Content uploaded by Krupa Rai
Author content
All content in this area was uploaded by Krupa Rai on Jul 09, 2020
Content may be subject to copyright.
Annals of “Dunarea de Jos” University of Galati
Fascicle I. Economics and Applied Informatics
Years XXV – no1/2019
ISSN-L 1584-0409 ISSN-Online 2344-441X
www.eia.feaa.ugal.ro
DOI https://doi.org/10.35219/eai158404094
Applications of Artificial Intelligence in Marketing
Vinchhi DEVANG, Shroff CHINTAN, Tanna GUNJAN, Rai KRUPA
A R T I C L E I N F O
A B S T R A C T
Article history:
Accepted February 2019
Available online April 2019
Artificial intelligence is the simulation of human intelligence processes by machines, which
covers a huge range of capabilities such as voice/image recognition, decision making,
machine learning techniques and semantic search. Through secondary research, this paper
aims at educating marketers on the present and future potential of artificial intelligence by
providing some real-world examples of few early-adopting firms and connecting them with
AI powered technologies that can improve marketing performance and transform their
businesses. Through primary research, this paper aims at studying the impact of artificial
intelligence on the overall marketing landscape. In the end the paper makes a modest
attempt to identify those sectors which have shown good acceptability for AI in marketing
and the ones which will benefit the most.
© 2019 EAI. All rights reserved.
JEL Classification
M31, L86
Keywords:
Adoption of AI by organizations, AI
applications in marketing, Artificial
intelligence, Use of AI applications
1. Introduction
Artificial intelligence can be defined as the simulation of human intelligence processes by machines,
which ranges across capabilities such as voice/image recognition, decision making, semantic search and
machine learning techniques. Artificial intelligence is the next big thing and is expected to create a wave of
disruption in digital technologies.
USA and China dominate the entire AI landscape with investments of $5-8 Bn and $1.5-2.5 Bn
respectively in 2016 followed by Europe with investments of $1.1-1.7 Bn. Although the investment in AI is
growing at the rate of approximately three folds since 2013, overall the adoption of AI remains quite low at
only 20% of the entire market (McKinsey, 2017). "In one of the studies published by IBM Institute for
Business Value, companies which outperformed their peers in various financial measures are more likely to
believe they are ready to adopt cognitive computing (88%) than other companies (57%). They are also more
likely to believe that the technology will be important to their organization’s future (91% vs 64%) and that AI
is mature enough to be market ready (93% vs 59%)" (John Ellett, 2017).
AI deployment will accelerate at the digital frontier which will in turn lead to expansion in the gap
between adopters and laggards across various sectors, industries and geographies. AI is finding its
application across the value chain; however major investments are directed towards the parts of value chain
which are to the core of the company. Peter Drucker once said business enterprise has two--and only two--
basic functions: marketing and innovation. Hence, sales and marketing which generates huge amount of data,
has been one of the critical component of value chain which has got a lot of acceptance of AI (McKinsey,
2017).
AI applications though being used across the value chain around the world, the benefits so provided
by them are yet to be analyzed from an Indian perspective. We are yet to understand how organizations in
India are adopting AI in their daily businesses and whether Indian customers are ready to accept AI.
2. Literature review
Artificial Intelligence defines technologies rising these days which will perceive, learn, so act
supported that information varieties of AI in use these days embrace digital assistants, chatbots, and machine
learning. Today, AI works in 3 ways, a) aided intelligence, wide offered these days, improves what individuals
and organizations are already doing. An easy example, rife in cars these days, is that the GPS navigation
program that gives directions to drivers and adjusts to road conditions, b) increased intelligence, rising these
days, allows individuals and organizations to try to to things they couldn’t otherwise do. For example, the
mixture of programs that organize cars in ride-sharing services permits businesses that might not otherwise
exist. c) Autonomous intelligence, being developed for the longer term, establishes machines that act on their
own. Associate example of this may be self-driving vehicles, once they acquire widespread use.
, , , , K.J. Somaiya Institute of Management Studies & Research, India. E-mail address: krupa.r@somaiya.edu (R. Krupa – Corresponding author)
29
With a market projected to succeed in $70 billion by 2020, AI is poised to own a transformative
impact on shopper, enterprise, and government markets round the world. Whereas there are actually
obstacles to beat, shoppers believe that AI has the potential to help in medical breakthroughs, democratize
pricey services, elevate poor client service, associated even unlock a burdened force. Some school optimists
believe AI might produce a world wherever human skills are amplified as machines facilitate human race
method, analyze, and judge the abundance of knowledge that makes today’s world, permitting humans to pay
longer engaged in high-level thinking, creativity, and decision-making.
