Analyzing Sales Performance Of Good Harvest

Problem definition and business intelligence required

Good Harvest is a business enterprise that is concerned with growing organic food. The Good Harvest thrives in;

• Empowering the individual to make conscious decisions
• Inspiring dynamic and passionate communities
• Believing integrity is the essence of everything

The company offers a number of products which include:

• Organic farm boxes for students and singles
• Organic farm boxes for couples
• Organic farm boxes for families
• Organic cold pressed juices
• Organic farm fresh extras

The company delivers the products to the customers and they have specific days of delivery

The CEO is greatly concerned with the business revenue (i.e. lead generation/new business), Cost of Goods (COGS margins) and average sales. This study sought to analyse the business standing of the company making recommendation to the CEO on how best the revenues can be improved.

The study sought to answer a number of research questions. The listed below are some of the research questions that this study sought to answer;

1. What are the top/worst selling products in terms of sales?

To answer this research question, analysis of the product class had to be done. The average sales of the different product classes are presented in order to compare the total sales received from each and every product class.

1. Is there a difference in payments methods?

Four different payments methods are provided, namely; cash, credit, visa and MasterCard payment options. To answer the research question, two independent t-tests will be applied. An independent-samples t-test, refers to an inferential statistical test that is used to test whether there is a statistically significant difference between the means of two unrelated groups. This approach is used analyze the differences two independent group means and their associated procedures (Derrick, Toher, & White, 2017). So cash and credit will be compared together while Visa and MasterCard will also be compared together hence the need for two independent t-tests.

1. Are the differences in sales performance based on where the product is located in the shop? How does this effect both profits and revenue?

The third research question that this study sought to answer was whether there are differences in sales performance based on where the product is located in the shop. This question will be answered using ANOVA test. ANOVA test is useful when it comes to comparing (testing) three or more means (groups or variables) for statistical significance (Hinkelmann & Kempthorne, 2010). Since there are more than 3 locations (factors), ANOVA would be an ideal methodology to test for the differences in sales performance.

1. Is there a difference in sales and gross profits between different months of the year?

This is another research question that will be answered using ANOVA. There are 12 months in a year and as such, we will try to verify whether there is variation in the mean sales and mean gross profits made in different months. Since ANOVA tests whether the means for three or more factors are vary statistically, the test (ANOVA) will be ideal for answering this research question (Gelman, 2005).

1. Are their differences in sales performance between different seasons?

The last research question to be tested will be to check whether the different seasons have differences in sales performance. ANOVA will be used to test for this since there are more than two seasons to be compared. 

Visualise the descriptive statistics

This section presents the summary statistics of some of the variables under investigation.

Table 1: Descriptive statistics

	Total Sales ($)	Total Profit
N	1033	1032
Minimum	2	0.48
Maximum	17276	8702.93
Mean	370.32	165.05
Std. Deviation	1015.145	482.52

Table 1 above gives the summary statistics for the total sales and total profit. As can be seen, the average sales made by the company was found to be $370.32 with minimum and maximum sales being $2 and $17276 respectively. On the other hand, the average total profit was found to be $165.05 with minimum and maximum total profit being $0.48 and $8702.93 respectively.

Analysis of seasons revealed that both Winter and Autumn were represented by 25.14% (n = 92) while Summer and Spring were both represented by 24.86% (n = 91).

In terms of the location of product in the shop, majority of products were located to the left of the shop (36.4%, n =376) and it was closely followed by products placed on the right (30.1%, n = 311). Outside front had the lowest proportion of products placed in it (1.2%, n = 12). 

A normality test of the gross sales was conducted since it was to be used later for the both the ANOVA and t-tests. Ideally, for the variable to be used in testing either ANOVA or t-test, the variable has to follow a normal distribution. Through the visualization of the histogram given below, we can conclude that the variable gross sales follows a normal distribution since the shape of the histogram appears to be bell-shaped-a condition attributed with normally distributed data

In terms of sales revenue, the top five selling products are Water, Fruit, Vegetable, Dairy and Drinks. These products have sales amounting to over $500. On the other hand the worst selling products (bottom five) are Juicing, Herbal Teas, Spices, Snacks, Salad greens. All the bottom five selling products had sales revenue less than $30.

