Data Analytics For Sales Data Of Computer World: Essay.

You are working as a Data Scientist in an ecommerce company.  The company is a market leader with range of product segments.  The main product segments are latest gadgets, books, toys, household items and clothes. 

However, the company’s performance especially its profitability is decreasing due to increased competition and changing customer expectations. The company has well qualified board of directors who believe in the power of data analytics in solving business problems and thus make important business decisions.  Due to your data analytics skills and data science capability, the company executives have given you a special task in analysing the company’s sales data for each product segment and each geographic region.  The executives have requested you the following: 

1. Data which has these data set ….Develop the dataset for at least one product segment with at least 1000 records and with attributes such as

• product name,
• product price,
• shipping type (free or customer paid)
• monthly sales ($)
• geographic region
• Of customers who bought the product,
• Customer type (New or existing)
• Tables , rows column of product name etc

You are also allowed to add any more attributes to describe product segment

Note:You must develop your own unique and original dataset - copying a dataset from another student or from the internet will result in reduced or zero marks.

2. Clustering techniques …Research any specific data mining or classification technique and propose a suitable technique or model to determine any association or relationships among the attributes.
3. Recommendations .,….Develop a predictive model to predict monthly sales for a given geographic region. You can use any of the methods such as Naive Bayes, decision trees or linear regression. You are also welcome to do a comparative analysis of all the methods you come across in your research and use the comparative analysis to justify your approach and research findings.
4. Impact…Based on your analysis, present recommendations to the board for the following business problems:

• What is the most likely geographic region to target new customers to increase sales and profit?
• Which products should be prioritised for increase in sales?
• What will be impact on product sales if free shipping is provided to all products?
• Any new innovative ideas to improve company’s profitability supported by your data analytics. 

Proposed Report Structure 

The report should be structured using the following headings (2000 words)

1. Executive Summary
2. Table of Contents
3. List of Abbreviations and assumptions made
4. Introduction – What is the problem?
5. Research Methodology used clustering technique , why you use it and why u are not using another techniques
6. Analytical Findings
7. Recommendations to the company
8. An implementation plan based on the recommendations you have provided
9. Conclusions
10. List of References
11. Appendix (E.g. Python code)

Note:

• You must submit your sample data as a CSV file.
• If you use Python code to analyse their data and draw graphs, then you must include all that code in the Appendix.
• Write csv file
• Coding
• upload and read that file. 

Assessment 1 ICT706 weighting 30% 

Assessment Criteria:

To write a report outlining the changes that have occurred in a business due to digital innovation using a work cantered analysis approach.

Criteria	Comments	Possible Mark	Mark awarded
Executive Summary: (1 page)		1
1.      Background.  You must be brief and to the point (about 1 page). Use your own words – you mustn’t just reproduce the case here.		2
2.      Introduction – Articulate the problem space.		2
3.       Assessment of the various data analytical methods Based on your analysis of various analytics approaches, justify your approach and especially discuss how your findings can bring innovative solutions to the business problems.  Your assessment must be supported by concepts and other material from the literature (text, readings, evidence etc.).		4
4.       Good use of Python to develop a predictive model of monthly sales		4
5.       Recommendations.  Provide recommendations on how your approach can answer the business problems.  You must convince company executives to adopt your recommendations (i.e. ‘sell’ them to management). You should support your recommendations by plotting graphs and using effective diagrams.		5
6.       Recommendation Implementation Plan.  Your plan for implementing the recommendation (this could include staff training, advertisement campaign design etc.). Again, you must convince management to adopt your implementation plan (i.e. ‘sell’ it to them).		5
7.       Conclusions		3
8.       References.  At the end of your report, you must provide a list of all cited references (using correct Harvard referencing style) (at least 5).		2
9.       Presentation.  Grammar, punctuation, expression and presentation must be of a high standard.		2
TOTAL		30

