Boosting E-commerce Sales: Data-Driven Strategies for Profitability

This report delves into the application of data analytics to enhance e-commerce profitability. Using a dataset of book sales, we explore data mining techniques, including clustering and linear regression, to identify key areas for improvement. The analysis reveals specific geographic regions and product categories that require strategic focus. Recommendations are provided for optimizing sales, including targeted marketing campaigns, free shipping strategies, and data-driven inventory management. The report also includes Python code examples and visualizations to illustrate the analytical process.

ICT 706

Executive Summary
The platform which allows organizations to do trading of their products through an online portal
is called e-commerce or electronic commerce. E-commerce offers so many services like selling
products, online reservations, transactions and so much more.
The main target of this assignment is to cover all the information that is related to e-commerce
and to analyse the methodology used in this assignment.
SWOT is the most effective way to make strategic plans for the profit for the organization. It
could be done by SWOT examination process, all the information regarding the profit domains
could be gathered by process.
The miles have an insight building which is known as shopper driver that is very beneficial. This
is useful in increasing the association execution. For your business endeavour, Improvement in
the site design and the substance should be facilitated.
The main aim of organization should be that they advertise their product in such a way that the
customer finds it easy and simple and they think that they actually need it. The costs of delivery
should not be as high or after a certain amount, it should be free of cost which lures the
customers to get to that amount.
1

Contents
Executive Summary.........................................................................................................................1
Background......................................................................................................................................3
Introduction......................................................................................................................................5
Research Methodology....................................................................................................................6
Descriptive Research....................................................................................................................6
Fundaments available...................................................................................................................6
Quantitative or Qualitative...........................................................................................................6
Conceptual or Empirical..............................................................................................................7
Analytical Findings:.........................................................................................................................8
Dataset..........................................................................................................................................8
Data Mining.................................................................................................................................9
Divisive method:........................................................................................................................11
Agglomerative method:..............................................................................................................11
Recommendation for the Company:..............................................................................................14
Implementation plan:.....................................................................................................................15
Python Codes with Screenshots.....................................................................................................16
Conclusion:....................................................................................................................................20
References......................................................................................................................................21
Appendex.......................................................................................................................................22
Working Python Codes..............................................................................................................22
2

List of Figures:
Figure 1: Sales Dataset....................................................................................................................9
Figure 2: Training Graph over Sales Data.....................................................................................10
Figure 3: Test Graph over Sales Data............................................................................................10
Figure 4: Output Prediction...........................................................................................................12
Figure 5: Product to be prioritized.................................................................................................12
Figure 6: Linear Regression Graph on Sales................................................................................16
Figure 7: Final Prediction Graph...................................................................................................16
Figure 8: Code for test and train....................................................................................................18
Figure 9: Main Module..................................................................................................................19
3

Background
It was long time ago when this organization was set up. The time when no one ever heard of this
organization before as it was new and had just started and as a newbie, they could only reach to a
very fewer region near them and that too with very fewer products. As the time passed this
organization started to get recognition and started to reach regions a bit far thus they started to
expand their business as wide as possible with more products in their hand and this was the time
when it was growing every day.
Their products varied from toys, gadgets, books to so much more and they also delivered all
these products to the doorsteps of their customers and with so many other factors they became
one of the top organizations of that time. Then there came the time when this organization too
had to go through bad graphs of sales and this was the time when they needed to work on
improving their strategic goals and it was all made possible when they used Data Analytics
which actually made me regain their position in the market thus they were once again leading the
market. Data Analytics gave them an exact graph of which regions were the one where the sales
graph was going downhill and where the organization needed to improve also from which region
there was the highest profit which helped them make their way to top.
4

Introduction
The current situation is like this that every organization wants to go e-commerce. So many
organizations are competing with each other to provide the best e-commerce experience. A
report will be made in this assignment which will focus on improving the profitability
of organization. To calculate the monthly sales of book is the main aim of organization.
Complete sales of products in the last month is recorded in report.
The aim of this assignment report is to focus on the techniques that will help the
organization increase its sales. The Technique widely used is Python programming language.
Techniques for producing the result are also present in the report (Turban et al., 2017).
5