Product Class	N	Minimum	Maximum	Mean	Std. Deviation
Water	12	15	6500	1866.88	2541.63
Fruit	54	3	17276	1048.68	2469.41
Vegetable	76	4	5554	871.49	1226.30
Dairy	66	10	10814	619.05	1473.79
Drinks	59	5	11910	574.25	1729.24
Coconut Water	11	21	1794	514.23	562.67
Bakery	44	7	3793	432.67	884.02
Fridge	51	9	1535	354.21	389.33
Dry Goods	84	2	3300	341.26	604.44
Health products	17	15	2914	332.78	757.81
Oils & Vinegars	25	9	1815	310.81	421.69
Snacks & Chocolates	110	4	2972	246.14	481.38
Ayurvedic	3	10	504	226.25	252.68
Milks non dairy	9	12	968	224.55	297.15
Freezer	62	5	3252	202.45	421.30
Household	25	7	987	196.23	255.52
Meats Small goods	34	5	1423	176.72	259.03
Pasta	15	8	488	114.22	138.52
Spreads, Sauces, Sweeteners	28	6	1310	113.60	296.33
Grocery	64	5	597	108.74	108.26
Market	2	20	158	88.75	97.23
Tea Coffee	24	5	583	88.55	147.16
Personal Products	96	2	676	84.38	114.32
Packaging	8	2	320	62.27	105.62
Tinned Goods	8	6	109	48.09	32.51
Harvest Kitchen	4	24	80	44.98	24.20
Chocolates & Slices	5	19	61	37.01	16.03
Pastas	1	36	36	35.80
Other	9	0	88	33.53	25.95
Stocks Sauces	6	20	49	32.29	12.17
Salad Greens	1	25	25	24.50
Snacks	2	20	21	20.33	0.74
Spices	14	4	129	18.99	32.06
Herbal Teas	4	2	54	17.96	24.37
Juicing	1	5	5	5.00

This is the second research questions that the study sought to answer.

Two different t-tests were conducted, the first t-test was to compare the cash and the credit methods. The results are given below;

Table 3: Two-Sample Assuming Equal

Variances
	Cash	Credit
Mean	404.2923	584.8115
Variance	23608.25	52380.18
Observations	366	366
Pooled Variance	37994.21
Hypothesized Mean Difference	0
df	730
t Stat	-12.5282
P(T<=t) one-tail	4.69E-33
t Critical one-tail	1.646944
P(T<=t) two-tail	9.37E-33
t Critical two-tail	1.963219

An independent samples t-test was done to compare the mean total cash received from either cash or credit payments. Results showed that the cash payments (M = 404.29, SD = 153.65, N = 366) had significant difference in terms of the total cash received when compared to the credit payments (M = 584.81, SD = 228.87, N = 366), t (730) = -12.53, p < .05, two-tailed. The difference of 180.52 showed a significant difference.  Essentially results showed that credit payment method attracts more total cash received as compared to the cash payment methods.

Normality test

Next, we compared the total cash received through the Visa payment method and the MasterCard payment method. An independent t-test was performed to compare the mean total cash received for the two payments methods. Results are in the table below;

Table 4: Two-Sample Assuming Equal Variances

	Visa	MasterCard
Mean	579.5983	152.5472
Variance	48734.1	12000.98
Observations	351	53
Pooled Variance	43982.56
Hypothesized Mean Difference	0
df	402
t Stat	13.81785
P(T<=t) one-tail	4.15E-36
t Critical one-tail	1.648653
P(T<=t) two-tail	8.29E-36
t Critical two-tail	1.965883

An independent samples t-test was done to compare the mean total cash received from either Visa or MasterCard payments. Results showed that the Visa payments (M = 579.60, SD = 220.76, N = 351) had significant difference in terms of the total cash received when compared to the MasterCard payments (M = 152.55, SD = 109.55, N = 53), t (402) = 13.82, p < .05, two-tailed. The difference of 427.05 showed a significant difference.  Essentially results showed that Visa payment method attracts more total cash received as compared to the MasterCard payment methods.