As a future with AI approaches, it’s vital to know however individuals consider it nowadays, however
it'll amplify the globe tomorrow, and what guiding principles are going to be needed to navigate this
monumental modification (PwC, 2017).
The applications mentioned below with respective real-world examples help us understand how few
early-adopting firms significantly improved customer satisfaction and customer retention; thereby increasing
their revenue. Thus, we can say that implementation of AI applications leads to significant improvement in
marketing performance and profitability. However, rate of adoption of AI in India is slow when compared to
other countries apart from few sectors like BFSI. BFSI as a sector has seen accelerated adoption of AI in the
last couple of years and is predicted to do so in the future (Accenture, 2017). Banks in India like ICICI have
been able to reduce their turnaround time for processes such as disputed transactions by 80% with the help
of AI (LiveMint, 2017). Most of the AI technologies implemented by firms in BFSI sector can also be used by
firms in other sectors like Retail, Telecom, Media/Entertainment etc. to not only improve their marketing
performance but also transform their businesses.
Table 1. List of AI applications
Name of the
Application
Feature
Case Example
Benefits Obtained
Saffron
Predictive
customer service
USAA tested it to enhance customer service by
effectively foreseeing how a certain customer
may next get in touch with USAA
Effectiveness of their
prediction increased from
50% to 88%
Albert
Marketing
automation
Harley Davidson made use of Albert to
centralize and automate their online marketing
campaigns right from optimization to execution
Dealership leads increased
by 29 times within three
months
Persado
1:1 Dynamic e-
mails
Citi used this cognitive content platform which
combines Machine learning and NLP to
generate exact words and phrases which can
inspire an individual to act
Citi’s e-mail open rate
increased by 70% and click
to open rate by 114%
Conversica
Lead scoring –
Automated sales
assistant software
CenturyLink, a telecom company, used this
application which reaches, engages, qualifies
and follows-up with leads in a two-way email
exchange
CenturyLink’s ROI came up
to $20 for every dollar
invested
Layer 6 AI
Predictive analytics
Deep learning platform for prediction and
personalization of suggestions using
transaction history, customer profile, etc.
Netflix’s AI
recommendation platform
helps them save $1bn
making a significant ROI
Wordsmith
Content generation
Yahoo uses Wordsmith which is a NLG platform
to help convert their data from fantasy football
into detailed match previews, recaps and
reports
Yahoo has added over 100
years equivalent of
incremental audience
engagement by working
with Wordsmith
IBM Watson
Content curation
Wimbledon used Watson which is a cognitive
system enabling new partnership between
people and computers
Watson helped Wimbledon
maintain its digital
viewership and increase
video views by 25% even
after having international
sports activities going on in
parallel
Centiment
Social semantics
Euler Hermes used Centiment to gain insights of
their customers’ feelings about their brand and
product
Euler Hermes’ inbound
leads increased by 38.1%
with projected conversion
of 51.98%
2.1 Research Question/Problem Statement
To understand how is AI disrupting marketing landscape?
To understand if organizations should adopt these AI powered applications for their marketing and sales
or is it just a fad?
30
2.2. Research Objective
The aim of this study is to understand the role AI powered applications played in increasing the sales of
an Indian organization by improving the effectiveness of marketing and sales plan. Towards this, research
objective has been divided into three parts:
i. To study how AI applications help improve 1:1 marketing and move away from mass marketing
ii. To study how AI enhances consumer convenience leading to increasing sales and market share for an
organization
iii. To identify the current adoption of AI applications by organizations in different sectors
2.3. Research Gap and Research Problem
AI applications are being used across the value chain in all organizations of various sectors from
different continents. However, what benefits these organizations have received has to be studied from an
Indian perspective. Should the other organizations be a naysayer and ignore this changing trend or adopt and
be a leader is yet unknown. A comprehensive study needs to be done to know which applications of AI have
been adopted and what effect does it have in transforming the marketing landscape.
2.4. Research methodology
To understand the impact of AI on overall marketing landscape in India, we will be using mixed method
research wherein we will be performing qualitative research (interview and observation) to understand the
perception of organizations towards adoption of AI in India and implications of the adopting AI; and
quantitative research (surveys) for understanding consumer’s acceptability of AI.
