Sales revenue of Amazon.com is showing
an increasing trend and it also has a seasonal component i.e. Q4 of every year
is exceptionally high as compared to rest of the three quarters. Triple Exponential
Smoothening also known as the Holt-Winters method is used for forecasting the
sales of Amazon from Q4 2017 till Q3 2018 as it has a trend and seasonal impact.
Quarterly data from Q1 2007 till Q3 2017
is used for calculating the forecast. Sales data from Q1 2007 till Q4 2018 i.e.
8 quarters is taken as the base for forecast. Average sales of this period are
taken as the starting point for calculating the starting level and seasonal
factor for the same period. (Hyndman and Athanasopoulos, 2013)
To keep a check on the forecast error
various indicators are calculated namely BIAS, MAD, MSE & MAPE. ?, ? &
? are then solved using the solver tool in excel to find out the optimum level
where MSE is the lowest. ?, ? & ? is kept in the range of 0.05 to 0.95. The
optimum solution provided by the solver tool and two other scenarios are shown
It can be observed from the above table
that solver’s solution is the optimum one with the lowest MSE. The optimum
solution has a low alpha factor which means that forecast have lower impact of
historical data and higher weightage of the average of the base quarters and
trend. High beta & gamma factors mean that trend and seasonality have
higher impacts on the forecast. Graph showing actual and forecasted values are
Amazon.com Investors Relations, 2017
Multiple Regression Analysis
Multiple regression analysis is used to
find if there is a statistically significant relationship between several predictor
variables and dependent variable and the strength of the relationship. It helps
in analysing and finding trends if any in different sets of data. (Statsoft,
Amazon is an e-commerce company and is
accessed through internet by its customers. Year on year worldwide internet
user data from 1996 to 2016 is tested as predictor variable against net revenue
of Amazon.com i.e. dependent variable. The data is plotted on a scatter graph
as shown below.
Source: Internetworldstats, 2017 &
Amazon.com Investors Relations, 2017
A regression analysis was run on the
data. The coefficient of correlation was 98.84% which indicates a very high
correlation between the variables. The coefficient of determination came out to
be 97.69% which indicates that ~ 98% variance can be explained through the
internet user statistics during the analysis period. The statistical p-test
shows that the probability of null hypothesis being true is less than 5% hence
null hypothesis was rejected. Linear equation generated through the multiple
regression analysis is as follows:
y = 50.629 * x – 62.565
The internet users line fit plot,
trendline and equation are shown in the below graph.
Regression analysis can be used by the
company to predict the growth or change in the dependent variable based on the
predictor variable. As in this case the company can predict its revenue based
on the increase in the number of internet users. This kind of analysis will
also help the company in taking strategic decisions like how they can
contribute or take steps to increase the number of internet users.
Correlation analysis is used to test the
relationship between two or more variables. It is important to understand the
correlation between different variables so that accurate predictions could be
made about the future. A positive correlation means that the variables move in
the same direction and negative correlation means that the variables move in
the opposite direction. It differs from regression in the terms that
correlation quantifies the degree to which two variables are related but it
does not fit a line through which the value of another variable can be
calculated whereas regression provides a line of best fit through which the
dependent value can be calculated. (Statisticshowto, 2017)
A correlation analysis has been run on
number of active users and net revenue. The correlation between the two comes
out to be 99.8% which means that they are highly positively correlated i.e. if
the company works towards increasing the number of active users the revenue
will increase in the same proportion. Below graph shows the net revenue and
active users’ data.
Basis the above analysis it is
recommended that data warehousing solution must be implemented in the company.
This will help the company in accurate forecasting and planning, assist in
taking strategic decisions and help in efficient utilisation of resources.
Implementation process, major risks and steps to overcome these issues while
implementing the data warehousing solution are mentioned in the report which
will help in effectively planning the implementation and will lead to a smooth
D. (2017). Alibaba vs Amazon: The race to $500 billion. online CNBC.
Available at: https://www.cnbc.com/2017/09/01/alibaba-vs-amazon-the-race-to-500-billion.html.
Editorial (2017). Amazon.com Inc (AMZN.O) Company Profile | Reuters.com. online
U.S. Available at: https://www.reuters.com/finance/stocks/company-profile/AMZN.O.
(2017). Amazon.com: Online Shopping for Electronics, Apparel, Computers, Books,
DVDs & more. online Amazon.com. Available at: http://www.amazon.com.
(2017). Amazon.com Annual Report 2016. online Seattle: Amazon.com. Available
Investor Relations (2017). Amazon – Investor Relations – Press Release.
online Phx.corporate-ir.net. Available at: http://phx.corporate-ir.net/phoenix.zhtml?c=97664&p=irol-newsArticle&ID=2311817.
M. (2017). Amazon Distribution Network Strategy | MWPVL International. online
Mwpvl.com. Available at: http://www.mwpvl.com/html/amazon_com.html.
Dunn, J. (2017). Amazon’s logistics
footprint in the US is only getting bigger. online Business Insider.
R. and Ray, J. (2017). Amazon’s epic 20-year run as a public company, explained
in five charts. online Recode. Available at: https://www.recode.net/2017/5/15/15610786/amazon-jeff-bezos-public-company-profit-revenue-explained-five-charts.
P. (2017). Amazon Will Make Up 50% of All U.S. E-Commerce by 2021. online
Fortune. Available at: http://fortune.com/2017/04/10/amazon-retail/.