Chapter 5
IMPLEMENTATION & RESULTS
Provides a succinct overview of the interview results in this chapter. The qualitative
character of this study necessitates a methodical examination of participant data in
order to spot recurrent patterns, as was indicated in the research systems section. The
themes in the participants' answers are used to create the headers. Using the same
terminology as the participants ensures impartial information presentation Bargain, P.
(2018) [61]. The thesis has completed the necessary steps to obtain relevant data for
this research and is now ready to map and compare these values.. The survey looked
into how familiar respondents were with AI and marketing automation. The practice
part provides specifics on how the research composed data after six individuals in
total. The participants were divided into two collections of three people each, based
on their occupations. The meeting directors were arranged and categorised based on
the importance of each section of the two data sections that were provided beforehand
for the inquiry objects.
The main conclusion emphasises how AI technology greatly improves marketing
strategies by quickly delivering client preferences and processing customer data
efficiently. A. M. Abdulazeez (2021) [62]. Over time, marketers hope to become
experts in both marketing analysis and the real-world use of artificial intelligence. As
a result, marketers are more likely to run effective campaigns if they understand the
underlying AI technology and how it relates to the marketing tools they are using.
Companies can use AI to build models that can adapt to the rapid changes in digital
commerce and to better grasp the complex links between information. Furthermore,
decision models can now more accurately and thoroughly predict events—even ones
that haven't happened yet—thanks to advances in AI and machine learning. In
addition to facilitating one-on-one interaction, this offers firms significant financial
benefits.
5.1 The influence of artificial intelligence on the evolving roles and
advancements in marketing
74
The conversation explores the application of artificial intelligence (AI) in marketing
and the corresponding obligations for marketers. It delves into the mechanisation
process's authoritative facets of novelty and collective consciousness as well. The
subsequent set of questions seeks to understand marketers' perspectives on integrating
artificial intelligence into their marketing endeavours Mari, A. (2019) [64]. One
marketing participant boldly stated, "It's going to revolutionize the whole thing and
become an essential part of our reality within the next 25 to 30 years." Another
participant emphasized the optimistic influence of AI on their work, highlighting its
ability to tailor products and services to individual customer preferences. AI
technology represents a significant enhancement in the realm of digital marketing,
with numerous benefits, including improved communication systems, a deeper
understanding of customer preferences, and expedited product delivery Marr, B.
(2019) [65]. Furthermore, this study, in line with previous research, underscores the
effectiveness of AI in providing swift feedback, enabling marketers to refine their
foodstuffs and ensure customer gratification.
Once requested about their dream for the upcoming of AI, participants expressed
enthusiasm, foreseeing remarkable improvements in marketing strategies. Future
campaigns and strategies are expected to leverage data from previous endeavours and
AI technology to gain a comprehensive understanding of market dynamics and
consumer preferences. According to one respondent, AI is poised to bring about
transformative changes across the marketing landscape, enabling personalized
customer interactions and emotional understanding & Zook, Z. (2019) [66].
Anticipations are high that AI will provide marketers with insights into customers'
emotional attitudes towards products, offering a profound understanding of
preferences from both psychological and emotional standpoints. This, in turn, will
equip marketers with valuable insights to enhance their marketing strategies.
Ultimately, the overarching goal of marketing is customer satisfaction, and access to
customer information is paramount in achieving this objective. AI technology is
anticipated to meet this requirement effectively, making marketers' tasks more
manageable. Marketers reported that their companies have successfully integrated AI
technology into their marketing systems, leading to significant improvements in
operations & Aditya, S. (2016) [68]. The AI technologies predominantly utilized by
these businesses include Google AI, Machine Learning, Python, AI Integration, and
75
Hub Spot. The respondents affirm that AI has streamlined marketing efforts, aligning
with their expectations. Marketing goals have, to a certain extent, been attained as a
result of AI implementation. As one marketer aptly put it, "We've incorporated AI
into our marketing strategies from the outset, and it has consistently improved as it
studies from statistics. Our marketing has been successful since our customers
understand our value proposition." This underscores the growing necessity for
marketers to embrace technology, including AI, to stay at the forefront of modern
digital marketing practices. AI is increasingly regarded as a facilitator of human
thought processes within the technological landscape, positioning tech-savvy
marketers for success in this evolving domain.
5.2 The Significance of AI in Marketing
According to According to the marketing interviewees, AI offers significant growth
potential by facilitating learning from processed data and saving time. In the domain
of contemporary marketing, a deep comprehension of clients' preferences and
interests, along with the capability to promptly and efficiently act on this information,
holds significant importance. Companies that do not harness AI in their marketing
endeavours often face challenges in making immediate, data-driven decisions.