3. Qualitative Analysis
Table 2 For the qualitative analysis, data was collected from organizations pertaining to different
sectors
Designation
Domain
Key Comments
Head
Emerging
Markets
Travel Industry
Low acceptance of digitization in India hampers AI’s growth in India
AI is currently more beneficial in customer retention rather than
customer acquisition
BFSI and e-commerce will benefit the most due to abundance of
existing data
“Smart AI, helps search Smart Customer, thus get Smarter Data, thus
make AI smarter”
CEO
Chatbot Platform –
HR Domain
Future of AI in India seems quite patchy and innovation in gathering
datasets should be concentrated on
BFSI and Telecom will be the first to adopt AI because of being under
constant profitability pressure
Telecom is actively tracking our social media data to enhance their
sales by effectively targeting customers for acquisition
Digital
Transformati
on Lead
Pharmaceutical
Industry
Companies in India don’t fully understand what AI is, how it works or
how to leverage its full potential
Some of the reasons for low adoption of AI in India include:
Data from various departments is siloed
Data is not frequently collected and speed of acting on it is not as fast
as it should be
Data collected may not be analyzed for the right insight
Machine learning as a concept has not evolved and companies in India
don’t know what to make out of the huge amount of data collected
Co-Founder
and CEO
Chatbot Platform –
Marketing & Sales
Domain
Most of the large organizations are adopting AI in India in a race
towards innovation
Organizations are not aware of AI’s potential and are therefore
adopting it to test its performance
Most of the Indian banks have adopted chatbots, not strategically but
to test its performance
BFSI and Automotive sector are the frontrunners in the adoption of AI
3.1. Quantitative Analysis
The questionnaire was designed in accordance with the research objectives to understand Indian
consumer’s acceptability of AI based on the following parameters of reference:
i. Willingness to use AI in personalization of services
31
ii. Ease of use of AI
iii. Actual adoption of AI
Personalization of services refers to the customization of services based upon consumer preference while
adoption of AI refers to the consumers’ acceptance of AI.
3.2. Demographics
The survey was conducted amongst the millennials (age group of 22-37). Millennials were chosen for
the survey since they are early adopters of latest technologies and are the future beneficiaries of the same.
There were 302 responses which were obtained out of which 26 were outliers and hence 276 responses were
analyzed. Convenience sampling was used.
3.3. Data Analysis
The data collected was analyzed using SPSS as a tool. The preliminary part of the questionnaire
recorded the information on Age, Gender, Income and the frequency of internet usage while the other
questions aimed at recording their level of agreement or disagreement towards adoption of AI in customer
support services, online shopping websites and virtual assistants.
4. Results and Discussion
4.1. Personalization and 1:1 Marketing
AI helps in providing a more personalized service experience by understanding the consumer
preference and tailoring the service based on the same. From the research point of view before
understanding how much impact adoption of AI in personalization of services would have from an Indian
perspective, it is imperative to understand Indian consumers’ willingness towards personalization of
services.
Hypothesis:
H0A: There is no willingness towards adoption of AI in personalization of services
H1A: There is willingness towards adoption of AI in personalization of services
Indian consumers’ preference towards personalization of experience with respect to e-commerce
websites was used while framing the research questionnaire. This was due to the high degree of familiarity of
the target audience with the same.
Table 3 a Preference for personalization* Purchase from recommendations on e-commerce websites
Crosstabulation
Purchase from
recommendations on e-
commerce websites
Total
Yes
No
Preference for personalization
Y
e
s
Count
24
66
90
% within Preference for
personalization
26.7%
73.3%
100.0%
% of Total
8.7%
23.9%
32.6%
N
o
Count
36
42
78
% within Preference for
personalization
46.2%
53.8%
100.0%
% of Total
13.0%
15.2%
28.3%
Table 3 b Chi-Square Tests
Value
df
Asymp. Sig. (2-
sided)
Pearson Chi-Square
7.720a
2
.021
Table 3c Symmetric Measures
Value
Asymp. Std.
Errora
Approx. Tb
Approx. Sig.
Nominal by Nominal
Phi
.167
.021
Cramer's V
.167
.021
Contingency Coefficient
.165
.021
32
From the above table it can be observed that maximum number of respondents who are currently
not purchasing from the recommendations provided on the e-commerce websites, feel that the
recommendations provided are too generic and should be more personalized.
