Assessment 2

Table of Contents
Introduction......................................................................................................................................3

Research objective for organizational decision making..................................................................3

Previous studies...............................................................................................................................3

Data Analysis approach...................................................................................................................4

Results..............................................................................................................................................5

Descriptive statistics....................................................................................................................5

Data visualization........................................................................................................................6

Recommendation and Conclusion.................................................................................................11

References......................................................................................................................................11
Introduction
Ecommerce data analytics is the systematic method of collecting data from all areas of have
strong impact on the store performance. It is used to perform business intelligence based decision
making to improve sales by 20% for next financial year. Ecommerce analytics has wide range of
metrics right from customer visit duration, frequency of total website visits, acquisitions, sale
conversion etc. The key advantages of the ecommerce analytics are to measure the effectiveness
of the sales campaign and marketing strategies, assessment of trends and patterns of the data in
order to perform sales forecasting accurately, to optimize the product pricing, up-sell and also
stockroom performance, to evaluate customer data to personalize the sales for individual
experiences and to assess the strategies implemented based data driven decisions on store
performance. The main aim of this study is to analyze the sales trends and patterns of technical
sales website which sells smart phones, tablets and desktops principally from leading brands
such as Apple, Samsung, Blackberry, Microsoft, HTC, Huawei, Google, Lenovo, LG etc. from
different countries in the world based on the ecommerce metrics such as clicks, bounce rate,
conversion, session, traffic channel and source, duration and engagement duration etc. The data
analysis was performed using Microsoft Excel and results are expressed in the form of tables and
graphs. The benefits of data visualization includes providing clear insights about data, to
transform data into actionable strategies to develop the sales performance of the ecommerce
website. It also offers insight driven data based decision making.

Research objective for organizational decision making

The main objective of this research is to assess the current sales performance of the ecommerce
website during Feb 2018 – June 2018 in order to develop strategies to increase sales by 20%
higher for next fiscal year.

Previous studies
In a study conducted by Abdul (2016) reported that ecommerce analytics offers merits to the
national economy. The study depicted the challenges of ecommerce in the economy. However, it
also favored the growth of IT in financial sector of the nation. Author reported the increase of
smart phone sales and service would possibly lead to customer behaviour patterns during online
shopping. It also adds value and enables simplicity to shop easily via online. (Shi, 2016) reported
that Business to Customer (B2C) ecommerce will accomplish rapid development and occupy the
key position of ecommerce sector. The direct selling pattern of ecommerce describes the
manufacturer directly sell products to the customers via their own ecommerce website instead of
traditional set up like shops, malls etc. The study concluded that wide range sales planning skills
are required to boost the store performance while using ecommerce business model. The
principal demands of the customer from ecommerce website includes user friendly interface,
search information about the products very quickly, personalized services, safe consumption,
secure payment, promotional commodities and its information in the website, comfortable
shopping experience etc. The ecommerce companies should develop highly interactive services
to permit the customers to shop in virtual reality.

Data Analysis approach

The study employed quantitative research methodology. Quantitative data analysis aim at
assessing the data based on numerical calculations using average, median and mode as well as
analyze the causal relationship between dependent and independent variables using methods
such as correlation and regression. Descriptive statistics usually describe the dataset using the
sample data. The data was gathered from the web sources on the sales revenue earned during 1,
Feb 2018 – July 1, 2018. The dataset has 22 variables such as date, first page topic, last page
topic, type of the device sold, browser name, manufacturer of the device, country, region, traffic
source, traffic channel, traffic description, product name, credit offered, merchant name, session,
total duration, total engagement duration, click, conversion and revenue. The variable country,
region are demographic variables. The first page topic, last page topic infers the website topics,
the variable type of device, browser name, product name, manufacturer name, credit offered,
traffic source, traffic channel, traffic description and merchant name are categorical variables
with three to six levels. This study conducted the descriptive analysis and data visualization
using MS Excel 2017. Data visualization is the characterization of the data using graph, chart or
other visual formats. It offers the visual context of the data. Data visualization aids in the
assessment of sales and marketing campaigning to increase the store performance. There are
different types of data visualization such as comparative analysis charts which elucidates the
similarities, differences and associations between variables in the dataset. Sentiment analysis
charts compares the customer preferences over the products or product types in the store. The
differences can be sales, customer satisfaction and credit type needed whether good, excellent or
no. The sentiment map aids in the customer feedback on the user’s shopping experience. Sankey
diagram offers overview of the workflow in the business process or any system. In this case,
Sankey diagram was employed to elucidate the customer journey to the website using the KPI
metrics related to search engine, campaigning such as traffic source, channel, description, total
duration, engagement duration, clicks and conversion. This type of visualization aids in the
improvement of the actionable strategies to optimize the search results to draw more users to the
website. It targets the improvement of sales and marketing campaigning. Sankey charts visually
describe how customers are moving through the sales funnel of the ecommerce website and their
impact in coordinating the ideal way to function the sales funnel. Visualization has strong impact
in data description since there is a strong relationship between the human brain and visual story
telling. Human brain can process the visual stories more effectively than the texts and statistical
outputs. It also enables to offer key insights and identify the trends and patterns within the
variables. The data visualization diagrams are designed using metrics and KPIs to reinforce the
accuracy and relevancy.