Today we purchase the most items online. We can get desired product at home within stipulated time. We can have more variety, quality in the desired products in online shopping than the offline shopping. Today we heard the worlds Amazon, Flipkart which is very common. There are many factors for shifting from offline shopping to online shopping.  Wolfinbarger and Gilly (2001) reported that "consumers report that shopping online results in a substantially increased sense of freedom and control as compared to offline shopping."  Lee and Lin (2005) developed the research model for examining the relationship among e?service quality dimensions and overall service quality, customer satisfaction and purchase intentions.

eCommerce becoming very raising and popular business in every corner of the world. Online shopping is most popular from the all eCommerce business.  According to an online survey, “Australians spent a total of $1.95 billion per month on online shopping alone”. Online shopping for electronic device is very common in over the world.

eCommerce business increasing exponentially in recent decade bring new challenges to the service provider. Business competition and customer satisfaction are the most important factors in the eCommerce business.

About Data:

For this case study, we have developed the data. We have data regarding the sale of laptops (1390 products) in the month. We assume that this data is from company Computer World. We considered the following attributes Product Name, Product Price (in $), Profit (in $), Sale Price (in $), Number of customers who bought this products, Shipping Type (Free or Paid), Customer Type (New or Existing), Region (NSW, QLD, WA, VIC, TAS, SA), Product Brand (Acer, Asus, Haier and HP), CPU Type (i3, i5 and i7), Operating System (Windows 10 Home and Windows 10 Pro), Screen Size in inches (11.1, 14, 15.6 and 17.1).

We need to define following variables for the study objectives as

Total Monthly sale amount (in $) = Sale Price (in $) × Number of customers

Total monthly profit (in $) = Profit (in $) × Number of customers

Project Problem:

We concentrate on the following

1. Profit analysis by shipping type, customer type, region, brand, CPU type, operating system and screen size.
2. Whether the mean number of customers are significantly different for different shipping type, customer type, region, brand, CPU type, operating system and screen size.

• Correlation analysis of variables

1. Regression analysis for total monthly sales 

Research Methodology

Statistical tool and techniques are important aspect of data analysis. In literature there are many statistical tools and techniques are available. But use of proper tools and technique is important part of analysis.

For the profit analysis, we presented the total monthly sale amount (in $) and total monthly profit (in $) for shipping type, customer type, region, brand, CPU type, operating system and screen size. We have also calculated the profit percentage in profit analysis. We presented the descriptive statistics for number of customers for shipping type, customer type, region, brand, CPU type, operating system and screen size. We analysed the mean number of customers for shipping type, customer type, region, brand, CPU type, operating system and screen size by the two sample t test and one way ANOVA. We studied the correlation between product price, profit, sale price and number of customers. We try to fit the regression model for total sale amount. We used Python 3.6.5 and MS-Excel for the data analysis. The sample code are given in appendixes.  We used Grus (2015), McKinney (2012), Pedregosa et al. (2011) and Schutt and O'Neil (2013).

Assessment of Various Data Analytical Methods

Analytical Findings

  Profit Analysis:

For the profit analysis we have given the total monthly sale amount (in $), total monthly profit (in $)  and profit percentage for shipping type, customer type, region, brand, CPU type, operating system and screen size. We referred Berenson et al. (2012), Black (2009), Groebner et al. (2008), Kvanli et al. (2000) and Mendenhall and Sincich (1993). Results are represented in Table 1. 

Table 1: Profit analysis according to different attributes

Attributes	Level	Total Monthly  Sale (in $)	Total Monthly Profit (in $)	Profit %
Shipping Type	Free	2916380	149853.5	5.14%
	Paid	4528836	232483.1	5.13%
Customer Type	Existing	3011425	153767.4	5.11%
	New	4433790	228569.2	5.16%
Region	NSW	1360477	69336.3	5.10%
	QLD	1420248	72803.5	5.13%
	SA	1467640	76496.2	5.21%
	TAS	411042.6	21301.6	5.18%
	VIC	1444593	74270.5	5.14%
	WA	1341216	68128.5	5.08%
Brand	Acer	1658146	86376.4	5.21%
	Asus	1604424	97642.3	6.09%
	Haier	1830076	86780.7	4.74%
	HP	2352569	111537.2	4.74%
CPU Type	i3	2127357	108809.4	5.11%
	i5	2898768	149383.6	5.15%
	i7	2419091	124143.6	5.13%
Operating  System	Windows 10 Home	4008167	206626.8	5.16%
	Windows 10 Pro	3437049	175709.8	5.11%
Screen Size (in Inches)	11.1	1579945	80963.0	5.12%
	14	1760526	91088.5	5.17%
	15.6	2010451	102975.4	5.12%
	17.1	2094294	107309.7	5.12%
Total		7445216	382336.6	5.14%

