Descriptive Analytics and Visualisation

Running head: MIS771DESCRIPTIVE ANALYTICS AND VISUALISATION
Descriptive Analytics and Visualisation
Name of the student
Name of the university
Author’s note

1
MIS771DESCRIPTIVE ANALYTICS AND VISUALISATION
Table of Contents
Introduction......................................................................................................................................2
Main Body.......................................................................................................................................2
Task 1 Descriptive Statistics........................................................................................................2
Task 2.1 - Identify the significant variable for the Model...........................................................3
Task 2.2 – Model Building..........................................................................................................4
2.2a - First Model........................................................................................................................4
2.2b - Second model....................................................................................................................5
Task 2.3 – Interaction Effect.......................................................................................................6
Task 3...........................................................................................................................................7
Task 3.1 – Predictive Model 1.....................................................................................................7
Task 3.2 – Predictive Model 2.....................................................................................................8
Task 3.3 – Predictive Model 3...................................................................................................10
Task 4.........................................................................................................................................11
Time series for predicting turnover...........................................................................................11
Conclusion.....................................................................................................................................11
Appendix........................................................................................................................................14

2
MIS771DESCRIPTIVE ANALYTICS AND VISUALISATION
Introduction
Mad Dog Craft Beer (MDCB), is an ale brewing micro-brewery, located in Australia. It
primarily supplies Pale Ale beer to the regions of Melbourne and Victoria. It has seen a
substantial growth in business for the last fifteen years. As a matter of fact, the people of the
region are liking the beer being produced by the organization. In fact, to meet the growing
demand of beer the organization was forced to increase the production to more than 3 ML/year
last year (2018).
The organization, is expecting a phenomenal growth in beer sales due to it being liked by
Victorians. More importantly there has been an increased interest in micro-brewing culture.
Thus, MDCB feels that it should investigate the relation between its product and customers. The
organization understands, that quality of their beer has been the force behind its growth. Herein,
we investigate the relation between the quality of MDCB’s beer and sales. The quality of beer
has been measured using nine different parameters. We explore which of the factors are highly
responsible for the success of MDCB. Further, we also examine the reasons behind customers
recommending Pale Ale Beer manufactured by MDCB. To investigate the success of beer,
information has been collected from 200 customers. The information is segregated into three
broad sections. Part of the information regarding loyalty and distribution channel is collected
from Warehouse of MDCB. The information regarding the perception of the customers is
collected from survey. The last group of information relates to the recommendation of the beer
and order quantity. The information is reviewed so that MDCB can increase its customer base.
Main Body

3
MIS771DESCRIPTIVE ANALYTICS AND VISUALISATION
Task 1 Descriptive Statistics
From the survey it is found that the average sales of Beer to each customer is 7670 bottles
with a variation of 890 bottles. The minimum and maximum number of bottles purchased by a
customer is 4300 and 9900 bottles respectively. The total number of bottles purchased by the 200
customers who were surveyed was 15330. Further it is found that 50% of the customers less than
7600 bottles (Table 1). Further, it is also visualised that the variable order quantity is normally
distributed (Figure 1). Hence, the variable order quantity can be used as the outcome variable.
The survey revealed that 51% of the customers surveyed would recommend beer
produced by MDCB (figure 2). Thus it is found that more customers would recommend
MDCB’s Pale Ale Beer. However, the difference in percentage of customers who would
recommend to would not recommend the beer is only 2%.
Task 2.1 - Identify the significant variable for the Model
In order to identify variables which might significantly affect order quantity of beer, 9
variables were identified. The distribution of dependent variable (order quantity) was tested. It
was found that order quantity was normally distributed. Thus, the relation between order quantity
and the nine variables was analysed. Karl-pearson correlation was used to reveal the relation
between order quantity and the variables. The analysis revealed that most of the customers
believed that competitive pricing and flexible pricing does not impact (has a negative impact) the
order quantity. Thus the two variables were dropped from further studies (Table 2).
From the seven variables which have a positive correlation it is found that Shipping cost
has the highest correlation (p = 0.5044) with order quantity. The correlation between Quality and

4
MIS771DESCRIPTIVE ANALYTICS AND VISUALISATION
Shipping speed with Order Quantity are 0.4334 and 0.4251 respectively. Hence, Quality,
Shipping speed and shipping cost are moderately correlated with order quantity.
The correlation between Brand image and Order fulfilment and Order quantity are found
to be 0.3380 and 0.3146 respectively. Moreover, SM presence and Advertisement have a
correlation of 0.2352 and 0.2370 respectively with order quantity. Thus from correlation
analysis it is found that brand image, order fulfilment, sm presence and advertisement have a low
correlation with order quantity.
Task 2.2 – Model Building
2.2a - First Model
The variables selected to predict the order quantity of beer are Product Quality, Social
Media Presence, Advertising, Brand Image, Order & Billing, Shipping Speed and Shipping Cost.
These seven variables have been selected from the initial process wherein they have been found
to be positively correlated with order quantity.
As a first step towards building a model to predict order quantity linear multiple
regression is used.
The order quantity can be predicted as:

Order Quanity=3.0334+0.2774∗Quality−0.1560∗SM Presence−0.0180∗Advert+0. .3219∗Brand Image−
Analysis of the regression coefficients reveal that keeping other variables constant for
unit increase in brand image, order quantity increases by a factor of 0.3219. Similarly, for
one-unit increase in shipping cost, order quantity increases by a factor 0.2568. In addition,

5
MIS771DESCRIPTIVE ANALYTICS AND VISUALISATION
for unit increase in quality of beer, order quantity increases by 0.2774 (Table 3). Further,
from the regression coefficients table it is found that the p-value for Quality, Brand Image
and Shipping Cost are 0.0000, 0.0000 and 0.0008 respectively. Thus at 5% level of
significance the coefficients of the variables are found to be significant.
Further, it is found that for unit increase in SM presence there is a decrease in
order quantity by 0.1560. Moreover, at 5% level of significance the regression coefficient is
not significant (p-value = 0.1195). For unit increase in Advertisement there is a decrease in
order quantity by 0.0180. Moreover, at 5% level of significance the regression coefficient is
not significant (p-value = 0.7384). For unit increase in order fulfilment there is a decrease in
order quantity by 0.1493. In addition, at 5% level of significance the regression coefficient is
not significant (p-value = 0.0713). Thus, it is found that the independent variables SM
presence, Advertisement and shipping cost reduce order quantity.
For unit increase in shipping speed there is an increase in order quantity by 0.1738.
However, at 5% level of significance the regression coefficient is not significant (p-value =
0.2026).
Thus the four variables (SM presence, advertisement, order fulfilment and shipping
speed) are dropped in further studies of order quantity.
The analysis revealed that 47.29% of order quantity can be predicted from the seven
variables. Moreover, it was also found that the independent variables significantly affected order
quantity, at 5% level of significance, F-value (7, 192) = 24.6039 (Table 3).