Roleof Artificial Intelligencein Social Media Marketing
Roleof Artificial Intelligencein Social Media Marketing
Roleof Artificial Intelligencein Social Media Marketing
net/publication/357705863
CITATIONS READS
9 14,406
1 author:
Anandvardhan Mishra
Deen Dayal Upadhyay Gorakhpur University
2 PUBLICATIONS 9 CITATIONS
SEE PROFILE
All content following this page was uploaded by Anandvardhan Mishra on 10 January 2022.
*
Research Scholar, Deen Dayal Upadhyaya Gorakhpur University, Gorakhpur, Uttar Pradesh, India.
Email: [email protected]
Role of Artificial Intelligence in Social Media Marketing 35
What is Artificial Intelligence? in defining how todays’ social media functions. Popular
social networks like Facebook, LinkedIn, Instagram, and
In the words of John McCarthy, the father of AI, artificial so on leverage machine learning models for suggesting
intelligence is “The science and engineering of making people/accounts to follow, job recommendations,
intelligent machines, especially intelligent computer identifying visuals, tracking current engagements, and
programs”. so on. According to Gartner’s report, 37% companies
are leveraging AI for content curation and management
AI refers to any human-like intelligence exhibited by a (Gartner Survey Shows 37 Percent of Organizations have
computer, robot, or other machines. In general terms, AI Implemented AI in Some Form, n.d.). As per the 2021
refers to the ability of the computer or machine to imitate State of Marketing Artificial Intelligence Report released
the capabilities of the human mind, like learning from by Drift and Marketing Artificial Intelligence Institute,
examples and experience, understanding and responding a majority of marketers know the importance of AI for
to what is said, decision making, solving problems, business, but 70% of respondents find lack of training and
and combining these and other capabilities to perform education a barrier in the adoption of AI in marketing
functions a human might perform, such as driving a car. (New Report Shows Marketers Lack the Education,
Demis Hassabis, CEO of Deepmind, an Alphabet Inc.- Training to Effectively Adopt Artificial Intelligence,
owned company, defines AI as the “science of making n.d.). To create customer recommendations, machine
machines smart”. Systems are being developed with the learning and natural language processing models derive
intellectual characteristics of humans, such as reasoning
social insights by analysing brand performance over time.
ability, discovering meaning, and generalising or learning
AI is a key component in the popular social networks we
from past experience.
use every day.
From the above discussion, it is clear that artificial intelli- focused on performing specific tasks; it enables some
gence (AI) is a machine displayed intelligence that mimics very impressive applications, including Apple’s Siri,
human behaviour or thinking, and can be trained to solve Amazon’s Alexa, and self-driving cars.
complex problems. Machine learning and deep learning
together Strong AI, also known as artificial general intelligence
Strongmake
AI, AI.
alsoAIknown
modelsasthat are trained
artificial usingintelligence
general big data (AGI), is an entirely theoretical concept
volumes have the required potential to make intelligent (AGI), is an entirely theoretical concept aimed at creating
aimed at creating intelligent machines that fully replicate the autonomy
intelligent machinesofthat
thefully
human brain.the autonomy of
replicate
decisions. Artificial intelligence has been basically classified
intoOptimists believe
2 categories, namelythatweak
AI and
would able to solve many the
be AI.
strong human brain.
types/classes Optimists and
of problems believe
willthat
evenAI would be able
to solve many types/classes of problems and will even
choose the problems it wants to solve, without human intervention.
Weak AI, also called narrow AI or artificial narrow choose the problems it wants to solve, without human
intelligence (ANI), is one where the AI is trained and intervention.
ARTIFICIAL
INTELLIGENCE
AI is the ability of a digital
computer or computer-
controlled robot to perform
tasks commonly performed by
intelligent beings.
MACHINE LEARNING
This is an AI application that
critically observes the data, finds
patterns, and applies its
understanding of the data in
making informed decisions.
Human interference is
sometimes required.
DEEP LEARNING
This is a subset of machine
learning that structures
algorithms in layers to create
an "artificial neural network"
that can learn and make
intelligent decisions on its
own. Human intervention is
not required.
