores Like Macy's And Kohl's Track You Like Never Before”, Forbes
http://www.forbes.com/sites/barbarathau/2014/01/24/why-the-smart-use-of-big-data-will-transform-the-retail-industry/
(accessed on 30 June 2015) 61
Martin (14 December 2014) “Big Data, Marketing Best Uses of Big Data in Marketing”, Entrepreneurial Insights,
http://www.entrepreneurial-insights.com/best-uses-big-data-marketing/ (accessed on 3 August 2015) 62
Arroyo, J. (24 February 2014) “Big Data and Advanced Analytics – The Key to the Market of One”, TFM& A Insights,
http://tfmainsights.com/big-data-advanced-analytics-key-market-one/ (accessed on 3 August 2015) 63
Hazan, E. et al (August 2013) “Leveraging Big Data to optimize digital marketing”, McKinsey,
http://www.mckinsey.com/client_service/marketing_and_sales/latest_thinking/leveraging_big_data_to_optimize_digital _marketing (accessed on 3 August 2015)
Page 9
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 9
purchase actions. These four sequential phases came together in the form of the AIDA
model.64
AIDA stood for Awareness, Interest, Desire, and Action [see Exhibit 6 for more details of the
model]. Traditionally, in the first phase of awareness, marketers needed to create brand
awareness among the target audience or affiliation with the company’s products. 65 They
should decide what kinds of outreach strategies and marketing platforms to use for promoting
products, and how to construct marketing messages. When it came to the phase of interest,
marketers aimed to arouse customers’ interests in the core benefits coming along with the
products, and had to ensure that the customers are sufficiently interested and attracted to
research further.66 After that, marketers should boost customers’ desires for possessing the
products through creating emotional connections between the customers and the brand.67 The
marketing campaign should bring out the brand personality and move the customers from
liking to wanting the products. Lastly, marketers should lead the customers to the next step of
action by making purchases.68
With the widespread implications of Big Data in the retail industry, the power of information
altered the traditional sequence of the four phases of the AIDA model. Awareness, interest,
desire, and action did not necessarily fall into the same order as before in this data-driven era.
A strong desire for owning a product so personalized to the needs of a customer would be
created at the first contact of the soon-to-be owner with the product. Big Data analytics would
also add vividness to the advertisements, which were now capable of moving the potential
customers to buy the product before they were aware of the needs of using it. The explosion
of data and digital technologies in merging the online and offline business of a retailer might
even dissolve the boundary of the four phases of A-I-D-A that they could happen
simultaneously.
Challenges Ahead
If a company wanted to draw upon Big Data for streamlining its business, it would need to
have a capability of collecting, storing, and analyzing a vast amount of data swiftly. The
company might either choose to invest in developing its own platforms of Big Data analytic
tools and training up its own teams of data scientists, or outsource the function of Big Data
analysis to a professional Big Data consulting firm for getting hidden insights revealed in the
datasets. Nonetheless, different challenges still made their presence along the way enterprises
tried to incorporate Big Data into their existing businesses [see Exhibit 7 for a survey’s
findings on the major challenges].
One of the major concerns about Big Data analytics was that it involved the use of a sheer
volume of data for deriving business insights, and this would cause data security and privacy
problems.69 With the recent data breaches at large retailers such as Target and Home Depot,70
issues about data security became a greater concern for retailers incorporating Big Data into
their businesses. Retailers using Big Data for benefits should also have the responsibilities for
upholding strict data security and privacy protocols to protect customers’ information,
64
DeMers, J. (8 May 2013) “How To Use The AIDA Formula To Boost Your Content Marketing Strategy”, Forbes,
“http://www.forbes.com/sites/jaysondemers/2013/08/05/how-to-use-the-aida-formula-to-boost-your-content-marketing-
strategy/ (accessed 20 July 2015) 65
Eagle, L.(2014) Marketing Communications, Routledge: New York. 66
Ibid. 67
Ibid. 68
Ibid. 69
Thuraisingham, B. et al (2014) “Workshop Report Big Data Security and Privacy for 16-17 September 2014”, The University
of Texas at Dallas 70
Sidel, R. (18 September 2015) “Home Depot's 56 Million Card Breach Bigger Than Target's”, Wall Street Journal,
http://www.wsj.com/articles/home-depot-breach-bigger-than-targets-1411073571 (accessed on 30 June 2015)
Page 10
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 10
especially personally identifiable information, and minimize fraud-related loss. Since laws
and regulations regarding the use of Big Data that varied across countries, companies should
get themselves familiar with these rules and guidelines for minimizing any risks.71
Besides, as found out by IBM researchers,72 a lack of skilled staff with sufficient Big Data
knowledge and training hindered enterprises from understanding how to use analytics to
improve their business operations.73 Since Big Data only gained its popularity in the past few
years across industries, it took time for more well-trained and experienced Big Data scientists
to turn up in the labor market. It was noteworthy though that as the Big Data and analytics
market continued to grow, the Big Data skills gap would continue to persist in the near future.