Beyond automating tasks, the most prominent benefit of AI lies in the heightened
relevance it empowers marketers to provide to their clients Abhishek, C. M. (2016)
[70]. In today's world, consumers are inundated with a vast amount of information
and choices. Personalization, driven by the collection of extensive data, equips
businesses with the ability to gain deeper insights into consumer behaviour and
interests across various platforms and touch points. This, in turner, enables the
alignment of customer interests with pertinent content, ultimately boosting sales.
Consumers have become increasingly discerning and less patient, especially when
they engage with brands during their whole e-commerce experience, from awareness
and investigation to assessment.
5.3 Implementation Result
Data Science (DS) offers a multitude of opportunities and methodologies for
statistical data analysis. Statistics provide a comprehensive set of criteria for
evaluating various data types, forming the theoretical foundation for Data Science
(DS) with Machine Learning (ML). It encompasses the development of arithmetic and
76
the development of DS, fundamentally rooted in statistical learning. This foundation
underpins three key aspects within the DS domain: functional analysis, exploratory
analysis, and result prediction based on market analysis Scott, J. (2018) [71]. These
methodologies are applied to analyse the data, serving as the core DS methods for
statistical learning.
Following a comprehensive exploration of the primary methods for investigating Data
Mining (DM) through DS approaches, we have uncovered the extensive uses and
applications of DS in tactical DM strategies and developments. One of the key
challenges within the DM Ecosystem is defining and controlling the success of a DS
strategy Wilson, D. (2016) [73]. Marketers and researchers are tasked with selecting
and comprehending the primary performance measures for assessing the effectiveness
of the employed approaches and models. Our Systematic Literature Review (SLR)
study has identified several key indicators for evaluating the successful deployment of
these DS strategies.
The primary research topics within DM employing DS are illustrated in Figure 2.
These topics were selected based on suggestions from the reviewed articles and
discussions on future field research. Each of these themes and their impact on the
development of DM methods utilizing DS is thoroughly explored. They encompass
the following: Medical Data Analysis: These analyses are crucial for identifying
trends, facilitating the innovation of new vaccines, and combating uncontrolled
epidemic infections, such as COVID-19, as well as predicting potential fatalities.
Medical Data and Health Strategies: Supporting organizations in collecting data for
future research is imperative for the advancement of health strategies. Smart City and
Governance: The automation and integration of artificial intelligence into large-scale
systems lay the foundation for efficient energy resource management, sustainable and
intelligent urban development J. G. (2017) [75]. Direct marketing communications
through digital platforms and social media, along with corporate social responsibility,
play a significant role in advancing social and responsible marketing.
Articles identified from Articles
excluded
Database using keywords after analysing title
77
Digital Marketing and abstract and
keyword
Data Science (n=84)
Articles excluded after analyzing research study objective, relation with research topic
description and specification of terms (n=140).
Databases Used
Total Identified
(ACM Digital Library, articles (n= 381)
IEEE Explore, Web of
Science, Science
Direct)
Potential
Identified Articles
(n=270)
Total Articles uArticles excluded
after analyzing research study
objective, relation with research
Figure 5.1 Flow chart for PRISMA SLR
Type of study Explanation
Evocative To condense the gathered information or
information datasets, descriptive statistics come into play.
These encompass metrics like variance and
range, as well as central tendency and
arithmetic mean measurements.
Byes’ law Bayer's rule is employed to characterize the
probabilities of events. This rule is rooted in an
understanding of the circumstances that may
have contributed to a specific occurrence.
Minimum The least squares method empowers you to
squares discern the ideal theoretical model comprising
method variables or structures that align with your
78
dataset and facilitate quantitative validation.
Linear A linear regression model explores the
weakening association between a scale and an independent
mutable. When there are multiple predictor
variables within linear regression, it is referred
to as multiple linear regressions.
Logistic Logistic regression is a reversion analysis
weakening technique used to forecast a mutable within
categories. When a variable is divided into
multiple categories, it becomes a categorical
variable. The fundamental principle for
modelling event probabilities involves
considering the components that contribute to it.
Network of Artificial neural networks are independent
reproduction knowledge systems composed of interconnected
neurons neurons with input and output nodes. They are
employed to identify solutions or characteristics
that are challenging to discern through
conventional programming.
Analysis of Multivariate analysis is an analytical approach
multivariate that concentrates on the examination of diverse
data involving multiple dependent variables.
Multivariate analysis finds applications in
variable and correlation analyses.