Fig. 1
AI AI Facts
Facts andand Figures
Figures ● As per Gartner, 37% of organisations have imple-
mented AI in some form. The percentage of en-
Accordingtotothe
● According thelatest
latest release
release of of International
International Data Corporation,
Data terprisesworldwide
employingrevenues
AI grewfor270%the over the past
Corporation, worldwidesoftware,
AI market, including revenueshardware,
for the AIandmar- yearsthe(Gartner
fourbreak
services, will Surveymark
$500 billion Showsby 37 Percent of
ket, including software, hardware, and services, will Organizations have Implemented AI in Some Form,
2024,the
break with a compound
$500 billion mark annual growth
by 2024, withrate of 17.5% (IDC Forecasts Improved Growth for
a com- n.d.).
Global
pound AI Market
annual growth in rate
2021,
of n.d.).
17.5% (IDC Forecasts
● According to an article published in AI Business, a
Improved Growth for Global AI Market in 2021,
As per Gartner, 37% of organisations have implemented AI leading content
in some form.portal for artificialofintelligence and
The percentage
n.d.).
enterprises employing AI grew 270% over the past four years (Gartner Survey Shows 37
Percent of Organizations Have Implemented AI in Some Form, n.d.).
According to an article published in AI BUSINESS, a leading content portal for artificial
Role of Artificial Intelligence in Social Media Marketing 37
its real-world applications in 2018, Servion Global ● Real-Time Recommendations: Retail and entertain-
Solutions report predicts that by 2025, 95% of cus- ment websites use neural networks to recommend
tomer interactions will be powered by AI (AI will additional purchases to a customer, based on his past
Power 95% of Customer Interactions by 2025 – AI activity and other factors, including time of the day
Business, n.d.). and the weather. Recommendation algorithms make
items more visible to customers. It increases t h e
A recent 2020 report from Statista reveals that the global likelihood of buying them. The importance of real-
AI software market is expected to reach a forecast size time recommendations is evident from the fact that
A recentof $126
2020 report billion byreveals
from Statista the Market
2025that(AI Size 2018-2025
global AI software | to
market is expected
reach a forecast size of $126 billion by 2025 (AI Market Size 2018-2025 | Statista, n.d.).
75% of what consumers watch on Netflix and 35%
Statista, n.d.). of what they purchase on Amazon are a result of
150 product recommendations based on such algorithms
Market revenues in billion U.S. dollars
125
126
(How Retailers can Keep up with Consumers |
100 94.41 McKinsey,
70.94
75
51.27 ● Image Recognition: It is the AI technology that en-
50
22.59
34.87
ables systems or software to identify people, ob-
25 14.69
10.1
jects, writing, and even actions, within the images.
0
2018 2019 2020 2021 2022 2023 2024 2025 It uses machine vision technologies typically driven
Revenues from the AI software markets worldwide from 2018-
2025
by deep neural networks and trained algorithms to
Source: Statista. recognise images through a camera system.
Source: Statista.
Various Artificial Intelligence Applications ● Virus and Spam Prevention: Today’s virus and spam
Fig. 2 detection software employs deep neural network AI
Some popular applications of AI are as follows.
to detect new types of viruses and spams. AI spam
Natural Language Processing (NLP): Natural language processing is a subfield of artificial filters scan incoming messages and put a label on
Various Artificial Intelligence Applications
intelligence that enables machines to comprehend human language. NLP uses the power of objectionable content. Intelligent learning capabili-
computer sciences and linguistics to create intelligent systems capable of understanding,
ties label a warning sign on malware. If any message
analysingSome
aspectspopular
like syntax,applications ofon,
semantics, and so AIand
areextract
as follows.
meaning from the text and
containing this malicious software is found in the
speech. The sentiment analysis feature of NLP enables marketers to analyse posts, product
● Natural Language Processing (NLP): Natural lan- inbox, it alerts users not to open it.