Abbreviations and Assumption List
The abbreviations used in this report are:
 Tab Separated Value
 Comma Separated Value
Assumptions made for assignment are as follow
 Test dataset development for process
 Training set and test set are two parts of Test dataset
 Geographic location is an addition
 Unique product available
 Delivery person required
6

Research Methodology
The process used to identify, select and analyze information and gather all the data available
about a particular topic. Research is done to improve the currently existing data or gather a new
one with the updated information. After a research is done it is easy to compare, observe and
experiment (Andam, 2017).
Descriptive Research
It is the research that includes all the surveys and all the inquiries about the facts that are being
found. This research basically focuses on what has happened and is happening at the moment. In
these research things like preferences of people, the frequency of shopping is being measured.
While on the other hand, the analytical research works on the data that is available at the current
time for more critical evaluation using it
Fundaments available
The research which works on the solution quickly to resolve the problem for the organization.
The research related to fundaments deals with theory.
Quantitative or Qualitative
Quantitative research deals with the measurement of quantity. Works with anything that may be
related to quantity in any way. Qualitative research, on the other hand, deals with the quality of
the subject which includes the science of behaviour in which research
about behaviour of human beings is being carried out. In comparison, the qualitative research is a
bit difficult to carry out and requires psychologist's guidance.
7

Conceptual or Empirical
Conceptual research is the one which mainly focuses on conceptual ideas. All the philosophers
before developing any idea or thought take conceptual research into consideration for better
results. Empirical research, on the other hand, focuses on the data that is being obtained by
experiments or observations. It works according to the data that is being verified by experiment
or observations.
8

Analytical Findings:
This assignment will be using a dataset of books. Information of the books sold by the
organization is in the dataset.
Dataset
The dataset includes:
 Books ID,
 Books Name,
 Type of Customer,
 Book Price,
 Monthly Sale,
 Geographic Region for Delivery,
 Number of Customers Who Bought the Book,
 Language,
 Shipping Type
9

Figure 1: Sales Dataset
Figure 1 shows the Sales Dataset Chosen for the process of prediction. The above figure shows
the part of the dataset which includes the metadata in a dataset of books.
Data Mining
Data mining is the strategy that organizations use to turn their unrefined data into the valuable
one. By using additional assistance in the form of programming, illustrations which help in a
very strong way to refine the data.
Data mining helps in choosing the arrangements and conveys high acquiring. It is used by
undertaking partnerships of net business. It depends on data amassing and setting ceaselessly.
The aim of this is to grab records from dataset and pass on it to such an extent in such a way
that it becomes sensible. Data mining helps by providing the ways to conquer loss and gain
profit. Data mining helps to focus on which areas does improvement is needed, what regions
10

must be focused to improve the business, what if detached transporting methodology can be
situated in real ways of life (Wu et al., 2014).
Figure 2: Training Graph over Sales Data
Figure 3: Test Graph over Sales Data
11

The output result of the data mining is in advance with the feeling body. Get-together joins the
dataset having a fundamentally indistinct sort of estimations. Different levelled Clustering directs
encircling packs that have a destined requesting for through and through. It is overseen in
structures:
Divisive method:
This strategy is in like way thought about as top-down where every conviction is chosen to an
individual built up arrange and after that opening the altogether finished. Into two alike
foundations. In examination with agglomeration, strategy offers the greater part of the more bona
fide inclines regardless while the idea can't abstain from being idea of it simply like more
troublesome (Amatriain and Pujol, 2015).
Agglomerative method:
Every idea is been dispensed their own exact bundle and they disengage closeness in each
foundation and join the 2 indistinguishable organizations. This approach is additionally named as
base up gathering
The geographic regions to target:
The regions which require working upon are:
 Roma
 Hamilton
 Taree
 Douglas
 Adelaide
 Mildura
12

The regions communicated above Adelaide are the region which must be toiled upon the
clearest from the figure given above. If the alliance is focused on expanding its business
endeavour, then the ones are the geographic areas wherein the straightforward spotlight
should be on by methods for monitoring Adelaide on the unprecedented.
Output:
Figure 4: Output Prediction
Adelaide should able to focus on the book's sales and mostly “Bedtime for Frances”.
The prioritized product:
Figure 5: Product to be prioritized
The Bedtime for Frances must be prioritized by the Booking Company.
13