 From Table 1 we can claim that Computer World earns on average 5.14% profit on each laptop. The profit percentage is significantly different for brand and region. On Asus laptops, company gets more profit about 6.09% whereas for HP and Haier laptops earns only 4.74%.  In SA region company gets 5.21% profit whereas only 5.08% from WA region. Profit percentage for different attributes like shipping type, customer type, CPU type, operating system and screen size. 

Descriptive statistics for number of customer:

Customer is the pillar of any business. Total monthly sale and profit are proportional to the number of customers. So in this subsection, we presented the summary statistics for the number of customers for shipping type, customer type, region, brand, CPU type, operating system and screen size. We used the well-known books for this section such as Bickel  and Doksum (2015), Casella and Berger (2002), DeGroot and Schervish (2012), Hodges Jr and Lehmann (2005), Papoulis (1990), Pillers (2002) and Ross (2014). Table 2 represents the size, mean, standard deviation, minimum and maximum of number of customers.

From Table2, we can observed following

6. i) Computer World get averagely 6.6 customers for each laptop
7. ii) mean number of customer for the laptops which has free shipping is more than those who has paid shipping

• iii) mean number of customers for laptops in TAS region is more than other region

1. iv) mean number of customers for HP brand laptops is more than other brand
2. v) mean number of customers for laptops having i5 CPU type is more than other CPU type
3. vi) mean number of customers for laptops having Windows 10 Home is more than Windows 10 Pro

• vii) mean number of customers for 15.6 inches laptop is more than other screen size laptops. 

Table 2: Summary statistics for numbers of customer for shipping type, customer type, region, brand, CPU type, operating system and screen size.

Attributes	Level	Size	Mean	SD	Min	Max
Shipping Type	Free	539	6.72	2.857	1	16
	Paid	851	6.53	2.844	1	18
Customer Type	Existing	561	6.61	2.792	1	16
	New	829	6.60	2.889	1	18
Region	NSW	264	6.38	3.011	1	18
	QLD	267	6.55	2.594	1	15
	SA	270	6.66	2.981	1	17
	TAS	71	7.17	3.247	1	14
	VIC	271	6.67	2.800	1	16
	WA	247	6.62	2.719	1	14
Brand	Acer	340	6.15	2.607	1	16
	Asus	349	6.11	2.654	1	15
	Haier	347	6.27	2.647	1	16
	HP	354	7.86	3.080	1	18
CPU Type	i3	460	6.10	2.529	1	14
	i5	470	7.61	3.045	1	18
	i7	460	6.09	2.673	1	17
Operating  System	Windows 10 Home	701	7.15	2.851	1	18
	Windows 10 Pro	689	6.05	2.741	1	17
Screen Size (in Inches)	11.1	342	6.31	2.840	1	17
	14	350	6.42	2.740	1	16
	15.6	338	7.12	2.996	1	18
	17.1	360	6.59	2.766	1	14
Total		1390	6.60	2.849	1	18

 Two Sample t-test:

Here we are interested to know whether there is significant difference between the mean of number of customers for shipping type, customer type and operating system. Our null hypothesis is that there is no significant difference between mean of number of customers for levels of attributes and alternative hypothesis is that there is significant difference between mean of number of customers for levels of attributes.  We used two sample independent test assuming unequal variances. In Table 4, we have represented the value of test statistic and p-value of two sample independent test assuming unequal variances.

