Entrepreneurship

ISSN: 1314-9598
Volume: VII, Issue:2, Year: 2018, pp. 298-304

ARTIFICIAL INTELLIGENCE AND MARKETING

Abstract
The application of Artificial intelligence (AI) in marketing is in order to
continuously follow and predict the next purchasing decisions of the target consumers
and to improve their consumer “journey”. The power of AI is reflected in its core
elements: big data, machine learning and powerful solutions. The concept of "big data"
means that marketers have ability to aggregate and segment huge amounts of data with
minimal manual work. By using this data, they will be sure that they would deliver the
right message to the right people at the right time, via the channel of choice. Machine
learning (deep learning) allows marketers to understand and draw logical conclusions
from large data collections. They can predict consumption trends, track and analyze
consumer purchases, predict the next consumer behavior. Making powerful solutions
means that we are living in an era when machines truly understand the world in the
same way as humans. Machines can easily identify concepts and themes across a range
of data, interpret emotions and human communications, and generate adequate
responses to consumers. They can easily predict the behavior and decisions of buyers
and use that data to solve issues in future.
In the following years, marketers can expect greater AI impact, through more
intelligent searches, smarter ads, refined content delivery, relying on bots, continued
learning, preventing fraud and data breaches, sentiment analysis, image and voice
recognition, sales forecast, language recognition, predictive customer service, customer
segmentation, etc. This paper attempts to discover the future relationship between
marketers and artificial intelligence machines.

Key words: Artificial intelligence, marketers, marketing, machine learning, big

1
European University-Republic of Macedonia, Skopje, Faculty of Economics, Full-time
professor in Marketing, PhD, [email protected]
2
European University-Republic of Macedonia, Faculty of Economics, Full-time
professor in Corporate Finance, PhD, [email protected]
3
National Bank of the Republic of Macedonia, Advisor, Statistics Department,
[email protected]
298
 1. Introduction
Nowadays, companies use and appreciate the values of business
intelligence. Business data are analyzed for multiple purposes: system log
analysis, analysis of comments and opinions published on social media, risk
assessment, consumer retention, brand management, etc. (Tjepkema, 2017).
These different tasks can be performed by different, separate systems. However,
separate systems do it in an expensive and slow way. There is a need of a large
infrastructure that will manage all heterogeneous large data of separate multi-
systems.
The “big data” changes the way people live and work, from doing
business to shopping in the store and buying movie tickets. Each piece of
information is segmented and used to analyze the way consumers think and
make purchasing decisions. In order to take advantage of this opportunity, one
has to forget the existing outdated and non-innovative solutions. We live in an
era when using new technologies can easily identify future consumer trends and
more deeply analyze the way consumers decide about purchases.
Developed economies increasingly use data-intensive technologies. There
are 4.6 billion subscribers to mobile phones worldwide and between 1 and 2
billion people using the Internet (Schulterbraucks, 2017). This means that
modern technology and artificial intelligence are increasingly used. Artificial
intelligence is a subset of computer science in which computers can undertake
reasoning and common sense tasks, such as vision and knowledge, which were
formerly only undertaken by humans. It includes tasks such as learning,
watching, observing, understanding, talking, socializing, planning, reasoning,
creativity and problem solving. It is thought that 63% of companies already use
artificial intelligence tools without knowing it. 47% of consumers, however, use
bots that communicate when buying online products. 40% of consumers do not
mind if they get an answer to their questions from an artificial intelligence tool,
as long as they receive a satisfactory, quick, helpful and simple answer. But
what does all this mean for marketers? Thanks to the bots, consumers can
receive personalized shopping recommendations and for them there is always
available consumer care.