reviews, and online surveys, and classify them as positive, negative, or neutral. Digital
guage processing is a subfield of artificial intelli- ● Ride-Share Services: Uber, OLA, Lyft, and other
assistants (such as the aforementioned Siri and Alexa), chatbots, and other text-based virtual
gence that enables machines to comprehend human
assistants work on the NLP technology of artificial intelligence. ride-share services use artificial intelligence to
language. NLP uses the power of computer sciences
Speech Recognition: Speech recognition is AI technology that converts spoken words into match up passengers with drivers to minimise wait
and linguistics to create intelligent systems capable
digitised text; it also has the ability to recognise individuals based on their voice commands. times and provide reliable ETAs (expected time of
of understanding, analysing aspects like syntax, se-
Tech giants like Google, Amazon, and Apple have developed many devices with voice arrival), and even curbs the surge in price due to
mantics, and so on, and extract meaning from the
recognition ability. Amazon’s Alexa, Google Home, and Apple’s Siri are a few examples. heavy traffic.
•
text and speech. The sentiment analysis feature of
Real-Time Recommendations: Retail and entertainment websites use neural networks to
NLP enables marketers to analyse posts, product re-
recommend additional purchases to a customer, based on his past activity and other factors,
● Household Robots: These robots are basically used
views, and online surveys, and classify them as pos-
including time of the day and the weather. Recommendation algorithms make items more
for cleaning, entertainment, and domestic security
itive, Itnegative,
visible to customers. increases t hor
e neutral.
likelihood ofDigital assistants
buying them. (such ofasreal-
The importance purposes. For example, iRobot’s vacuum cleaner
the aforementioned
time recommendations Siri
is evident from the fact that and
75 perAlexa), chatbots,
cent of what and on
consumers watch Roomba has advance decision making capabilities.
other text-based virtual assistants work on
Netflix and 35 per cent of what they purchase on Amazon are a result of product the NLP It uses AI to scan the size of a room, identify, and
recommendations technology
based on suchofalgorithms
artificial intelligence.
(How Retailers Can Keep up with Consumers | avoid obstacles. New robots are also embedded with
Speech
McKinsey,●n.d.). Recognition: Speech recognition is AI tech- voice recognition technology.
nology that converts spoken words into digitised text; ● Autopilot Technology: AI offers autonomy to drones
it also has the ability to recognise individuals based and aircrafts. By using a combination of sensors,
on their voice commands. Tech giants like Google, collision avoidance technology, robotics, GPS tech-
Amazon, and Apple have developed many devices nology, image recognition, and natural language
with voice recognition ability. Amazon’s Alexa, processing, AI guides an aircraft safely through the
Google Home, and Apple’s Siri are a few examples. skies. It provides the ability to estimate the posi-
38 International Journal of Business Analytics & Intelligence Volume 9 Issue 1 & 2 April & October 2021
tion, orientation, and velocity of the vehicle; pro- SMM saves a lot of time and money spent on reviewing
vides the ability to compute and execute necessary analytics, creating performance reports, drafting social
commands, map the environment, and also provides media posts, building promotion strategies, and so on.
real-time navigation facility. The marketers who leverage AI technology in their social
media marketing campaigns have the opportunity to
E-commerce, automated stock trading, navigation, create a significant and sustainable competitive advantage
robotics, human resource management departments, for their businesses. AI enables marketers to:
healthcare, agriculture, gaming, automobiles, banking,
and social media marketing are some of the major areas ● Decrease cost: AI can automate some of the tasks
where AI applications are predominantly used. which require more time and toil. This includes ac-
tivities like scheduling and uploading posts, pulling
Advertisers can take AI services of open sources relevant links and hashtags, and conducting research
like Google or can avail the services of a third party into what to post easier and faster.
like Persado, Lately, Socialbakers, Cortex, Phrase, ● Increase revenue: AI can help someone to better de-
Worldstream, Linkfluence, and so on for a price mostly termine what to post, whom to target, which medi-
determined by lead conversion rate. ums of promotion will be beneficial, and so on. AI
can even write more effective advertisements than
Key Differences between AI Marketing humans.
and Non-AI Marketing ● Create social media posts faster: The AI tool uses
historical data to learn what really works and what
Table 1
social media posts to share next; they help marketers
produce posts on a large scale.