Table 4: Two sample independent test for shipping type, customer type and Operating system

Attributes	Levels	Test Statistic	p-value
Shipping Type	Free and Paid	1.16	0.245
Customer Type	New and Existing	0.02	0.985
Operating System	Windows 10 Home and Windows 10 Pro	7.30	0.000

 From Table 4, we can claim that there is no significant difference between the mean of number of customers between Free and Paid shipping. A similar conclusion for customer type. We observed a significant difference between the mean of number of customers for laptops having Windows 10 Home and Windows 10 Pro. Mean of number of customers for laptops having Windows 10 Home is significantly more than laptops having Windows 10 Pro.

Good Use of Python to Develop a Predictive Model of Monthly Sales

One way ANOVA:

Here we are interested to know whether there is significant difference between the mean of number of customers for region, brand, CPU type, Screen size. We used one way ANOVA for testing the null hypothesis that there is no significant difference between mean of number of customers for levels of attributes and alternative hypothesis is that there is significant difference between mean of number of customers for levels of attributes. In Table 5, we have represented the value of F statistic and p-value. 

Table 5: Output of one way ANOVA for region, brand, CPU type and Screen size

Attributes	Level	F Statistic	P Value
Region	NSW, QLD, WA, VIC, TAS and SA	0.96	0.440
Brand	Acer, Asus, Haier and HP	32.87	0.000
CPU Type	i3, i5 and i7	46.81	0.000
Screen Size (in inches)	11.1, 14, 15.6 and 17.1	5.47	0.001

 From Table 5, we observed that there is no significant difference between the mean of number of customers for different region whereas there is significant difference between the mean of number of customers for different brands, CPU type and screen size (in inches). Mean number of customers for HP brand laptop is significantly more than other brand, for laptops having i5 CPU type is significantly more than other CPU type and for laptop having screen size 15.6 inches is significantly more than other screen type.

Correlation Analysis:

In this subsection, we studied the correlation between product price, Sale price, profit and number of customers. Table 6 represents the correlation coefficient between product price, Sale price, profit and number of customers.

Table 6: Pearson’s correlation coefficient for Product Price, Sale Price, Profit and Numbers of customers

	Product Price	Sale Price	Profit	Numbers of customer
Product Price	1	0.999	0.683	0.043
Sale Price	0.999	1	0.713	0.038
Profit	0.683	0.713	1	-0.057
Numbers of customer	0.043	0.038	-0.057	1

 From Table 6, we observed that

1. i) Product price is positive correlated with sale price, profit and number of customers.
2. ii) Sale price is positively related with profit and number of customers.

• iii) Profit is negatively correlated with number of customers.

Regression analysis:

In this subsection, we have fitted the simple regression model to the total profit. We used number of customers as a predictor variable. Table 7 presents the output of regression analysis. 

Table 7: Output of Regression Analysis

Regression Statistics
R Square	0.862291
ANOVA
	df	SS	MS	F	Significance F
Regression	1	18906460	18906460	8691.225	0
Residual	1388	3019386	2175.35
Total	1389	21925846
	Coefficients	Standard Error	t Stat	P-value	Lower 95%	Upper 95%
Intercept	4.636421	3.158995	1.467688	0.142415	-1.5605	10.83334
Number of Customer	40.94684	0.439218	93.22674	0	40.08524	41.80844

From Table 7, we observed that there is significant relationship between total monthly profit and number of customers. We also observed R² as 0.86 suggests that model fitting is adequate.  Each addition customer gets a profit of $ 40.96484.

Recommendations to the company

1. i) Company needs to concentrate on the products which earns more profit.
2. Company needs to attract the number of customers so that profit will be maximize.

• Company should give the offers so that more number of customers will be attracted.

15. There is more demand of HP laptops, laptops with Windows 10 Home Operating system, laptops with i3 CPU type, laptops with 15.6 inches screen size.

An implementation plan based on the recommendations you have provided

1. i) Company can use more marketing techniques for the brands which earns more profit.
2. Company can give the offers to the customer like money back, free delivery certain discount, pay later.

