Profit Analysis And Descriptive Statistics For E-commerce Clothing Sales

Research Methodology

Online shopping is most popular from the all eCommerce business. eCommerce started in 1971 but recently it becomes very popular in developed and in developing countries. Australians spent a total of $1.95 billion per month on online shopping alone.

In this case study, we have data of 3990 clothing products. We considered variables like Product Name, Product Price, Sale Price, Profit, Number of customers, Shipping Type (Free or Paid), Customer Type (New or Existing), Region (NSW, QLD, WA, VIC, TAS, SA), Product Brand and Size (S, M, L, XL, XXL).

In the profit analysis, we observed that for the average company gets 6.87% profit on each product. Products which are shipped freely gives less profit than the products which has shipping charges. In the region, TAS shows maximum profit, whereas WA shows minimum profit. Arrow, Polo and Woodland brand give more than 10% profit among the all brands.

We observed that there is no significant association in the pairs of attributes at 5% level of significance, whereas the customer type and size are significantly associated at 10% level of significance.

There is no significant difference between mean total profit earned on products which are shipped freely and shipped by paying shipping whereas there is a significant difference between mean total profit earned from new customers and existing customers. We observed that, there is a significant difference between mean total profit earned from region, brand and size. We observed that product price and number of customers have significant negative correlation. A similar relation is observed in product price and profit. We also observed positive significant correlation between number of customer and profit. Regression analysis suggests that there is a significant relation between total profit (response variable) with product, price and number of customers (predictor variable).

Table of Contents

Sr. No.	Topic	Page No.
1	List of Abbreviations and assumptions made	4
2	Introduction – What is the problem?	5
3	Research Methodology	6
4	Analytical Findings	7
5	Recommendations to the company	17
6	An implementation plan based on the recommendations you have provided	17
7	Conclusion	18
8	List of References	19
9	Appendix	20

List of Abbreviations and assumptions made

L : Large size

M : Medium size

Max : Maximum

Min : Minimum

NSW : New South Wales

Q1 : First Quartile

Q3 : Second Quartile

QLD : Queensland

S : Small size

SA : South Australia

TAS : Tasmania

VIC : Victoria

WA : Western Australia

XL : Extra-large size

XXL : Double extra large

Introduction – What is the problem?

In Today’s world, eCommerce becoming very raising and popular business in every corner of the world. Online shopping is most popular from the all eCommerce business. ECommerce had very past history, starting from 1971. But recently it has become very popular in developing countries. Australians spent a total of $1.95 billion per month on online shopping alone.

Analytical Findings

In eCommerce service, customer can buy or sell the service or product online. Today, there are many service providers for online shopping like Amazon, Flipkart, Ebay, etc. The service provider gives the attractive offer to attract the customers.

As the Online Shopping business increased in the exponentially manner it offers many challenges for service providers like competition and customer total satisfaction.

About Data:

In this case study, we develop the data sets regarding to Cloths (Jacket, Suit and Jeans) for the 3990 products. We considered the following attributes

1. Product Name
2. Product Price

• Sale Price

1. Profit
2. Number of customer
3. Shipping Type (Free or Paid)

• Customer Type (New or Existing)
• Region (NSW, QLD, WA, VIC, TAS, SA)

1. Product Brand
2. Size (S, M, L, XL, XXL)

We define following variables as

Total Monthly sale amount= Sale Price × Number of customer

Total monthly profit= Profit × Number of customer

Project Problem:

We are interested to know the following things

1. Profit analysis by brand, size, shipping type, customer type and region wise.
2. Whether there is any association between shipping type, customer type, region, brand and size.

• Whether the price is significantly different for different shipping type, customer type, region, brand and size.

1. Whether the sale price is significantly different for different shipping type, customer type, region, brand and size.
2. Whether the profit is significantly different for different shipping type, customer type, region, brand and size.
3. Whether the rating is significantly different for different shipping type, customer type, region, brand and size.