Sr. AI Marketing Non-AI Marketing
● Develop right message for each platform: AI can
No.
automatically develop creative messages suiting a
1. Creates its own algorithm Built on algorithm, human-
and lays down a path to ad- coded set of instructions that brand across the platforms to increase social media
vanced marketing. tell a machine what to do. engagement.
2. AI can read and write natu- It cannot read natural lan- ● Measure brand and trends across social media: AI-
ral language. guage; traditional social me- powered social media intelligence can help compa-
dia tools help in creating con-
nies measure and improve brand equity, detect con-
tent for social media.
sumer trends, and understand target audiences.
3. Automates decision mak- Marketing teams take the de-
ing. cision. ● Determine what to post for maximum effect: AI-
4. One-to-one approach. One-to-many approach. powered social media tools can analyse marketer’s
5. Programmatic demand-side Manual advertising cam- posts, and the posts of other companies, to recom-
platforms can use hundreds paigns (even those performed mend what to post.
of targeting signals to indi- with professional tools) take
vidualise the advertisement, into account three or four
● Find product logos, brands in posts online: Using
and even the target, accord- targets: the keyword, time of image and face recognition tools, marketers can
ing to lifestyle or behaviour, day, and location. identify logos and brands on the Internet.
when integrated with cus-
● Find the right influencer: AI-powered influencer re-
tomer data platforms.
search platforms analyse a variety of social media
analytics, to understand which accounts can provide
Advantages of using Artificial the most engagement, reach, and influence for a spe-
Intelligence in Social Media Marketing cific industry.
● Manage and improve social media advertising: AI
AI has touched almost all facets of life. However, helps pay-per-click advertising by displaying the
marketing is the sector that is the most affected by this right advertisement to the right people at the right
advanced technology. Employing AI technology in time, and has also taken over some repetitive tasks
Role of Artificial Intelligence in Social Media Marketing 39
of PPC management. AI can analyse PPC (pay-per- high-performance computers and software are required.
click) advertising across Facebook and Instagram, This is an obstacle, especially for smaller organisations.
and offers recommendations to improve perfor-
mance-based data. Investment of Resources - AI requires complex software
and high-performance hardware, which is expensive to
● Improved personalisation and recommendations: deploy and maintain. However, a large number of AI
Consumers are more likely to interact with person- developers are coming out, which means organisations
alised marketing messages. AI enables marketers to no longer have to rely on developing in-house solutions.
personalise their communications on an individual In addition, organisations like Google, Amazon, and IBM
level. are offering their algorithms to the world; some third-party
● Dynamic pricing: AI can be used to set the price services are open source, while others are available for
of the products dynamically, depending on demand, some consideration. A few organisations may need to
availability, popularity, and other factors, to maxi- allocate budget towards hiring AI experts.
mise both sales and profit.
Lack of Talent: Presently, there is a huge AI skill gap. This
● Customer service chatbots: Organisations are us-
issue is expected to become worse as the number of AI
ing chatbots to deal with frequently asked ques-
technology organisations develop.
tions. Chatbots provide replies 24×7. These are pro-
grammed to provide a set of replies to FAQs and Lack of Data or Poor Data Quality: AI requires a large
direct the conversation to a human agent if the con- amount of standard data, to get insight about the users.
versation is complicated. Chatbots are cost effective Substandard or inadequate data will result in poor results
and sometimes can be more effective and polite than from the AI software.
humans.
Lack of Trust in AI Software:AI is a new and suspicious
● Search engine optimisation: Integrating AI into
technology. Laggards in technology and even technical
search engines can help pick up misspellings and
employees not trained in AI are suspicious of its use.
suggest alternatives, influenced by users’ past
Privacy of users is at stake as well, as data collected for
browsing or shopping behaviour. Marketers can use
one purpose can be sold to some other marketer for a
AI to optimise voice search, improve SEO, and site
different purpose.
traffic.
