1

Leveraging Data Analytics for Strategic Advantage

Name of the Student:
Student ID:
Name of the Author:
Course Code:
Date Submitted:
 2

Executive Summary
The report is aimed at describing the two examples of successful data analytics initiatives in
Coca-Cola Company. This information is required to evaluate on why the company has
continued to be at the top- notch of their competitors. The two initiatives described in this
report include: the use of AI in Vending machines and Client Reward programmes
Experience Using OCR. Coca cola sells most of its beverages through vending machines and
due to this it has implemented AI algorithms in the vending machines allowing them to
advertise their flavours and collect the opinions of the customers regarding the tastes and
preferences. This allows the company to stay at the top of the competitors in the market
since they are able to use the data collected to improve their brands and products. This
application is of benefit to the marketing department since it builds up their reputation
when utilised well by increasing the sales. Coca-Cola recently encountered another obstacle
in developing an innovative approach for its well-known loyalty programmes, which
facilitate clients to use the item's codes written on drinks bottle caps for an opportunity to
win various giveaways or sweepstakes. The loyalty programme was at MyCokeRewards.Com
but due to the launch of smartphones the programme was implemented in the apps. This
study resulted to the creation of Convolutional Neural Networks (CNNs), an alternative
machine learning technology with the possibility for higher precision than present optical
character recognition (OCR) approaches. In conclusion because of the AI algorithms in the
vending machines, Coca-Cola can gain an improved awareness of how preferences and
drinks tastes vary between consumers worldwide. Furthermore, the business asserts that
client participation has risen substantially since the incorporation of Tensor Flow's artificial
intelligence system into their OCR solution. The ability of the company to conquer in the
above detailed data should enable them to produce new brands and products into the
market.
 3

Introduction
Coca-Cola is the globe's biggest beverage organisation, supplying over 1.9 billion drinks per
day throughout its 500 brands. Because the organization is so large and influential in a
variety of nations within the world, AI powers every detail the organisation does and serves
as its foundation. Marketing is a large area of business where AI is being utilised thoroughly
(Aaker and Moorman, 2023). With more than 200 distinct brands all providing numerous
goods in over 200 various nations there are local variations in marketing preferences,
calories, flavours, and sugar. As a consequence, so as to maintain its 'top' position, it needs
to gather and analyse enormous quantities of data from multiple sources in order to
determine which products are most probable to be received favourably in which areas (Ma,
2021). Even taste can differ from places to places, so handling these minor details is critical
to the brand's effectiveness. The report is aimed at looking into two data analytics initiative
that the company has employed to make the business a success. The company has
employed the data analytics initiatives in the marketing department using the AI systems in
the vending machines and Client Reward programmes Experience Using OCR.
AI Usage in the Practices of Coca-Cola
Vending Machines
Coca-Cola sells an extensive number of its beverages via vending machines. Clients are able
to communicate with numerous novel machines to a stage whereby they can select
various customizations of their drinks with shots of completely distinct drinks added. Coca-
Cola Freestyle initially launched in America in 2009, and based on data, there are now
50,000 units in use worldwide, providing 14 million drinks per every day (Shi, 2019). Coca-
Cola has started implementing artificial intelligence (AI) algorithms to operate these
machines, allowing them to advertise the drinks and flavours that are more common all
around and make recommendations to consumers based on the precise spot of the
beverages dispenser (Vyas et. al., 2019). Additionally, the organization announced that they
are always looking for new ways to remain significant and customers are going to soon be
capable of develop customized drinks by employing an app, permitting them to swiftly get
their particular drink poured when they show up at the vending machine (Tang et. al., 2022).
Coca-Cola gathers data on native drink tastes for various areas via the user interfaces in its
electronic touchscreen machines (there are now over 1 million of these in Japan alone)
(Marr, 2019). Coca-Cola utilises the AI systems in the vending machine to enable them
understand why the customers buy their products and why have the customers shifted from
a certain brand to another so that they can use the information to improve their brands and
products in the market.
Client Reward programmes Experience Using OCR
Tensor’s Flows Machine Learning Framework
Coca-Cola recently encountered another obstacle in developing an innovative approach for
its well-known loyalty programmes, which facilitate clients to use the item's codes written
 4

on drinks bottle caps for an opportunity to win various giveaways or sweepstakes (Istrefi
and zdravevski, 2020).

Figure 1: Examples of Codes Printed on bottle Caps to facilitate the company Promotions
Source: Google Developers

Coca-Cola's fundamental loyalty programme was once available at MyCokeRewards.com,

but in response to an upsurge in prominence of mobile devices, they replaced it with an app
accessible to users of Android as well as iOS (Unal et. al., 2021). Recognizing that manually
inputting a 14-character code on a smartphone or tablet may turn off certain consumers,
Coca-Cola tried to make simpler the procedure of gaining reward points for its clients.
This study resulted to the creation of Convolutional Neural Networks (CNNs), an alternative
machine learning technology with the possibility for higher precision than present optical
character recognition (OCR) approaches.
OCR improves employee morale, makes the workplace simpler to navigate, and is less
paper-centric. Text recognition reduces documentation. Businesses and entities are more
probable to function easily when technological information is readily available because
employees have access to multiple customers at once (Unal et. al., 2021). Because OCR has
affordable processing, it enhances the way organization functions. Mobile OCR allows
individuals to keep track of their utilities through an easy-to-use app. There is no reason for
customers to continually record readings from the utility metre in order to track their
consumption. In this case the marketing department was the one benefiting from the OCR
solutions since it would be able to understand the tastes and preferences of their customers
and attract them through promotions of scanning the codes at the bottle caps.
Tensor Flow is a platform for machine learning that simplifies the creation of neural network
models such as CNNs, allowing firms such as Coca-Cola to rapidly build, evaluate, and
enhance effective connections for their OCR solutions. Patrick Brandt, Senior Solutions
 5