• Company can be used more advertising hording to know about company and products.

1. Company should keep the stock of laptops with proper specification which has more demand.

Conclusions

From Profit Analysis, the profit percentage is significantly different for brand and region. On Asus laptops, company gets more profit about 6.09% whereas for HP and Haier laptops earns only 4.74%.  In SA region company gets 5.21% profit whereas only 5.08% from WA region.

From two sample t test, we observed that there is no significant difference between the mean of number of customers between Free and Paid shipping. A similar conclusion for customer type. We observed a significant difference between the mean of number of customers for laptops having Windows 10 Home and Windows 10 Pro. Mean of number of customers for laptops having Windows 10 Home is significantly more than laptops having Windows 10 Pro.

Recommendations

From one way ANOVA, we observed that there is no significant difference between the mean of number of customers for different region whereas there is significant difference between the mean of number of customers for different brands, CPU type and screen size (in inches). Mean number of customers for HP brand laptop is significantly more than other brand, for laptops having i5 CPU type is significantly more than other CPU type and for laptop having screen size 15.6 inches is significantly more than other screen type.

From Correlation analysis, we observed that (i) Product price is positive correlated with sale price, profit and number of customers (ii) Sale price is positively related with profit and number of customers and (iii) Profit is negatively correlated with number of customers. From Regression analysis, we observed that there is significant relationship between total monthly profit and number of customers. We also observed R² as 0.86 suggests that model fitting is adequate.  Each addition customer gets a profit of $ 40.96484.

We have also provided recommendations and plan for company.

References

Berenson, M., Levine, D., Szabat, K.A. and Krehbiel, T.C., (2012). Basic business statistics: Concepts and applications. Pearson higher education AU.

Bickel, P.J. and Doksum, K.A., (2015). Mathematical statistics: basic ideas and selected topics, volume I (Vol. 117). CRC Press.

Black, K., (2009). Business statistics: Contemporary decision making. John Wiley & Sons.

Casella, G. and Berger, R.L., (2002). Statistical inference (Vol. 2). Pacific Grove, CA: Duxbury.

DeGroot, M.H. and Schervish, M.J., (2012). Probability and statistics. Pearson Education.

Groebner, D.F., Shannon, P.W., Fry, P.C. and Smith, K.D., (2008). Business statistics. Pearson Education.

Grus, J., (2015). Data science from scratch: first principles with python. " O'Reilly Media, Inc.".

Hodges Jr, J.L. and Lehmann, E.L., (2005). Basic concepts of probability and statistics. Society for Industrial and Applied Mathematics.

Kvanli, A.H., Pavur, R.J. and Guynes, C.S., (2000). Introduction to business statistics. Cincinnati, OH: South-Western.

Lee, G. G., & Lin, H. F. (2005). Customer perceptions of e-service quality in online shopping. International Journal of Retail & Distribution Management, 33(2), 161-176.

McKinney, W., (2012). Python for data analysis: Data wrangling with Pandas, NumPy, and IPython. " O'Reilly Media, Inc.".

Mendenhall, W. and Sincich, T., (1993). A second course in business statistics: Regression analysis. San Francisco: Dellen.

Papoulis, A., (1990). Probability & statistics (Vol. 2). Englewood Cliffs: Prentice-Hall.

Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V. and Vanderplas, J., (2011). Scikit-learn: Machine learning in Python. Journal of machine learning research, 12(Oct), pp.2825-2830.

Pillers Dobler, Carolyn. "Mathematical statistics: Basic ideas and selected topics." (2002): 332-332.

Ross, S.M., (2014). Introduction to probability and statistics for engineers and scientists. Academic Press.

Schutt, R. and O'Neil, C., (2013). Doing data science: Straight talk from the frontline. " O'Reilly Media, Inc.".

Wolfinbarger, M. and Gilly, M.C., (2001). Shopping online for freedom, control, and fun. California Management Review, 43(2), pp.34-55.