Estimate Maximum probability estimation is a
maximum statistical parameter estimation method that
probability
determines the highest probability estimate
within an observation. The outcomes provide
evidence of the most likely parameter values
under consideration.
Analyzes of The identification of discrimination is referred
discriminatio to as categorization or pattern recognition. It
n automatically detects patterns or regularities
within a dataset and is extensively employed for
research in Data Mining (DM) or Knowledge
Discovery in Databases (KDD) S. (2017) [81].
Sequence The theory of information is applied to analyse
Theory and assess quantifications and communication
to define the boundaries of communication
processing.
Table 5.1 Methods used for digital marketing in data science
Category Explanation
79
Content Analysis of publicly available content on social
analysis media and digital platforms encompasses the
generated by examination of content such as Tweets, reviews,
the user comments, information sharing, likes, metadata,
and more.
Customer Enhancing user preferences through the analysis
preferences of extensive databases, including consumer
optimize preferences for products or services.
Customer Digital channels enable the monitoring and
Control prediction of customer behaviour, allowing
Online companies to anticipate and gain a deeper
Tracking understanding of their marketing and
advertising efforts.
Tracking Monitoring interactions and responses between
comments/int users and companies within digital ecosystems.
eractions in Marketers can then deliver more tailored
social media messages based on the results.
Stock levels Enhancing stock optimization on e-commerce
on websites to efficiently manage sales
ecommerce information, products, and services, as well as
websites to proactively request supplies from suppliers..
should be
optimized.
Online Sales analysis aids in uncovering patterns that
selling data elucidate current trends, customer demand, and
analysis interest in specific products.
Presentation Conducting market research to identify
of new emerging trends that can lead to the
products introduction of new products within existing
sectors.
Trend Engaging in market research to discover
analysis of emerging trends that have the potential to
social media introduce new products within established
sectors.
Product Consumer reviews evaluate essential aspects of
analysis companies, such as products, services, and
recommendat customer service features, and offer
ions and recommendations Kumar, E. P. (2021) [83].
reviews
Customizatio Enhancing customer purchasing experiences by
n of client delivering personalized content based on the
experience historical data of website visitors.
online
Recommenda Creating automated recommendation systems
tion of that rely on analysing the purchase history of
building specific customers to prioritize successful
systems purchases and make predictions.
80
Clicks online Utilizing clicks on websites and social networks
measure and to predict user behaviour. With this
predict information, companies can optimize their
advertisements for increased profitability and
enhance the visibility of impressions and clicks.
Real-time Facilitating direct decision-making in specific
data analysis sales, visibility, or distribution scenarios.
Identify Identifying leaders within online communities
communities and analysing user loyalty to opinion leaders
online and their messages.
Table 5.2 Mean uses in digital marketing techniques of data science
Pointer Explanation
Dependabilit Precision, often mentioned to as accuracy,
y signifies the superiority and precision of the
model or procedure's hits.
Accuracy Accuracy is a measure of the effectiveness and
success of a method or approach when used in a
data source in Data Science
Soundness Cogency is the statistical amount that assesses
whether the results and findings are valid and
accurate, providing support for them.
Consistency Consistency assesses whether the standards in
one dataset are concurrent with the values in
another dataset at the same time.
Remember Recalling the number of precise findings
divided by the number of dismissed values is
also known as sensitivity.
Compassion Compassion, often referred to as true positive
rate, measures the ratio of positive values or
dataset points.
Specificity Specificity is the metric that assesses the
predictive properties of actual negatives in
variables within the data category.
Commonness Occurrence relates to the amount of people
with specific shared qualities during a certain
period.
Table 5.3 Mean performance measurements to measure DS success in DM
81
Enhancing Public Mobility involves personalized evaluation of people's movements
and organizational patterns, utilizing big data from citizens or vehicle purchases. This
task proves challenging for Digital Marketers (DMs) aiming to customize mass
communications and discern specific behavioural patterns using Data Science (DS)
techniques, relying on demographic and psychographic characteristics to optimize
Return on Investment (ROI) for digital advertising Sarvary, M. (2011) [86]. The
development of novel machine learning models is paramount. Companies can
incorporate these models into their projects to address specific challenges.
Constructing, training, and refining these models are essential for achieving specific
objectives. Furthermore, the creation of user-friendly algorithms and model libraries
is vital to simplify technical complexities for marketers and data professionals.
Another key aspect is Operating Customer Relationship Management (CRM) systems.
This encompasses the establishment of automated Information Management (IM)
systems that enable real-time data handling for organizations, facilitating more
informed decision-making by marketers. Among these methods, CRM stands out as
the most effective approach.