Strategist at Coca-Cola at the time, describes how the firm utilized Tensor Flow to develop
an application that is easy to use. The organization developed a learning app for both
Android and iOS smartphones that 'trainers' could employ to photograph and label codes;
the labelled pictures were subsequently uploaded to cloud storage for training (Vasilache et.
al., 2018). The company then printed a number of coded products on bottle caps and fridge
packs and gave them to various vendors, who then utilized the app to develop the first real-
life instruction set. The data gathered from the aforementioned training set gave an initial
assessment of the app’s accuracy in forecasting in both replicated and actual image
illustrations. Brandt asserts that the user interface additionally provides for the creation of a
feedback cycle when consumers correct the app's forecasts, progressively enhancing the
precision of the image acknowledgment approach.

Figure 2: Coca-Cola use of the Tensor Flow System. Source: Google Developers

Brandt asserts that as a consequence of this OCR implementation, Coca-Cola was in a

position to not only "finally accomplish a long-sought proof-of-purchase capability," but also
fuel "more than a dozen advancements and... over 180,000 scanned codes" in the six
months adhering to the release of the app in 2017 (Unal et. al., 2021). In an interview from
2018 with Tensor Flow, Brandt addresses how this led to one of the company's largest
online advertising campaigns.
Conclusion
Because of the AI algorithms in the vending machines, Coca-Cola can gain an improved
awareness of how preferences and drinks tastes vary between consumers worldwide. This
information is essential for deciding how, when and where novel goods should be
manufactured and sold. Because of AI analysis of images, Coca-Cola is able to develop social
media content that is more probable to have an impact with its intended audience in
different parts around the world, leading to an increased effectiveness of marketing
 6

expenses as well as higher sales margins. Furthermore, the business asserts that client
participation has risen substantially since the incorporation of Tensor Flow's artificial
intelligence system into their OCR solution. Coca-Cola's capacity to capitalise on such
comprehensive and detailed data should enable the company to keep innovating novel
goods and novel methods to meet the needs of consumers.
 7

References

Aaker, D.A. and Moorman, C., (2023). ‘Strategic market management’. John Wiley & Sons.
Available Online at: https://www.amazon.com/Strategic-Market-Management-
David-Aaker-ebook/dp/B0785W5L6V (Accessed on 11th October 2023).

Istrefi, D. and Zdravevski, E., (2020). ‘Making vending machines smarter with the use of
Machine Learning and Artificial Intelligence: Set-up and Architecture’. Available
(Online) at:
https://www.researchgate.net/publication/348311074_Making_vending_machines_
smarter_with_the_use_of_Machine_Learning_and_Artificial_Intelligence_Set-
up_and_Architecture (Accessed on 11th October 2023)

Ma, B., (2021). ‘Faculty of Informatics and Management Department of Management

(Doctoral dissertation, University of Hradec Králové)’. Available (Online) at:
https://www.uhk.cz/en/philosophical-faculty/about-faculty/departments/
department-of-political-sciences/student/doktorske-studium (Accessed on 11th
October 2023).

Marr, B., (2019). ‘Artificial intelligence in practice: how 50 successful companies used AI and
machine learning to solve problems’. John Wiley & Sons. Available (Online) at:
https://www.amazon.com/Artificial-Intelligence-Practice-Successful (Accessed on
11th October 2023)

Shi, Y., (2019). ‘Glocalization in China: An Analysis of Coca-Cola’s Brand Co-Creation Process
with Consumers in China (Doctoral dissertation, The University of Western Ontario
(Canada))’. Available (Online) at: https://ir.lib.uwo.ca/cgi/viewcontent.cgi?
article=8470&context=etd (Accessed on 11th October 2023)

Tang, Z., Xu, X., Song, Y. and Yang, H., (2022, March). ‘Data Analytics Applications in the
Soda Industry’. In International Conference on Business and Policy Studies (pp. 677-
688). Singapore: Springer Nature Singapore. Available (Online) at:
https://www.atlantis press.com/proceedings/icfied-22/125971633 (Accessed on 11th
October 2023)
 8

Ünal, A.E., İnce, G., Ay, K., Öztürk, S. and Gezer, C., (2021, August). ‘An Easily Configurable
Customer Loyalty Framework Using Distributed Ledger Technologies’. In
International Conference on Intelligent and Fuzzy Systems (pp. 767-774). Cham:
Springer International Publishing. Available (Online) at: https://www.aeunal.com/
(Accessed on 11th October 2023)

Vasilache, N., Zinenko, O., Theodoridis, T., Goyal, P., DeVito, Z., Moses, W.S., Verdoolaege,
S., Adams, A. and Cohen, A., (2018). ‘Tensor comprehensions: Framework-agnostic
high-performance machine learning abstractions’. arXiv preprint arXiv:1802.04730.
Available (Online) at: https://arxiv.org/abs/1802.04730 (Accessed on 11th October
2023).

Vyas, S., Jain, S.S., Choudhary, I. and Chaudhary, A., (2019, February). ‘Study on use of ai
and big data for commercial system’. In 2019 Amity International Conference on
Artificial Intelligence (AICAI) (pp. 737-739). IEEE. Available (Online) at:
https://www.researchgate.net/publication/332756151_Study_on_Use_of_AI_and_B
ig_Data_for_Commercial_System (Accessed on 11th October 2023)