McKinsey Global Institute (MGI) projected that there would be a lack of 1.7 million big-data
related workers in the US in 2018.74 The IDC also estimated that 181,000 deep data analysts
and another 900,000 workers with related data management and interpretation skills would be
needed in the US by that year.75
Another major challenge posed to enterprises was that Big Data was indeed disruptive by
nature. Business leaders should be well prepared for any infrastructure, personnel and cultural
changes that their companies would need to go through.76 For instance, the workflow might
need to be changed with the installation of new computer software, and training should be
provided for staff in how to operate with the use of Big Data analytics. During the adoption
process, it would not be uncommon to see that some staff would be reluctant to cope with the
changes brought about by Big Data when integrating with the existing system.77 Moreover,
since Big Data analytics involved the use of data from various sources, collecting data from
departments (e.g. sales, marketing, and information technology) which housed the data would
require them to work closely together so that the latest data could be extracted in a timely
manner.78
Conclusion
Changes were something that cannot be avoided, and so were the waves and opportunities
brought about by Big Data. Companies that embraced the trend and successfully derived
insights from their datasets would enjoy a distinct advantage over their competitors. In the
future, when data was generated at a higher speed, cloud or other storage technologies became
well-developed, and more efficient data analytic tools turned up, Big Data would be nothing
but the central part of almost every business in the world.
71
Selby, J. (3 April 2015) “Legal Issues to Consider Before Starting Big Data Projects”, Datafloq,
https://datafloq.com/read/#!/Legal-Issues-Consider-Starting-Big-Data-Projects/1006 (accessed on 30 June 2015) 72
IBM is the abbreviation of International Business Machine, a multi-national information technology corporation. 73
Hayes, B. (8 February 2013) “Big Data has Big Implications for Customer Experience Management”, IBM,
http://www.ibmbigdatahub.com/blog/big-data-has-big-implications-customer-experience-management (accessed on 30 June
2015) 74
McKinsey Global Institute (June 2011) “Big Data: The next frontier for innovation, competition, and productivity” 75
Framingham, M. (11 December 2014) “IDC Reveals Worldwide Big Data and Analytics Predictions for 2015”, IDC,
http://www.idc.com/getdoc.jsp?containerId=prUS25329114 (accessed on 30 June 2015) 76
EMC (2011) “Big Data: Big Opportunities to Create Business Value” 77
Proformative (28 February 2013) “HR Must Make Better Use of Big Data”,
http://www.proformative.com/articles/hr-must-make-better-use-big-data (accessed on 20 July 2015) 78
Tectonic (2015) “Analytics Strategy”, http://www.gettectonic.com/big-data-analytics-old/analytics-strategy/ (accessed on 20
July 2015)
Page 11
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 11
Learning Objectives
This case study aims to introduce the basic characteristics of Big Data to students. It also
showcases how the use of Big Data could revolutionize the operations of the retail industry.
After completing the case study, students should be able to:
- understand the 4Vs (i.e. Volume, Variety, Velocity and Veracity) of Big Data;
- illustrate the uses of Big Data analytics in different retailing contexts;
- explain how the implementation of Big Data could alter the sequence of the AIDA
model; and
- critically evaluate the short- and long-term impacts of Big Data on the retail industry.
Suggested questions for students
How could Big Data revolutionize the original sequence of the four phases of the
AIDA model?
What are the short- and long-term impacts of Big Data on the retail industry?
How could retailers leverage the use of Big Data to create their own competitive
advantages?
Page 12
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 12
EXHIBIT 1: DIFFERENT TYPES OF EXPLOITABLE UNSTRUCTURED AND STRUCTURED DATA BY BUSINESSES
Source: PwC Australia (2012) “Big Data – The next frontier for innovation”
Page 13
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 13
EXHIBIT 2: FACTORS DRIVING COMPANIES TO IMPLEMENT BIG DATA IN THEIR BUSINESS
Source: Adapted from the Information week 2015 analytics & BI survey
Page 14
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 14
EXHIBIT 3: THE ACTUAL AND PROJECTED AMOUNTS OF UNSTRUCTURED AND STRUCTURED DATA GENERATED FROM 2006 TO 2020
Source: Adapted from http://lenovodev.com/blog/tech-modder/2013/01/27/welcome-to-the-tech-
modder-blog/ (accessed on 30 June 2015)
Page 15
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 15
EXHIBIT 4: CUSTOMERS’ PREFERENCE ON IN-STORE MOBILE OFFERINGS
Source: Accenture, “Accenture Seamless Retail Study”, April 2013
Page 16
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 16
EXHIBIT 5: A WIDE VARIETY OF DATA RETAILERS COULD USE TO OPTIMIZE THEIR MARKETING EFFORTS
Source: IBM (2013) “Capitalizing on the power of Big Data for retail”
Page 17
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 17
EXHIBIT 6: THE AIDA MARKETING MODEL
Source: http://strategimarketing-wow.blogspot.hk/2013/12/mengenal-istilah-aida-dalam-
digital.html (accessed on 20 July 2015)
Page 18
Big Data: Revolutionizing the Retail Industry (Ref.: 15/010C)
CENTENNIAL COLLEGE Page 18
EXHIBIT 7: MAJOR CHALLENGES CONFRONTED BY COMPANIES WHEN INCORPORATING BIG DATA INTO THEIR OPERATIONS
Source: Accenture, “Accenture Big Success with Big Data Survey”, April 2014