Medical and Health Data
Smart City and Governess
Public Mobilization
New
Operating CRM Machine
Learning
Sustainable Data base
strategies
Social Media Listening
IoT Devices
Figure 5.2 Research Areas that can be used to improve the industry performance
using DS and Digital Marketing
82
Fostering Sustainable Data-Driven Approaches: This involves the examination of data
sources and the management of globalization to bolster sustainable strategies and
data-centric initiatives. Within this realm, areas such as Social Marketing and Green
Marketing hold significant relevance.
Harnessing the Power of Social Media Monitoring: Automated analysis delves into
the ever-evolving trends within social networks and messaging platforms, tracing the
responses of opinion leaders and communities to significant messages, particularly
those related to environmental and social movements or in response to disasters and
diseases De Bock, K. W. (2020) [89]. It is crucial for Digital Marketers (DMs) to
grasp the dynamics and effectiveness of these communities when delivering
compelling and responsible messages. Leveraging these diverse research domains can
substantially enhance industry performance through the synergy of digital marketing
and data science.
5.4 Experimental Results of Digital Marketing Pre-Processing Stage
The research study will initiate by validating the research theories through a PLS-
SEM examination, which includes dimensional fit testing, model robustness
assessment, and precision evaluation. Subsequently, an assessment of the digital
advertising performance index will be carried out.
The hypotheses under investigation in this training are delineated as follows:
Hypothesis 1: The readiness of the novelty bionetwork has a important influence on
numerical alteration.
Hypothesis 2: The willingness within the novelty bionetwork significantly influences
the adoption of digital marketing.
Hypothesis 3: The willingness within the innovation ecosystem significantly affects
the effectiveness of digital marketing capabilities.
Hypothesis 4: Digital alteration has a significant impact on the acceptance of digital
advertising.
Hypothesis 5: Digital alteration significantly influences the capacity of digital
marketing.
Hypothesis 6: The acceptance of digital marketing significantly impacts the capacity
of digital marketing.
83
Figure 5.3 Machine Learning use cases in Marketing
84
Table 5.4 Area of the impact of AI on marketing mix source
The authors scrutinized the collected instances and distilled their observations
regarding how these cases align with the marketing framework. These insights have
been systematically organized and are now presented in Table 1. Each validated case
underscores the profound impact of AI on all aspects of marketing strategies. This
insight holds particular significance for marketing professionals who are at the
forefront of pioneering new approaches, as AI's influence permeates every facet of
marketing activities Hinson, R. E. (2020) [90]. Furthermore, the incorporation of AI
to enhance core products necessitates a fresh and innovative design perspective,
facilitating the generation of ideas that transcend not only product boundaries but
entire categories. This emerges as a critical element in the integration of AI within the
domains of "product" and "promotion" in the marketing mix program.
Figure 5.4 More accurate result we can find in less learning rate in
comparison
85
Table 5.5 Training Time and Accuracy of the proposed AI model
5.5 Several Segments for AI Applications in Digital Marketing
Figure 5.5 Artificial Intelligence for Marketing
86
Figure 5.6 Rework of the category system and findings into structure
5.6 A Machine Learning Approach for Consumer Demand
Forecasting
In this paper presents the results of the implemented framework, showcasing the
accuracy of demand forecasts compared to traditional methods. Insights into the
adaptability of the model to different market conditions and its efficiency in real-time
decision-making are discussed. Sample Data Set containing Telco customer data and
showing customers left last month.
Figure 5.7 Senior Citizen is actually categorical Target Value and count
87
A. Data Exploration
Plot distribution of individual predictors by churn
Figure 5.8 Univariate Analysis gender and count
Figure 5.9 Correlations between Strumming TV and Strumming Movies
88
Figure 5.10 Correlation between contract and Paperless billing after Testing
Figure 5.11 Relationships between Monthly Charges and Total Charges
89
Figure 5.12 Density and Monthly Charge between customer and Products
This comprehensive study delves into the realm of Digital Marketing and employs
Decision Tree Algorithm concepts through training and testing to comprehensively
evaluate AI-driven cost strategies. The investigation extends its focus to the logistics
delivery industry, shedding light on how fake data can significantly enhance cost
strategies. These findings contribute valuable insights to the existing knowledge base,
affirming the potential of AI to enhance cost analysis methodologies. Beyond this,
ML classification emerges as a powerful tool for predicting the costs associated with
producing specific products or services [15]. Leveraging data such as price opinions
and manufacture costs, the algorithm estimates and informs decision-makers about the
expenses linked to the offered product or service. This predictive information plays a
pivotal role in guiding pricing decisions, ensuring the company's competitiveness
within the market.