• Correlation analysis of variables
• Regression analysis for total monthly sales

Research Methodology

Any analysis is incomplete without use of statistical tool and techniques. There are many tools and techniques available in literature. But selection of proper tool and techniques is the main task. For profit analysis, we summarised the total sale amount and total profit amount with profit percentage for shipping type, customer type, region, brand and size.

We used Chi-square test for association for testing whether there is any significant association between different pairs of attributes. (Shipping type, customer type, region, brand and size). We used two sample t-test for testing the mean total profit for shipping type and customer type. We used one way ANOVA for testing the mean of total profit for region, brand and size. We carry the correlation analysis for studying the relation between price, profit and number of customers. We used multiple regression analysis for studying the relation between total profits with product price and number of customers.

Analytical Findings

We used the well-known books for this analytical findings such as  Berenson et al. (2012),  Bickel  and Doksum (2015), Black (2009), Casella and Berger (2002), DeGroot and Schervish (2012), Groebner et al. (2008), Grus (2015) Hodges Jr and Lehmann (2005) Kvanli et al. (2000), McKinney (2012), Mendenhall and Sincich (1993), Papoulis (1990), Pedregosa et al. (2011), Pillers (2002), Ross (2014),Schutt and O’Neil (2013).

Profit Analysis:

In profit analysis, we have given the total sales amount, total profit, profit percentage for the shipping type, customer type, region, brand and size. Table 1 gives the profit analysis according to shipping type, customer type, region, brand and size.

From Table 1, in the profit analysis, we observed that on the average company gets 6.87% profit on each product. Products which are shipped freely gives less profit than the products which has shipping charges. In the region, TAS shows maximum profit whereas WA shows minimum profit. Arrow, Polo and Woodland brand gives more than 10% profit among the all brand.

Recommendations to the Company

 Table 1: Profit analysis according to shipping type

Attributes	Level	Total Sale (Amount)	Total Profit	Profit %
Shipping Type	Free	10204025.0	697103.7	6.83%
Paid	15325158.8	1057032.9	6.90%
Customer Type	Existing	10330624.5	695405.8	6.73%
New	15198559.3	1058730.7	6.97%
Region	NSW	3891978.6	274803.1	7.06%
QLD	4307534.8	302954.0	7.03%
SA	4547671.6	324598.1	7.14%
TAS	4224240.6	303815.7	7.19%
VIC	4156360.7	292153.7	7.03%
WA	4401397.4	255811.9	5.81%
Brand	Adidas	1728242.9	147497.6	8.53%
Arrow	2003625.0	268470.5	13.40%
Burberry	1731031.8	66194.6	3.82%
Chanel	1721340.5	54891.5	3.19%
Dior	1582728.8	75917.3	4.80%
Dolce And Gabbana	1594762.5	90168.8	5.65%
Giorgio Arman	1412979.9	63204.9	4.47%
Gucci	1545161.7	63100.3	4.08%
Hermes	1703220.1	64626.0	3.79%
Nike	1866710.0	177350.2	9.50%
Polo	1874379.1	212890.1	11.36%
Prada	1645254.2	68008.0	4.13%
Ralph Lauren	1604872.1	95173.7	5.93%
Versace	1612508.2	105967.9	6.57%
Woodland	1902367.1	200675.2	10.55%
Size	S	5243458.8	346558.7	6.61%
M	4795094.4	316434.5	6.60%
L	5049065.6	352380.9	6.98%
XL	5129794.9	355895.8	6.94%
XXL	5311770.1	382866.6	7.21%
Total	25529183.8	1754136.5	6.87%

Descriptive statistics for number of customer:

From Table 2, we have reported the summary statistics for numbers of customer for the shipping type, customer type, region, brand and size. In Table 2, we have given sample size, mean, standard deviation, minimum and maximum for numbers of customer for the shipping type, customer type, region, brand and size. We observed that average numbers of customer for WA region is minimum among the all considered region.  We can have similar type of conclusion from other attributes.