299
 2. Big data
The term "big data" denotes an explosion in the quantity and diversity of
high-frequency digital data. In a company, large amounts of data come from
everywhere: posts and comments on social networks, digital images and videos,
electronic catalogues, reports on cash and consumer transactions, GPS signals
from mobile phones, etc. These data are very large and complex which
traditional data processing software can not process. The main challenges
associated with large data are their collection, storage, analysis, search, sharing,
transfer, visualization, updating and their privacy. With the exponential increase
in the amount of data, the current technologies become obsolete. Big data
processing requires complex coding skills, domain knowledge, and statistical
knowledge.
Big data can be explained by the 3V dimensions - Volume, Velocity,
Variety (Laney, 2001).
Volume: Companies produce terabytes of data every day! With the
increase in the volume of data, while using the traditional ways to process them,
many data remains unprocessed. Although the company has all the information
about what consumers are doing and want, they can not answer them
accordingly. This is a sure alarm for the company to change something.
Variety: the volume of data is only an initial problem. The real problem
for traditional tools is the diversity of data. Traditional processing tools are best
used for structured data. They process only data that is well structured and
formatted. But today's data collected from various sources, e-mails, consumer
comments, social media, consumer web site experiences and call centers are
unstructured or semi-structured.
Velocity: the speed at which data is generated is critical as the previous
two factors. The speed at which the company analyzes data can be its
comparative advantage.
Each of these V-dimensions can be achieved with traditional solutions.
For example, if in the company most data is structured, it can achieve 80-90%
of the business value through traditional tools. But if the company faces all 3V
dimensions at once, then it has a problem of "big data". Accordingly, the "big
data" technologies can be described as a new generation of technologies and
architecture designed to economically draw value from large amounts of diverse
data with high speed and deep analysis. These data are so large that current or
300
traditional technologies can not cope with them (to collect, store, manage and
analyze it efficiently and in a shorter time).
3. Machine learning
Machine learning is used to predict or provide calculated suggestions that
are based on analyzing large amounts of data (Tjepkema, 2017). The best
examples are Netflix algorithms that offer suggestions for watching movies to
consumers based on movies they have seen in the past or Amazon's algorithms
that recommend books to consumers based on purchased books in the past.
Machine learning is an advanced area of artificial intelligence that allows
programs to absorb large amounts of data and create predictable algorithms that
improve over time. With the use of machine learning, marketers can provide
customized content to consumers, as well as suggestions for products that they
could buy. Marketers have data that guarantee success for offering consumers
what they want to buy.
Machine learning algorithms are divided into three major categories:
supervised learning, unsupervised learning and reinforcement learning (Sterne,
2017). Supervised learning is used in cases where a particular set of data lacks
certain data and it should be adequately predicted. Unsupervised learning is
used when the connection between unrelated, diversified data is to be
discovered in a huge set of data. Reinforcement learning is located between
these two extremes, i.e it predicts what will happen if there is a connection
between two or more events or data.
Machine learning is interdisciplinary in nature and involves techniques
from various fields, such as computer science, statistics, mathematics, artificial
intelligence, etc. The main feature of machine learning is to obtain experience
data by applying algorithms that rely on computer vision, artificial intelligence
and data mining.
4. Powerful solutions
Decision making consists of two main tasks: tree introduction and tree
pruning (Davis, 2016). Tree introduction and creation of a "tree" of data is a
task when a multitude of data is segmented by characteristics, then selects,
separates and categorizes it. The goal is that the data tree consists of the purest
minimal data that is split across the entire set of data. The purity is measured
according to the concept of information that shows how well one needs to know
301
about the particular topic. The complete data tree can be too complicated and
contains unnecessary structures, it is difficult to understand and difficult to
interpret. Tree pruning is the process of removing unnecessary structures from
the decision tree so that people can make more efficient, more accurate and
more understandable decisions.
5. The future of aritifical intelligence marketing
In future, marketers can expect the following impacts from artificial
intelligence marketing (Tjepkema, 2017):
A) More intelligent searches: As technology solutions become wiser and
more perfect, it's important to keep in mind that the public becomes more
sophisticated in their requirements. Thanks to social media and fast search
engines like Google, people find what they need in a very short time. Artificial
intelligence and big data can analyze these search models and help marketers
identify the key areas they need to focus their efforts on.
B) Smarter Ads: Marketers are still trying to attract consumers with smart
advertisements today, but with using artificial intelligence, they can do it faster
and more efficiently. With big data, online ads are becoming smarter and more
effective. Artificial intelligence can dig deeper into the data, social networks,
profiles, and other online content for human solutions.
C) Refined Content Delivery: With the help of artificial intelligence,
marketers will be able to target consumers to a completely different level. By
analyzing the targeted consumers, marketers, besides the demographic
characteristics, will understand the consumers on an individual basis. Now,
marketers can use artificial intelligence on two levels: identify potential
customers more easily and deliver the ideal content that is most relevant to
them. This is achieved through big data, machine learning and their
combination.
D) Relying on Bots: Consumer care and retention is another area where
artificial intelligence can play a significant role in the future. Very soon, the
conversation functions and other direct-to-consumer engagement will be carried
out by artificial intelligence bots. In this way, companies will save time for
employees and reduce costs. The artificial intelligence bots will have access to
the entire internet of data, information, a search history so that they become
more effective than humans.