5.7 Algorithms
The future AI algorithm is to categorize the Customer data to achieve with the
prospect technique. The intended method is analysed through the track superiority
factor to be old to study its staging.
Step 1: Load Data
time_series_data = load_time_series_data()
Step 2: Preprocessing
preprocessed_data = preprocess_data(time_series_data)
Step 3: Determine Model Parameters
90
p, d, q = choose_arima_params(preprocessed_data)
P, D, Q, s = choose_seasonal_params(preprocessed_data)
Step 4: Initialize SARIMAX Model
model = initialize_sarimax_model(order=(p, d, q), seasonal_order=(P, D, Q, s))
Step 5: Fit SARIMAX Model
fitted_model = fit_sarimax_model(model, preprocessed_data)
Step 6: Model Evaluation
evaluate_model(fitted_model)
Step 7: Forecasting
future_forecast = forecast_future_values(fitted_model, steps=n_steps)
Step 8: Model Refinement (if needed)
refined_model = refine_sarimax_model(fitted_model, preprocessed_data)
Step 9: Final Forecasting
final_forecast = forecast_with_refined_model(refined_model, steps=n_steps)
End
91
5.8 Performance Comparison of the Current result and previous
results
The best performance achieved in Digital Marketing using the dataset is 89.30%. This
remarkable result signifies a high level of success in leveraging data for digital
marketing purposes. Such a performance level indicates effective data-driven
strategies and potentially superior marketing campaigns when compared to lower-
performing attempts. Achieving an 89.30% success rate can be considered a
significant accomplishment in the field of digital marketing, suggesting that the
dataset and associated methodologies have been highly effective in delivering desired
outcomes. This result sets a benchmark for future digital marketing efforts and
emphasizes the importance of data analysis and its impact on campaign success.
Digital marketing using Artificial Intelligence (AI) can yield various results and
benefits for businesses and marketers. AI has transformed the digital marketing
landscape by providing advanced tools and capabilities for more efficient and
effective campaigns. Here are some of the key results and advantages of using AI in
digital marketing: Improved Personalization: AI allows for highly personalized
marketing campaigns. It can analyse user data to understand individual preferences
and behaviour, enabling businesses to tailor their content and offers to specific
audiences. This leads to higher engagement and conversion rates. Enhanced
92
Targeting: AI can segment and target audiences with precision, ensuring that you’re
marketing efforts are directed at the right people. This reduces ad spend wastage and
maximizes ROI.
Predictive Analytics: AI can predict trends and consumer behaviour, helping
businesses stay ahead of the competition. It can also forecast sales and identify
potential areas for growth.
Customer Support: AI-powered provide instant responses to customer inquiries,
improving customer support and user experience. They can handle routine questions
and tasks 24/7.
Content Generation: AI can assist in creating content, including writing blog posts,
generating social media updates, and even designing visuals. This can save time and
resources while maintaining quality.
Optimized Advertising: AI can automatically optimize advertising campaigns by
adjusting budgets, bidding strategies, and targeting in real-time. This maximizes the
performance of ads across various platforms Bressgott, T. (2020) [92].
A/B Testing: AI can hurry A/B testing processes by rapidly analysing data and
making recommendations on what elements to test and adjust Liu, J. (2018) [94].
Email Marketing: AI can improve email marketing by personalizing subject lines
and content, sending emails at the best times, and segmenting email lists effectively.
Search Engine Optimization (SEO): AI tools can analyze keywords, monitor
website performance, and recommend SEO improvements to boost search engine
rankings.
Data Analysis: AI can procedure and examine vast amounts of data, if valuable
insights for decision-making and campaign optimization.
Fraud Detection: AI can help identify and stop ad fraud, safeguarding that your
marketing cheap is not wasted on fake clicks or impressions.
Customer Insights: AI can assistance trades gain a deeper sympathetic of their
customers' preferences, behaviors, and needs, allowing for more effective marketing
strategies.
93
Social Media Management: AI tools can schedule posts, curate content, and monitor
social media trends, making it easier to maintain an active and engaging social media
presence.
Lead Scoring: AI can help prioritize leads by assigning scores based on their
likelihood to convert, allowing sales teams to focus on the most promising prospects.
Content Recommendation: AI-driven content recommendation engines can increase
user engagement by suggesting relevant products or articles to website visitors.
The use of AI in digital marketing can lead to increased efficiency, higher conversion
rates, better targeting, and a deeper understanding of customers Day, S. (2019) [100].
The specific results will vary depending on the strategies and tools employed, but AI
has the potential to transform and optimize various aspects of digital marketing.
94