Table 2: Summary statistics for numbers of customer

Attributes	Level	Sample Size	Mean	SD	Min	Max
Shipping Type	Free	1596	261.95	44.35	151	401
Paid	2394	261.52	45.05	137	403
Customer Type	Existing	1619	261.00	44.26	141	400
New	2371	262.16	45.12	137	403
Region	NSW	608	269.37	39.92	183	401
QLD	675	269.88	42.26	187	400
SA	711	271.67	40.74	195	403
TAS	657	273.30	41.05	186	400
VIC	651	268.78	38.24	194	386
WA	688	218.77	39.08	137	340
Brand	Adidas	265	279.82	30.09	202	353
Arrow	271	331.93	32.26	251	403
Burberry	269	228.66	30.21	141	298
Chanel	266	223.21	30.42	137	291
Dior	261	242.45	32.54	160	314
Dolce And Gabbana	267	252.76	30.26	176	320
Giorgio Arman	253	241.62	32.03	165	314
Gucci	253	235.01	30.81	160	302
Hermes	274	229.64	30.21	149	303
Nike	282	291.04	29.64	203	364
Polo	265	310.22	31.60	227	380
Prada	258	237.29	30.47	151	306
Ralph Lauren	267	255.28	30.32	169	324
Versace	260	258.81	30.08	176	329
Woodland	279	300.81	29.29	225	370
Size	S	801	284.08	43.30	185	403
M	757	267.84	43.85	158	384
L	792	258.76	41.86	156	375
XL	807	252.75	42.53	151	371
XXL	833	246.02	42.34	137	364
Total	3990	3990	261.69	44.77	137

Association between different attributes under study:

We used chi-square test of association between different attributes. In this test, we have following null and alternative hypothesis as

Null hypothesis: There is no significant association between two attributes.

Against

Alternative Hypothesis: There is significant association between two attributes.

We have following ten pair of attributes for possible significant association

1. shipping type and customer type.
2. shipping type and region.

• shipping type and brand.

1. shipping type and size.
2. customer type and region.
3. customer type and brand.

• customer type and size.
• region and brand.

1. region and size.
2. Brand and size.

Following Table 3 shows the chi-square statistic and p-value for chi-square test of testing association. We observed that there is no significant association in the pairs of attributes at 5% level of significance whereas customer type and size are significantly associated at 10% level of significance. 

Table 3: Chi-squared test for association

Pairs of attributes	Chi-Square Statistic	P-Value
shipping type and customer type.	0.923	0.337
shipping type and region.	1.223	0.943
shipping type and brand.	14.541	0.41
shipping type and size.	1.61	0.807
customer type and region.	5.933	0.313
customer type and brand.	11.321	0.661
customer type and size.	8.619	0.071
region and brand.	49.492	0.97
region and size.	16.723	0.671
brand and size.	59.146	0.361

 Two Sample t-test and One way ANOVA:

We are interested to know whether the total profit is significantly different for shipping type, customer type, region, brand and size or not. We used two independent sample t test for shipping type and customer type and ANOVA for region, brand and size.

Shipping type:

In the following Table 4, we presented the output of two sample t-test. We observed the null hypothesis that is no significant difference between mean total profit earned on products which are shipped freely and shipped by paid shipping is accepted. (Here P-value=0.564343 > 0.05)

Table 4: t-Test: Two-Sample Assuming Unequal Variances for mean total profit of shipping type

	Free	Paid
Mean	436.7817	441.5342
Variance	64342.54	66199.61
Observations	1596	2394
Hypothesized Mean Difference	0
df	3451
t Stat	-0.57646
P(T<=t) one-tail	0.282172
t Critical one-tail	1.645295
P(T<=t) two-tail	0.564343
t Critical two-tail	1.960652

Customer Type:

In the following Table 5, we presented the output of two sample t-test. We observed the null hypothesis that is no significant difference between mean total profit earned on products which are new customer and existing. We reject the null hypothesis means there is significant difference between mean total profit earned from new customer and existing customers. (Here P-value = 0.038117 < 0.05)

Table 5: t-Test: Two-Sample Assuming Unequal Variances for mean total profit of customer type

	Existing	New
Mean	429.528	446.5334
Variance	63069.76	66978.3
Observations	1619	2371
Hypothesized Mean Difference	0
df	3543
t Stat	-2.07437
P(T<=t) one-tail	0.019058
t Critical one-tail	1.645284
P(T<=t) two-tail	0.038117
t Critical two-tail	1.960634

Implementation Plan

Region:

In this section, we test whether the mean of different region is not significantly different. We run the one way ANOVA for testing whether the different region have different mean or not for total profit. Table 5 represent the result.

Table 5: One way ANOVA for testing mean of total profit of region

SUMMARY
Groups	Count	Sum	Average	Variance
NSW	608	274803.1	451.9787	62371.36
QLD	675	302954	448.8208	67469.82
SA	711	324598.1	456.5375	68789.55
TAS	657	303815.7	462.4288	71415.85
VIC	651	292153.7	448.7768	68338.67
WA	688	255811.9	371.8196	49064.8
ANOVA
Source of Variation	SS	df	MS	F	P-value	F crit
Between Groups	3912545	5	782509.1	12.12329	1.11E-11	2.216343
Within Groups	2.57E+08	3984	64545.96
Total	2.61E+08	3989

Here P-value < 0.05, so we reject null hypothesis. There is significant difference between the mean of total profit for all the region. We observed that WA region has lower total profit among the all-region.

Brand:

In this section, we test whether the mean of different brand is not significantly different. We run the one way ANOVA for testing whether the different brand have different mean or not for total profit. Table 6 represent the result.

Table 6: One way ANOVA for testing mean of total profit of brand

SUMMARY
Groups	Count	Sum	Average	Variance
Adidas	265	147497.6	556.5948	10486.29
Arrow	271	268470.5	990.666	20174.06
Burberry	269	66194.56	246.0764	4935.439
Chanel	266	54891.46	206.3589	4597.23
Dior	261	75917.3	290.8709	6499.589
Dolce And Gabbana	267	90168.84	337.711	7405.071
Giorgio Arman	253	63204.89	249.8217	6201.763
Gucci	253	63100.34	249.4085	6432.721
Hermes	274	64626	235.8613	5620.921
Nike	282	177350.2	628.9013	11180.99
Polo	265	212890.1	803.359	15261.4
Prada	258	68007.96	263.5967	6271.566
Ralph Lauren	267	95173.72	356.4559	7057.027
Versace	260	105967.9	407.5687	7681.169
Woodland	279	200675.2	719.2659	12708.41
ANOVA
Source of Variation	SS	df	MS	F	P-value	F crit
Between Groups	2.26E+08	14	16124791	1814.9	0	1.694258
Within Groups	35316576	3975	8884.673
Total	2.61E+08	3989

Here P-value = 0 < 0.05, so we reject null hypothesis. There is significant difference between the mean of total profit for all the brand. We observed that woodland, arrow and polo has significantly higher total profit.

Size:

In this section, we test whether the mean of different size is significantly different or not. We run the one way ANOVA for testing whether the different size have different mean or not for total profit. Table 6 represent the result.

Table 6: One way ANOVA for testing mean of total profit of size

SUMMARY
Groups	Count	Sum	Average	Variance
L	792	352380.9	444.9254	66542.37
M	757	316434.5	418.0113	62882.49
S	801	346558.7	432.6575	63045.95
XL	807	355895.8	441.0109	64500.34
XXL	833	382866.6	459.6237	69370.15
ANOVA
Source of Variation	SS	df	MS	F	P-value	F crit
Between Groups	749477	4	187369.3	2.868328	0.021856	2.374162
Within Groups	2.6E+08	3985	65323.51
Total	2.61E+08	3989

Here P-value = 0 < 0.05, so we reject null hypothesis. There is significant difference between the mean of total profit for all the size.