Table 4 a Usefulness of recommendations provided on e-commerce websites * Purchase attitude
towards websites with virtual trial Crosstabulation
Purchase attitude towards
websites with virtual trial
Most Likely
Likely
Neutral
Usefulness of
recommendations
provided on e-
commerce websites
Useful
Count
18
36
15
% within Usefulness of recommendations
provided on e-commerce websites
24.0%
48.0%
20.0%
% of Total
6.5%
13.0%
5.4%
Neutral
Count
36
51
24
% within Usefulness of recommendations
provided on e-commerce websites
26.7%
37.8%
17.8%
% of Total
13.0%
18.5%
8.7%
Not
Useful
Count
6
9
15
% within Usefulness of recommendations
provided on e-commerce websites
13.3%
20.0%
33.3%
% of Total
2.2%
3.3%
5.4%
Table 4 b Chi-Square Tests
Value
df
Asymp. Sig. (2-
sided)
Pearson Chi-Square
65.927a
16
.000
Table 4 c Symmetric Measures
Value
Asymp. Std.
Errora
Approx. Tb
Approx. Sig.
Nominal by Nominal
Phi
.489
.000
Cramer's V
.244
.000
Contingency Coefficient
.439
.000
From the above table it can be observed that the respondents who are neutral about the usefulness
of the recommendations provided on the e-commerce websites, have shown a higher preference towards
buying online if there is an option of virtual trial provided.
Hypothesis Test Result
Assuming the confidence level of 95%, it can be observed that in both the cases (referring to the
significance value of PHI and Cramer’s V) the significance value is less than 0.05. Therefore, null hypothesis is
rejected. Thus, it can be concluded that there is willingness towards adoption of AI in personalization of
services.
4.2. Customer service
AI can not only help reduce cost at the organizations end but also enhance customer convenience by
automating customer support service. From the research point of view before understanding how much
impact adoption of AI in Customer Support Service applications would have from an Indian perspective, it is
imperative to understand Indian consumers’ comfortability with using automated customer support service
and their preference towards using the same.
33
Hypothesis:
H0B: There is no willingness towards adoption of AI in Customer Support Service applications
H1B: There is willingness towards adoption of AI in Customer Support Service applications
Indian consumers’ preference towards mode of communication with the customer support and
perceived ease of use of each mode of communication was used while framing the research questionnaire due
to the nature of current AI powered customer support service applications (mostly uses texting as a mode of
communication).
Table 5 a Preference for communication * Ease of use for texting over call Crosstabulation
Ease of use for texting over call
Very Easy
Easy
Neutral
Cumbersome
Preference for communication
Phone
Count
15
54
93
30
% within Preference for
communication
7.5%
26.9%
46.3%
14.9%
% of Total
5.4%
19.6%
33.7%
10.9%
Text
Count
18
18
21
12
% within Preference for
communication
24.0%
24.0%
28.0%
16.0%
% of Total
6.5%
6.5%
7.6%
4.3%
Table 5 b Chi-Square Tests
Value
df
Asymp. Sig. (2-
sided)
Pearson Chi-Square
18.367a
4
.001
Table 5 c Symmetric Measures
Value
Asymp. Std.
Errora
Approx. Tb
Approx. Sig.
Nominal by Nominal
Phi
.258
.001
Cramer's V
.258
.001
Contingency Coefficient
.250
.001
From the above table it can be observed that even though people find texting easier than calling, their
preference for using phone is still higher than texting. This can be attributed to the fact that there is higher
emotional connect established through calling due to the presence of a human factor. AI can help integrate
the human factor with texting, making it preferred mode of communication.
Hypothesis Test Result:
Assuming the confidence level of 95%, it can be observed that the Cramer’s V significance value
comes out to be 0.001 which is less than 0.05. Therefore, null hypothesis is rejected. Thus, there is willingness
towards adoption of AI in customer support service applications.
4.3. Virtual Assistants
AI can create a huge impact on how daily chores are carried out by helping automate various day to
day applications. From the research point of view before understanding how much impact adoption of AI in
day to day applications would have from an Indian perspective, it is imperative to understand Indian
consumers’ preference towards using AI in day to day applications.
Hypothesis:
H0C: There is no willingness towards adoption of AI in day to day applications
H1C: There is willingness towards adoption of AI in day to day applications
Indian consumers’ preference towards use of mobile/virtual assistants in day to day applications and
the perceived ease of use of the mobile/virtual assistants was used while framing the research questionnaire
due to the high degree of familiarity of the target audience with the same.
