Case Analysis of Happy Cow: by - Gazal Doshi 20030241065
Case Analysis of Happy Cow: by - Gazal Doshi 20030241065
Case Analysis of Happy Cow: by - Gazal Doshi 20030241065
Cow
By –
Gazal Doshi 20030241065
From next type of visualization, we are getting a similar result. We notice that the sale of
S.caramel for staff is the highest among all other flavours.
Conclusion: The dataset suggests that the sale of S. caramel flavoured ice-cream is better than
other flavours throughout the season. So, there is need to focus less on this particular flavour and
focus should be on increasing sales of other ice cream flavours as well. Also, the sale is almost
constant throughout the given dates, that also shows that irrespective of season and months, S.
caramel sale is best among others and customers like it more than other flavours.
A similar pattern is observed once again while using other analysis technique and plotting data in
a different way.
It is depicted in the diagram below:
Some insights gathered by using Forecasting techniques:
We have used some forecasting techniques using the Forecast function of Microsoft excel along
with 3- months moving average and 5-months moving average. The insights after the analysis of
data using these techniques are:
For Staff: There is a considerable increase in sales in the month of May (compared to April),
which is obvious due to the summer season. But after that there is a decreasing pattern that
occurs in the analysis. This shows, that as the season changes from summer to rainy, there is a
decrease in sales. Further, the sales decrease more in winter season. Similar pattern is also found
when we have used some other forecasting techniques like, 3-months moving average and 5-
months moving average.
For Students: Though the sales started high from April itself (probably due to start of new sessions,
students interacting over small parties), but it decreases in further months. However, there is again
an increase in sales in the month of September (as compared to previous months). This maybe due
to start of a new semester or because of student’s likeness to eat ice cream in such seasons. So, in
such season more focus should be given on students. Further, the sales remain almost constant and
there are more sales when compared to staffs. Similar pattern was also found with other forecasting
techniques like 3-months moving average and 5-months moving average.
For Tourists: Tourism is more in Winter season as compared to summer season. This is because
people love winter season and it is their favourite season to explore new places. A similar pattern
is also supported by the graphs generated by forecasting techniques using given data. It is noticed
that the sales are low in the month of April. It is very low in the month of June (i.e., people don’t
like to travel in the month of June at all, probably due to the high temperature). However, as rainy
and winter season approaches, there is a sudden rise in sales of ice creams. This shows that people
get out from their homes in these seasons and love to travel in these months. Similar pattern was
also found with other forecasting techniques like 3-months moving average and 5-months moving
average.
Ques-02: Explore the sales performance of three consumer groups (students, staff
and tourists)
Looking at the Happy cow ice cream data we could infer that, the favorite flavor among the group
is S. Caramel and Mango throughout the season.
Also, we got to know that the Tub Ice Cream and YY Sesame are the least favorable flavor
among all.
25000
20000
15000
10000
5000
26%
61%
3 Topping 20 56 333
YY Sesame 920 0 0
150000
100000
72079
50000 37388
0
Total Revenue
When looked into the overall Sales performance, we could say that students are the highest
contributor to the Happy Cow’s Revenue. About 61% revenue part is from Student and 26%, 13%
is from Staff and Truism respectively.
We can see that tourist spent less for ice cream or in this season there were very less tourist visited
the store. For attracting more Tourist there has to be some specific offers pertaining to the tourism
group.
DISCOUNT LANDSCAPE
Tourism Staff Student
TRIPLE SCOOP
S INGLE SC OOP
FINGER B 100
DOUBLE SCOOP
Double Scoop Finger B Gift coupon 3 cone Single Scoop Triple Scoop
100
Tourism -1440 -720 -261 0 -280
Staff -3930 0 0 0 -342
Student -9802 0 -390 -208 -1118
The discount strategy could actually work in flavor of Happy Cow Ice cream, as we know that the
student group has spent highest on ice cream even though we found that the highest redemption of
discount Is taken by them (Students).