302
 E) Continued Learning: With the help of artificial intelligence, not only
will it reveal some hidden data, but it will teach them and incorporate them into
new promotional campaigns and optimize the messages to the most relevant
users. Over time, the artificial intelligence solutions will become more
intelligent, effective and will promote decision making in real time.
6. Conclusion
Artificial intelligence, through big data, machine learning and perfect
solutions, are already transforming the technological landscape of the
companies. Companies change the way they do business by making it more
responsive, more productive, and more competitive. The technological advances
have always created new opportunities for marketing. Just as the advent of TV
brought an era of truly mass advertising and reach, and the Internet and mobile
phones brought a new level of targeting and context, so artificial intelligence
will change how people interact with information, technology, brands and
services.
Artificial intelligence is a study of how to make machines intelligent or
capable of solving problems as well as people can. At its core, machine learning
is a new way of creating those problem-solving systems. For decades,
programmers manually coded computer programs to provide outputs when
given a certain input. With machine learning, we teach computers to learn
without having to program them with a rigid set of rules (Sterne, 2017).
What does this mean for marketers? The further integration of technology
into the physical, real world, new interactions are created with consumers that
are simpler and instantaneous. High customer expectations will be higher than
ever. This will pose a new challenge, a new opportunity for companies and
marketers. Artificial Intelligence helps marketers to realize full personalization
and relevance. With platforms like Search, Facebook, YouTube, Google
reaching billions of people every day, and with digital ad platforms, it will
finally achieve communication at scale. This scale combined with a
customization possible through artificial intelligence means that companies will
soon be able to tailor personalized campaigns in real time. The world of the
future is in the application of artificial intelligence.

303
REFERENCES
Davis Ben (2016). 15 Examples of Artificial Intelligence in Marketing, Econ-
Consultancy, Manchester, UK, 19 April 2016
Johnsen Maria (2017). The Future of Artificial Intelligence in Digital Marketing – The
next big technological break, USA
Korada Prasanthi (2018). Impact of Artificial Intelligence on Digital marketing, Andhra
University, India
Laney Doug (2001). Controlling Data Volume, Variety and Velocity, in 3D Data
Management, Meta Group, (https://blogs.gartner.com/doug-
laney/files/2012/01/ad949-3D-Data-Management-Controlling-Data-Volume-
Velocity-and-Variety.pdf)
Lasse Shulterbraucks (2017). A short history of Aritificial Intelligence, December 2017,
https://dev.to/lschultebraucks/a-short-history-of-artificial-intelligence-7hm
Marvin Chow (2017). Artificial Intelligence and Machine Learning get us one step
closer to relevance at scale, Emerging Technology Mobile, September 2017
Sterne Jim (2017). Artificial Intelligence for Marketing- Practical Applications”, Wiley,
New Jersey and Canada,
https://books.google.mk/books?id=o_YtDwAAQBAJ&printsec=frontcover&dq=
artificial+intelligence+and+marketing&hl=en&sa=X&ved=0ahUKEwiiycCUrO
DdAhVNaVAKHX_ACkoQ6AEIJjAA#v=onepage&q=artificial%20intelligence
%20and%20marketing&f=false
Tanase George Cosmin (2018). Artificial Intelligence: Optimizing the Experience of
Digital Marketing”, Romanian Distribution Committee Magazine, vol. 9, issue 1,
24-28, Econpapers,
https://econpapers.repec.org/scripts/search.pf?ft=artificial+intelligence+and+mar
keting
Tjepkema Lindsay (2017). What is Artificial intelligence marketing and why is it
powerful?, Emarsys
Utku Kose, Selcuk Sert (2017). Imrpoving Content Marketing Processes with the
Approaches by Artificial Intelligence, EcoForum, vol. 6, issue 1, 31

304