Correlation Analysis:

Table 7 shows the Pearson’s correlation coefficient and P-value of its significance for Product Price, Numbers of customer and Profit.

Table 7: Pearson’s correlation coefficient and P-value of its significance for Product Price, Numbers of customer and Profit

	Product Price	Numbers of customer	Profit
Product Price	1
	0
Numbers of customer	-0.640	1
	0.000	0
Profit	-0.282	0.553	1
	0.000	0.000	0

From the correlation analysis, we can say that product price and numbers of customer have significant negative correlation. Similar relation is observed in product price and profit. We also observed positive significant correlation between numbers of customer and profit.

Regression analysis:

We used multiple regression for studying the relation between total profit with product price and numbers of customer.    In Table 8, we represented the output for the multiple linear regression.

We observed that P-value = 0 < 0.05, suggest that there is significant relation between total profit (response variable) with product price and numbers of customer (predictor variable). We also observed that all the coefficient are significant. R square is not too large but not too small also suggest that model fitting is good.

Conclusion

 

Table 8: Multiple regression output

Regression Statistics
Multiple R	0.74115
R Square	0.549303
Adjusted R Square	0.549077
Standard Error	171.7878
Observations	3990
ANOVA
	df	SS	MS	F	Significance F
Regression	2	1.43E+08	71701423	2429.647	0
Residual	3987	1.18E+08	29511.05
Total	3989	2.61E+08
	Coefficients	Standard Error	t Stat	P-value	Lower 95%	Upper 95%
Intercept	-1037.77	46.43045	-22.3511	2.4E-104	-1128.8	-946.743
Price (In $)	10.97155	1.264613	8.675814	5.89E-18	8.492201	13.4509
No. of customer	4.640963	0.079049	58.70962	0	4.485982	4.795944

Recommendations to the company

1. From profit analysis, we observed that new customer’s sales amount is larger than existing customer. So it recommended that company should attract their existing customer.
2. From profit analysis and one way ANOVA of total profit for region wise suggest that WA contributes very little as compare to other. So company should attract the people from WA region.

• From profit analysis and one way ANOVA of total profit for brand wise suggest that Arrow, Woodland, Adidas, Nike and Polo contributes more in total sales and have higher profit percentage than other brand. So should go with these product.

1. We observed there is significant relation between total profit with price and numbers of customer. So company need to attract more customer.

An implementation plan based on the recommendations you have provided

1. Company can give some offer for attracting existing customers such as free delively, cash back, etc.
2. As WA region have less numbers of customers compared to the other region which resulted in lower the total sale and low profit percentage. To increase the numbers of customer from WA region, company should offer free delivery, price cut, money back etc. offer to WA people so that they attract for online shopping.

• It is desired that company should use more advertising spots in WA region than other region.

1. Overall company should appoint more marketing executive for marketing the product in the company.

Conclusions

In the profit analysis, we observed that on the average company gets 6.87% profit. Customer paid delivery gives the profit than free delivery. New customers resulted in more profit than existing customer. In the region, TAS shows maximum profit, whereas WA shows minimum profit. Arrow, Polo and Woodland brand give more than 10% profit among the all brand.

We observed that there is no significant association in the pairs of attributes at 5% level of significance, whereas customer type and size are significantly associated at 10% level of significance.

There is no significant difference between mean total profit earned on products which are shipped freely and shipped by paid shipping whereas there is a significant difference between mean total profit earned from new customers and existing customers. We observed that, there is a significant difference between mean total profit earned from region, brand and size. We observed that product price and numbers of customer have significant negative correlation. A similar relation is observed in product price and profit. We also observed positive significant correlation between numbers of customer and profit. Regression analysis suggests that there is a significant relation between total profit (response variable) with product price and numbers of customer (predictor variable).

List of 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.

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.”.