The discount strategy for the student group could be maintained as same because it is giving what it
can for revenue but for Staff and Tourism there should be some alternative for more redemptions.
There is one interesting thing observed that the lowest discount redemption is from Tourism group
but Tourist have taken discount in all the categories unlike other groups.
The highest discount redemption took place from ‘Double Scoop’ category from all three groups.
Also we see that the discount Category ‘Single Scoop’ and ‘Finger B100’ did not go well for all
group only one group has taken the advantage of this category.
Discount Insights
1. Overall, Lowest Discount was availed by Tourism.
2. Among all three groups, the highest category which was redeemed was
Double Scoop.
Happy Cow could actually add some more benefit for staff members and their families in discount
categories so that the overall revenue could go higher from this category too.
Ques-03: Both Marry and Prem, senior sales assistant belief that different groups of
flavors sell better at different times of the year. Do the data backs it up? please
propose your groupings and visualize them to generate insight into the ice cream
sales. Regarding flavor groups, does grouping gives a better level of analyses than
individual level.
Both Marry and Prem, senior sales assistant belief that different groups of flavors sell better at
different times of the year. Yes, grouping of various flavors can give better results at different time
of year. For this, we have taken weekly data of Student, Staff and Tourism.
Let’s understand each of the weekly data and visualize them to generate insight into the ice cream
sales.
Student Data
Here, we have generated graph for sales of ice cream. Below are the insights obtained:
Grouping:
1st Group: From 1st April 2017 to 28th April 2017, flavors preferred more are: Waffle Cone,
Strawberry, S. Caramel, Pure Coco, Mint Choco, Mango
2nd Group: From 29th April 2017 to 19th May 2017, YY Sesame, Waffle Cone, Vanilla Bean,
Strawberry,
S. Caramel is sold more.
3rd Group: 1st September 2017 to 29th September 2017, YY Sesame, Waffle Cone, S. Caramel, Mint
Choco, Mango, Hazelnut is preferred. Ice Cream sand, Mango, Hazelnut is sold more than other
time of year.
Sales of Student Weekly data YY Seasame St
14000
Waffle Cone St
Vanilla Bean St
12000
Up 3 Toppings ST
Triple Scoop ST
8000 Strawberry St
Single Scoop ST
6000
S. Caramel St
Red Bean St
Pure Coco . St
Pistachio St
2000
Pina Colada St
Mint Choco St
Mango St
Out of all 28 flavors, not all the flavors have great demand. To understand the preference, we have prepared
Pareto Chart, that gives us the weightage of ice cream flavor of Student weekly data.
As per graph below, the highest weightage is given to S. Caramel. This flavor is
demandedthroughout the season. Next 5 flavors are Pure Coco, Mango, Mint Choco, and
Strawberry.
Apart from the groups detected, Management should focus that these ice cream should be
available all the time in inventory.
To understand if grouping gives a better level of analyses than individual level, we have created
Staff Data
Here, we have generated graph of staff weekly for their product sales of ice cream. Below are the
insights obtained:
Grouping:
1st Group: From 2nd June 2017 to 23rd June 2017, flavors preferred more are: Double Scoop, Red
bean, Apricot, Coffee, Green Tea, Hezelnut, Waffle Cone, Strawberry, Ice Cream Sand
2nd Group: From 23rd June to 7th July Chai Tea, Chocolate, Mint Choco, Green Tea, Vanila
Bean, Banana is sold more.
3rd Group: 1st September 2017 to 29th September 2017 Double Scoop, Vanila Bean, Pina Colada,
YY Sesame, Strawberry, Mint Choco, Pure Coconut is preferred.
Staff
Out of all 28 flavors, not all the flavors have great demand. To understand the preference, we have
prepared Pareto Chart, which gives us the weightage of ice cream flavor of Staff weekly data.