Results
Descriptive statistics
Statistics Sessions bounced_sessi pages_view total_durati engagement_dura clicks conversio Revenu
ons ed on tion ns e
Mean 10.10563 0.133802817 36.5915493 174516672 2136843.848 16.035 10.26056 3529.53
21 34 5
Standard Error 1.539306 0.046049481 5.87719490 37588772.2 410708.0881 2.4763 1.565111 546.030
3 9 88 31 7
Median 3 0 11.5 13684095.5 546460 5 3 1070
Mode 1 0 2 #N/A #N/A 1 1 485
Standard Deviation 18.34295 0.548742899 70.0348601 441568092. 4824727.602 29.509 18.65045 6506.70
1 9 56 38 7
Sample Variance 336.4639 0.301118769 4904.88163 1.94982E+1 2.3278E+13 870.81 347.8394 423372
7 44 27 36
Kurtosis 13.51395 46.0422243 15.6633573 15.9117114 34.27566423 12.376 13.40603 14.5688
5 2 38 14 6
Skewness 3.464968 6.078166246 3.69696460 3.88172739 5.173597857 3.3513 3.458775 3.58654
6 6 68 42 5
Range 112 5 456 252389137 41346256 179 113 40781
7
Minimum 1 0 1 4552 599 1 1 75
Maximum 113 5 457 252389592 41346855 180 114 40856
9
Sum 1435 19 5196 240833007 294884451 2277 1457 501194
33
Count 142 142 142 138 138 142 142 142
Confidence Level 3.043103 0.091036669 11.6188116 74329211.5 812146.9925 4.8956 3.094117 1079.46
(95.0%) 7 5 49 84 5
The mean sessions among nations from different users is 10 minutes during Feb – June 2018
which ranges from 1 to 113 minutes. The mean bounced sessions is 0.13 minutes. The mean
pages viewed are 37 during Feb – June 2018. The mean total duration spent by the users during
Feb – June 2018 is 174516672 minutes. The mean engagement duration is 2136843.848 and the
mean clicks made by the users is 16. The mean conversion is 10.26. The mean revenue earned
during Feb – June 2018 is 3529.53 USD.

Data visualization

The above graph clearly depicts that highest sales was earned via chrome browser from PC and
Safari browser from iPhone. Customers using Google chrome from mobile, Firefox and internet
explorer earned revenue.
The daily revenue earned during Feb – June 2018 observed to increase. However, the sales were
poor during end of May – June 2018.

Total Clicks vs Browser

1200

1000

800

Total
600

400

200

0
Chrome Chrome Edge Firefox Internet Opera Safari Samsung
Mobile Explorer Browser

The total clicks made by the users at different browsers during online shopping were described at
the above graph. The highest clicks were made using Google Chrome PC and Safari browser.
Most of the customers were using desktop for accessing the ecommerce website. Nearly 100000
customers were using mobile for online shopping and less than 5000 users were using tablet for
online shopping.

The total average duration spent by the user at different browser indicates that most of the
customer used Google Chrome (PC), Edge and Safari. Chrome mobile earned lowest average
duration spent by the customer. Opera and Samsung browser are rarely used by the browser.
 3000

2500

2000

1500
Sum of sessions
1000 Sum of bounced_sessions
Sum of pages_viewed
500

0
e il e ge x
re
r a ri r
m ob Ed
fo lo er fa se
ro re Op Sa ow
Ch M Fi Ex
p
Br
e t
ro
m ne ng
Ch ter su
In m
Sa

The total sessions and pages viewed were higher for Google chrome than Chrome mobile, Edge,
Firefox, Internet explorer, Samsung browser, Opera. The sum of bounced sessions were higher
for safari. This infers that website should be more optimized for apple users.

Revenue by Traffic Source

400000

350000

300000

250000
Total
200000

150000

100000

50000

0
Direct Internal Organic Paid

The revenue earned by organic traffic sources are higher than direct, paid and internal sources.
 Revenue vs Traffic Channel
350000
300000
250000
200000
150000 Total
100000
50000
0
ct l ok l l er hoo
Bi
ng re G o m ai og
le
ra
m na ra
m ra
Di uck E e bo o g t er g f er w
itt
Ya
ac G a In o Re T
c kD F nst Pr
Du
I er
a rtn
P

The search contributed highest revenue than that of direct, referral, internal and social web traffic
channels employed to direct the customers to the website. The email and native earned very low
revenue.

Revenue vs Traffic Description

350000
300000
250000
200000
150000 Total
100000
50000
0
ct l ok al l er hoo
Bi
ng re G o m ai og
le
ra
m
rn ra
m ra
Di uck E e bo o g te g f er w
itt
Ya
c G a In o Re T
ck
D
Fa st Pr
u In e r
D
rtn
Pa

The revenue earned via traffic routing by Google is highest than partner affiliate program,
referral program, Facebook, Direct and Internal traffic description types.
Recommendation and Conclusion
The study concludes the store performance is high during Feb – April 2018 and experienced
sudden drop during middle of May till end of June 2018. Google chrome and Google was mostly
used by the users and hence, marketing should focus on these browsers. The user purchase
history should be evaluated to offer personalized shopping experience. This study offers
following recommendation for increase in sales by 20% for next fiscal year. They are

 The social media and email marketing should be conducted to increase the traffic from
the social media and email sources
 The personalized shopping experiences should be offered by the customers based on the
customer choices data gathered from the user’s cookies
 The sales appear to drop during May and June. The promotions and discount offers as
well as combo offers should be provided as in the name of summer sale to boost the sales
performance during this period.
 The trending products should be used for the promotion to increase the customer visit
duration in the website
 Strategies should be developed to improve conversion rates, increasing the average
purchase value of customer and bounce rates should be reduced by referral ad marketing.

References
Abdul G. (2016). Electronic Commerce: A Study on Benefits and Challenges in an Emerging
Economy, Global Journal of Management and Business Research: Economics and Commerce
16:1.

Shi, Y. (2016). Analysis on the B2C Ecommerce Pattern in China, Journal of Service Science
and Management, 9, 443 – 452.