34
Table 6 a Frequency of use for mobile assistants * Ease of use for mobile assistants Crosstabulation
Ease of use for mobile assistants
Very Easy
Easy
Neutral
Cumbersome
Frequency of use for
mobile assistants
Neutral
Count
12
24
36
0
% within Frequency of
use for mobile assistants
16.7%
33.3%
50.0%
0.0%
% of Total
4.3%
8.7%
13.0%
0.0%
Sometimes
Count
3
27
36
24
% within Frequency of
use for mobile assistants
3.3%
30.0%
40.0%
26.7%
% of Total
1.1%
9.8%
13.0%
8.7%
Never
Count
0
15
36
15
% within Frequency of
use for mobile assistants
0.0%
19.2%
46.2%
19.2%
% of Total
0.0%
5.4%
13.0%
5.4%
Table 6 b Chi-Square Tests
Value
df
Asymp. Sig. (2-
sided)
Pearson Chi-Square
143.346a
16
.000
Table 6 c Symmetric Measures
Value
Asymp. Std.
Errora
Approx. Tb
Approx. Sig.
Nominal by Nominal
Phi
.721
.000
Cramer's V
.360
.000
Contingency Coefficient
.585
.000
From the above table it can be observed that even though the respondents were neutral about the
ease of use for the virtual assistants, the frequency of usage for these types of assistants is quite low.
Table 7 a Ease of use for mobile assistants * Preference for usage of virtual assistants Crosstabulation
Preference for usage of virtual
assistants
Most Likely
Likely
Neutral
Ease of use for mobile
assistants
Very
Easy
Count
24
0
6
% within Ease of use for
mobile assistants
80.0%
0.0%
20.0%
% of Total
8.7%
0.0%
2.2%
Easy
Count
21
42
12
% within Ease of use for
mobile assistants
25.9%
51.9%
14.8%
% of Total
7.6%
15.2%
4.3%
Neutral
Count
27
42
18
% within Ease of use for
mobile assistants
23.7%
36.8%
15.8%
% of Total
9.8%
15.2%
6.5%
Table 7 b Chi-Square Tests
Value
df
Asymp. Sig. (2-
sided)
Pearson Chi-Square
157.652a
16
.000
35
Table 7 c Symmetric Measures
Value
Asymp. Std.
Errora
Approx. Tb
Approx. Sig.
Nominal by Nominal
Phi
.756
.000
Cramer's V
.378
.000
Contingency Coefficient
.603
.000
From the above table, it can be observed that the respondents who found virtual assistants easy or
were neutral about the easiness of use, were more likely to adopt these types of assistants for practical
repetitive tasks which can be automated.
Hypothesis Test Result:
Assuming the confidence level as 95%, it can be observed that in both the above cases the Cramer’s V
significance value comes out to be less than 0.05. Therefore, null hypothesis is rejected. Thus, there is
willingness towards adoption of AI in day to day applications.
5. Managerial Implications
This section describes how AI can assist marketers and which applications of AI can be used for what
purpose. As we infer from the quantitative inference that there is a gap between willingness and actual
adoption of AI. In order to bridge this gap AI helps to move away from mass marketing and get towards 1:1
marketing.
AI can help marketers segment targets more accurately. Knowing segments accurately will help them
make content more personalized for each individual. From the survey conducted by Forrester, 57% would
shop more if a feature of personalization were made available which is supported with 73%, which we
obtained from our primary research. This was from customers’ point of view, coming towards business
executives, 61% of them feel that AI can offer a superior one-to-one personalized experience. Example: AI
application Persado writes dynamic emails.
Another way of filling this gap is by taking action in real time AI delivered insights. Example: IBM
Watson, Albert helps in everyday planning, executing and reporting on campaigns. It helps in everything from
optimization to execution.
Consumer Convenience
In order to address the issue of enhancing the consumer convenience, AI plays a critical role. AI can
help in building emotional connect, though it is just a machine. For example, Centiment is a social semantics
tool which helps marketers gain insight into the emotions of the customers thereby helping them understand
how customers feel about their brand and product. It uses NLP, which is a component of AI to do so.
Content Generation
(NLG) is an AI application, which can generate stories on its own using existing sources of
information like scores & names of players of Fantasy football club. This brings convenience to user as they
see stories on topics they love that too with little efforts from marketer’s end, thereby proving superior
consumer convenience. Thus, as marketers it becomes necessary to make maximum use of data to bring
personalization and enhanced consumer experience.
Lead nurturing
Conversica (Automated Sales Assisting Software) is an AI driven software that automatically reaches,
engages, qualifies and follows up with leads via a two-way email exchange.