As per graph below, the highest weightage is given to Mango Ice Cream. This flavor is demanded
throughout the season. Next 5 flavors are Chocolate, Pure Coconut, Mint Choco, Strawberry, Ice
Cream Sand.
Apart from the groups detected, Management should focus that these ice cream should be available
all the time in inventory
1200 2017
Vanilla Bean
Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri
Waffle Cone
7th 14th 21st 28th 5th 12th 19th 26th 2nd 9th 16th 23rd 30th 7th 14th 21st 28th 4th 11th 18th 25th 1st 8th 15th 22nd 29th Apr
Apr Apr Apr May May May May Jun Jun Jun Jun Jun Jul Jul Jul Jul Aug Aug Aug Aug Sep Sep Sep Sep Sep 2017 2017
2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 S. Caramel
As per graph below, the highest weightage is given to Mint Choco Ice Cream. This flavour is
demanded throughout the season. Next 5 flavours are Chocolate, S. Caramel, Pure Coco,
Strawberry, Hazelnut, Mango.
Apart from the groups detected, Management should focus that these ice cream should be available
all the time in inventory.
Chart Title
Fri 7th Apr 2017
Fri 14th Apr 2017
Fri 21st Apr 2017
Fri 28th Apr 2017
Fri 5th May 2017
Fri 12th May 2017
Fri 19th May 2017
1200
Fri 26th May 2017
1000 Fri 2nd Jun 2017
Fri 9th Jun 2017
800 Fri 16th Jun 2017
Fri 23rd Jun 2017
600
Fri 30th Jun 2017
400 Fri 7th Jul 2017
Axis Title
Red Bean
3 Topping
S. Caramel
Single Scoop
Tub Ice Cream
Waffle Cone
Chocolate
Pure Coco .
Mint Choco
Cherry Almond
Finger B 100
Double Scoop
Strawberry
Triple Scoop
Chai Tea
Coffee
Green Tea
Banana Caramel
Lime Coconut
1 Topping
YY Seasame
Pina Colada
Granola
Mango
Chocoate Cookies
Ice Cream SAND
Apricot
Ginger
Ques-04: What outliers can be identified from the daily sales of happy cow? Please
define outliers and how to address them.
An outlier is a data point that is significantly different from the remaining data. Outliers can be
very informative about the subject-area and data collection process. In case of Happy Cow sales,
we are able to visualize some outliers with respect to various ice cream flavors and other beverages
which are being sold by Happy Cow university outlet.
As per the above visualization majority of the outliers are because of natural progression of the
data hence there is no much deviation from the upper bound value of that particular product sales
quantity. Except for beverage item green tea. For sales quantity of green tea there is extreme
deviation from the range of sold quantity particularly in the last week of the April as that week is
last week of the lectures and summer season hence to avoid any health issues students must have
preferred green tea. This outlier suggests that green tea should be kept as accompany to ice cream so
as to increase the revenue.
When it comes to staff, vanilla bean flavor of the ice cream has comparatively large deviation from
the range of the quantity sold. Vanilla being the favorite ice cream flavor of the West, no wonder
is famous ice cream flavor in the world and it goes with everything be it syrups, chocolates or
candies! It is also very old flavor and hence considering the varied age group of the staff it is
obvious that majority of the people are choosing vanilla. Also, this quantity has risen in June month
where there are no regular lectures and hence staff members have ample amount of time during
this month.
Vanilla also adds taste if eaten with fudge flavor hence this outlier may be indicating that no matter
what Happy cow inventory must have vanilla flavor for all 365 days.
For tourists visiting Happy Cow university outlet chocolate flavor sales quantity is very evident
outlier. Considering the large number of tourists who visits the university outlet chocolate being
the favorite flavor of all age groups. The sales of chocolate flavor is good throughout the year
somehow in 2017 it had commendable increase than that of its range.
Other than the mentioned flavors all other outliers are natural progression of the data and can be
ignored as they are not having any impact on the data analysis.
We can address outliers in various ways depending upon how are they generated:
If outliers are due to errors in data entry then thy must be corrected or removed from the
dataset.