Free Shipping Impact On Product Sale:
Giving your clients their articles on their doorsteps that too everlastingly is something which
each support shows up for. Executing this could help the relationship in building up its customer
so it'll interchange the conveyance costs if the expense of the difference is moreover darkened
then it needs to guarantee that it doesn't decrease the man or lady of the factor to protect up the
reputation of the endeavour. Since undeniable work environments charge a couple of additional
money for transportation it'll help our director a creating amount of and could pull in additional
obvious clients (Zhao, 2015).
To improve profit of company:
To adorn the pickup of affiliation the geographic districts should be dealt with. Two or 3 systems
ought to be done inside the ones geographic locales which can be the area the foundations are
poor. Those issues which aren't being gotten the need to now not be organized away for
protracted and at whatever point the one's contraptions are requested that should be gained round
than in a way. Progressed advancing must be done in the one's regions which have fewer
customers. Those articles which are appropriately gotten less for those detached transporting
must be accessible.
14

Recommendation for the Company:
The offers will be broadened if bargains ascend in cheerful seasons. Things must be
outperformed on forever if the buyer is unsatisfied at that thing there must be a system for re-
setting up the request. The request for a method for systems for the buyer should achieve the
supporter on time. On the off risk that the site isn't that attracting, by means of then, its UI should
be pushed ahead. Distinctive suggestion consolidates:
 Scalable Storage of Data
 Better cleaning of Data
 Training Team
 Better Advertising Team
 Better Predictive Sales
 Focusing on the regions with fewer Customers
15

Implementation plan:
The areas which can be truant in the back of in diversion designs must be better through the use
of understanding numerous plans so it can build up some buyer in that region. This will help
inside the celebrated distinction in the undertaking. By then the things that have been procured
the irrelevant or three raising frameworks should be amassed to collect the offer of those units.
Further, the company should be able to implement a data storage model that could be helpful in
storing the books in stock and could help in making the better and suggestive model.
16

Python Codes with Screenshots
Figure 6: Linear Regression Graph on Sales
Figure 6 shows the Monthly Sales Linear Regression Graph for the Books of the company. This
is helpful in making the prediction over the Sales and could be helpful in making the better sale
model. As this is the first regression model the sales prediction need more testing.
17

Figure 7: Final Prediction Graph
Figure 7 shows the Final Prediction model over the Monthly Sales data and this is going to help
in making the predictions and is going to make a better model for the company using which the
company could be able to get the profit in the next quarter.
18

Code:
Figure 8: Code for test and train
19

Figure 9: Main Module
20

Conclusion:
This is the report which joins every single one of the approaches that the affiliation requires with
a particular genuine objective to build its course of action over each one of the zones in which its
working appropriate now the locales wherein the affiliation is deficient in the back of. A dataset
of a part component has been given on this report which may help the affiliation the partition to
expand the game plans inside the region wherein more work is required and the issues that
necessities all the more exceptional exchange to unite the offers of affiliation.
An exact records mining has been discovered at the dataset that transformed into given and a
brief span later bunching has been executed in a way that enables the association with
acknowledging which factor require additional idea recalling the correct objective to profit
salary. The area which wishes additional critical exchange to make the gives of alliance is
Hamilton region.
An execution setup has been given to the relationship to each one of the elements that may come
extends to accomplishing more plans for the alliance.
21

References
 Amatriain, X. and Pujol, J. (2015). Data Mining Methods for Recommender
Systems. Recommender Systems Handbook, pp.227-262.
 Andam, Z. (2017). E-Commerce and E-Business. [online] Doer.col.org. Available at:
http://doer.col.org/handle/123456789/4128 [Accessed 7 Jun. 2018].
 Turban, E., Outland, J., King, D., Lee, J., Liang, T. and Turban, D. (2017). Marketing and
Advertising in E-Commerce. Springer Texts in Business and Economics, pp.361-401.
 Wu, X., Zhu, X., Wu, G. and Ding, W. (2014). Data mining with big data. IEEE
Transactions on Knowledge and Data Engineering, 26(1), pp.97-107.
 Zhao, Y. (2015). DATA MINING TECHNIQUES Review of Probability Theory. [online]
Ccs.neu.edu. Available at:
http://www.ccs.neu.edu/home/yzsun/classes/2015Fall_CS6220/Slides/prob-review.pdf
[Accessed 7 Jun. 2018].
22

Appendix
Working Python Codes
23