Data Governance
With the increase in the data which is collected, AI can facilitate in generating bigger insights of the
massive information by serving to perceive deeper levels of shopper insights and the way to effectively build
shoppers have interaction. Advanced systems and advanced analytics can amendment the approach we tend
to expertise data via our mobiles, our wearable and therefore the net of Things (IoT).
Sectors Identified
BFSI and Telecom will be the first to adopt AI, or as a matter of fact any kind of innovation since they
are constantly under profitability pressure.
6. Conclusion
While there are several challenges to overcome, AI possesses the potential to solve many of today’s
problems and push Martech even further. This change can only be brought about if individuals and
businesses co-create a man-machine hybrid which entails to be more powerful than either entity acting alone.
The government of India has taken proactive action towards digitization and is driving the AI agenda
as well – from setting up a policy group for AI by the MEIT, to the Karnataka government’s recent
36
announcement of a data science and AI Centre of Excellence. NITI Aayog will be establishing a national
program in the area of AI which will include research and development of its applications.
Altogether, these developments will eventually propel the evolution of an AI ecosystem and its
applications in our day to day lives. Soon we will be witnessing ‘augmented age’. To remain competitive in the
current industrial scenario, it is necessary for the companies to adopt AI.
References
1. Accenture, 2017. [Online], Available at: https://www.accenture.com/us-en/perspectives-banking-tech-vision-2017-
mcintyre-westland
2. Automated Insights, 2017. Natural Language Generation | Yahoo! - Automated Insights, Inc.. [Online], Available at:
https://automatedinsights.com/case-studies/yahoo
3. Centiment, 2017. Centiment & Euler Hermes: Risk & Social Application Case Study. [Online], Available at:
https://centiment.io/beta/blog/wp-content/uploads/2017/06/EulerHermesCaseStudy.pdf
4. IBM, 2016. Wimbledon 2016. [Online], Available at: http://ecc.ibm.com/case-study/us-en/ECCF-WWC12371USEN
5. LiveMint, 2017. [Online], Available at: http://www.livemint.com/Industry/Vu7w0N8yUaCqsJwGkyS0GN/Private-sector-
banks-looking-to-leverage-robotics-AI.html
6. McKinsey, 2017. Artficial Intelligence the Next Digital Frontier?, s.l.: McKinsey & Company.
7. Anon, (2017). [online], Available at: http://www.martechadvisor.com/articles/customer-experience/recommendation-
engines-how-amazon-and-netflix-are-winning-the-personalization-battle, [Accessed 3 Nov. 2017].
8. Forbes.com. (2017). Forbes Welcome. [online] Available at: http://www.forbes.com/sites/johnellett/2017/09/13/ibm-
study-finds-outperforming-companies-more-ready-to-adopt-ai-for-marketing/#b2a9c3a5952e, [Accessed 3 Nov. 2017].
9. AI software for marketing & sales | Conversica | Convert more leads into opportunities. (2017). Customer Success Story:
How CenturyLink gets $20 return for every $1 spent with Conversica. [online], Available at:
http://www.conversica.com/customer-success-story-how-centurylink-gets-20-return-for-every-1-spent-with-
conversica/, [Accessed 3 Nov. 2017]
10. Haworth, E. (2017). Meet Albert, the AI that's better at your marketing job than you. [online] Cityam.com, Available at:
http://www.cityam.com/268916/meet-albert-ai-thats-better-your-marketing-job-than-you [Accessed 3 Nov. 2017].
11. Harvard Business Review. (2017). How Harley-Davidson Used Artificial Intelligence to Increase New York Sales Leads by
2,930%. [online], Available at: http://hbr.org/2017/05/how-harley-davidson-used-predictive-analytics-to-increase-
new-york-sales-leads-by-2930 [Accessed 3 Nov. 2017].
12. TechRepublic. (2017). Persado uses artificial intelligence to help digital marketers write better copy. [online], Available
at: https://www.techrepublic.com/article/persado-uses-artificial-intelligence-to-help-digital-marketers-write-better-
copy/, [Accessed 3 Nov. 2017].
13. imFORZA. (2017). Online Marketing with Artificial Intelligence | imFORZA. [online] Available at:
https://www.imforza.com/blog/online-marketing-with-artificial-intelligence/ [Accessed 3 Nov. 2017]
14. PWC, 2017. Bot.Me: A revolutionary partnership. [online], Available at:
https://www.pwc.com/consumerintelligenceseries, [Accessed 3 Nov. 2017].