Sampling problems such as sample size is too small, random sample or sometimes you may
collect data from the population which is not your target population. In these cases, it is
suggested to remove the outlier from the data.
Natural variation: All data distributions have a spread of values and it is natural to have
some extreme values in such dataset. You need to remove such outliers as they are
depicting some potential surprises and unusual observations which can be used to analyse
data and infer something from it. This was the case with Happy cow sales data. Hence we
haven’t removed the outliers as they produced because of the natural variation and they can
be used for deriving sales strategy.
Ques-05: Give characteristics of Happy cow dataset? what re the purposes of time
series analysis (i.e predictive vs descriptive)?
Descriptive statistics are brief descriptive coefficients that summarize a given data set, which can
be either a representation of the entire or a sample of a population. Descriptive statistics are broken
down into measures of central tendency and measures of variability (spread). Measures of central
tendency include the mean, median, and mode, while measures of variability include standard
deviation, variance, minimum and maximum variables, kurtosis, and skewness.
Median 0
Mode 0 Measure of central
tendency
Average 180.329
Standard
deviation 127.5413
Variance 16266.78
Kurtosis 23.49407
Minimum -31
Maximum 2150
Since the kurtosis here is greater than 3 that means that the data set has heavier tail than a normal
distribution. Here positive vale of skewness tells that the data are skewed to right side. Since we
have obtained a high standard deviation here this means that data is more spread out.
Time series analysis methods can also be split into linear and non-linear and univariate and
multivariate.
Trend analysis
• Student
Trend Analysis for Student
14000
12000
10000
8000
6000
4000
2000
0
Fri 28th Apr…
Fri 7th Apr…
Fri 14th Apr…
sum total
Linear (sum total)
Because of the restrictions by the university on the stores operations we can see the
fluctuations in the revenue for different months. As there is price difference for different
customers, it is possible that price of ice creams will be low for students and hence the
revenue generation is less in this case. The period between May, June and July are the
examination time and hence this could be the reason that less students come to the store
to grab ice cream and hence less revenue is made and this we can see in the trend line.
• Staff
10000 Staff trend analysis
8000
6000
4000
Revenue
2000
Linear ( Revenue)
0
Fri 28th…
Fri 7th…
Fri 23rd…
Fri 14th…
Fri 4th…
Fri 11th…
Fri 18th…
Fri 25th…
Fri 8th…
Fri 15th…
Fri 7th Jul…
Fri 22nd…
Fri 2nd…
Fri 21st…
Fri 1st…
Fri 28th…
Fri 29th…
Fri 14th…
Fri 12th…
Fri 19th…
Fri 26th…
Fri 21st…
Fri 5th…
Fri 9th…
Fri 16th…
Fri 30th…
June and July are the month when exams are generally conducted and teachers and staff gets most of
the free time. This is the reason when most of the staff comes in the campus and visit stores. So this
can be the reason why revenue from staff is generated more in the month of June and July andthen it
decreases. Since the cost of ice cream for staff may also has been kept low this is the reason the trend
line is continued to be decreasing.
Tourist
Tourist trend analysis
6000
5000
4000
3000
2000
1000
Fri 7th Apr…
Fri 14th Apr…
June, July and August are considered to be the summer months in Hong Kong and that is the time
period when tourists from nearby countries like China arrives Hong Kong. The temperature in these
months can go above 35 degree Celsius and hence comes the need for some cold dessert. Since
Happy cow has various kind of flavors that are not being sold by another ice cream company. So
this makes the benefit for the revenue for happy cow in the month of July and august as most of
the tourists come to visit and have happy cow ice cream and hence the revenue for sales is
increasing and we can see the trend line increasing linearly. It is also known that October and
November are season when European tourist mostly visit Hong Kong and since this season also
consist of same temperature as July and august the demand for ice cream will increase and that we
can in the graph that trend for revenue is an increasing trend line.