● Pay-per-click advertising optimisation: AI adver- Certain AI Applications Seem Unethical: For example,
tising optimisation is used on social networks like figuring out an unmarried customer is pregnant by
Instagram. The algorithm analyses the accounts that analysing her purchase patterns, preferences, browsing,
a user is following and shows advertisements that and so on; this is a practice that many users dislike, and
are most relevant to this user. which develops a negative perception of AI.
Unemployment – As we all know, AI-powered tools start leveraging AI-powered marketing technologies and
have the capability to perform many functions which are cognitive computing, among many others, to enjoy the
usually performed by intelligent beings. Therefore, there competitive edge.
is a fear of job loss in many industries which rely on
technology. References
AI Might Limit Creativity: Excessive use of technology AI Market Size 2018-2025 | Statista. (n.d.). Retrieved
will eventually degrade individuals’ creativity. April 28, 2021, from https://www.statista.
com/statistics/607716/worldwide-artificial-in
Suggestions telligence-market- revenues/
AI in Social Media Market Worth 2,197.1 Million USD by
● Stakeholders must make sure that existing data sets 2023. (n.d.). Retrieved April 24, 2021, from https://
are cleaned and that the data being collected is of www.marketsandmarkets.com/PressReleases/ai-in-
standard quality. social-media.asp
● By explaining AI algorithms and educating users AI will Power 95% of Customer Interactions by 2025
about when, where, and how information about – AI Business. (n.d.). Retrieved April 28, 2021,
them is gathered, customers’ trust and loyalty can be from https://aibusiness.com/document.asp?doc_id
=760184
increased.
Gartner Survey Shows 37 Percent of Organizations have
● AI cannot be left all alone to work on itself. Human Implemented AI in Some Form. (n.d.). Retrieved
intervention and control are always required. April 25, 2021, from https://www.gartner.com/en/
● Marketing managers should make AI training man- newsroom/press-releases/2019-01-21-gartner-sur-
datory for their marketing teams. vey-shows-37-percent-of-organizations-have
● If possible, companies should create in-house mar- How Retailers can Keep up with Consumers | McKinsey.
keting solutions and should not remain dependent (n.d.). Retrieved April 28, 2021, from https://
www.mckinsey.com/industries/retail/our-insights/
on third-party services for AI.
how-retailers-can-keep-up-with-consumers#
IDC Forecasts Improved Growth for Global AI Market
Conclusion in 2021. (n.d.). Retrieved April 28, 2021, from
https://www.idc.com/getdoc.jsp?containerId=
AI is revolutionising industries with its vast applications prUS47482321
and helping in solving complex problems. There is no
Kaplan, A. M., & Haenlein, M. (2010). Users of the world,
proof required to accept the growth of AI in marketing. unite! The challenges and opportunities of social
With the help of AI, marketers are able to identify media. Business Horizons, 53(1), 59-68. doi:https://
potential customers, create content, and follow the leads. doi.org/10.1016/j.bushor.2009.09.003
Incorporating AI can assist marketers, individuals, and New Report Shows Marketers Lack the Education,
advertising agencies in making social media marketing Training to Effectively Adopt Artificial Intelligence.
more efficient. Machine learning models, advance (n.d.). Retrieved April 23, 2021, from https://
analytical tools, and deep learning network algorithms www.marketingaiinstitute.com/blog/2021-state-o
have the potential to increase the benefits of using AI f-marketing-ai-report-launch
in social media over time. The functions and popularity State of Artificial Intelligence for Enterprises. (n.d.).
of artificial intelligence are soaring day by day. AI has State of Marketing Insights and Trends from 3,500 Global
significantly evolved over the past few years and has Marketing Leaders. (n.d.).
found its applications in almost every business/sector. AI
Top 10 Artificial Intelligence Applications [Updated
has widened its reach in all the fields of life, but has made 2021]. (n.d.). Retrieved April 24, 2021, from https://www.
its way deeper in the marketing sector. The marketing simplilearn.com/tutorials/artificial-intelligence-tutorial/
sector has witnessed rapid technological changes over artificial-intelligence-applications
the last decade. There is no better time for marketers to