Are the differences in sales performance based on where the product is located in the shop? 

The next research question that this study aimed at answering was on whether there are differences in sales performance based on where the product is located in the shop. Using ANOVA (Howell, 2007), we checked whether the mean sales varies across the different locations. Results are given below;

Table 5: ANOVA

Total Sales ($)
	Sum of Squares	df	Mean Square	F	Sig.
Between Groups	134299725.02	4	33574931.256	37.18	.000
Within Groups	929333380.82	1029	903142.255
Total	1063633105.84	1033

A one-way between subjects ANOVA was conducted to compare the effect of difference locations of the product on the sales (Gelman, 2005). There was a significant effect of location where the product is located on the total sales at the p<.05 level for the five conditions [F(4, 1029) = 37.18, p = 0.000]. 

Table 6: Test of Homogeneity of Variances

Total Sales ($)
Levene Statistic	df1	df2	Sig.
47.870	4	1029	.000

Test of homogeneity of variances showed that the groups we are comparing do not similar population variances. For this reason, post hoc analysis employed would be the one that does not assume equal variances. It is for this reason that Games-Howell was used. 

Table 7: Multiple Comparisons

Dependent Variable:   Total Sales ($)
Games-Howell
(I) Location of product in shop	(J) Location of product in shop	Mean Difference (I-J)	Std. Error	Sig.	95% Confidence Interval
Lower Bound	Upper Bound
Front	Left	357.512*	117.700	.023	32.86	682.17
Outside Front	-2808.054	1367.254	.303	-7218.81	1602.70
Rear	40.243	140.538	.999	-345.63	426.11
Right	336.423*	119.784	.043	6.27	666.58
Left	Front	-357.512*	117.700	.023	-682.17	-32.86
Outside Front	-3165.566	1362.538	.208	-7571.64	1240.50
Rear	-317.269*	82.914	.002	-545.45	-89.09
Right	-21.089	38.365	.982	-126.07	83.90
Outside Front	Front	2808.054	1367.254	.303	-1602.70	7218.81
Left	3165.566	1362.538	.208	-1240.50	7571.64
Rear	2848.297	1364.700	.290	-1559.90	7256.50
Right	3144.477	1362.719	.213	-1261.77	7550.73
Rear	Front	-40.243	140.538	.999	-426.11	345.63
Left	317.269*	82.914	.002	89.09	545.45
Outside Front	-2848.297	1364.700	.290	-7256.50	1559.90
Right	296.181*	85.846	.006	60.19	532.18
Right	Front	-336.423*	119.784	.043	-666.58	-6.27
Left	21.089	38.365	.982	-83.90	126.07
Outside Front	-3144.477	1362.71	.213	-7550.73	1261.77
Rear	-296.181*	85.85	.006	-532.18	-60.19
*. The mean difference is significant at the 0.05 level.

Post hoc comparisons using the Games-Howell test indicated that the mean total from front location (M = 572.75, SD = 1430.66) was significantly different from the left location (M = 218.22, SD = 427.61) and the right location (M = 239.89, SD = 553.00). However, front location (M = 572.75, SD = 1430.66) did not significantly differ from outside front (M = 3384.37, SD = 4719.35) and rear location (M = 536.07, SD = 1072.153). In overall, these results suggest that the location of the product in the shop have effect on the total sales. In particular, total sales from outside front was much higher compared to all other locations with left location resulting to the lowest total sales. 

Results of the selected analytics methods and technical analysis

The third research question we sought to answer was whether there is a difference in sales and gross profits between different months of the year. Using ANOVA, we compared the mean sales between the different months of the year. Results are given below;

Table 8: ANOVA

Gross Sales
	Sum of Squares	df	Mean Square	F	Sig.
Between Groups	1508892.47	11	137172.04	1.300	.222
Within Groups	37349615.46	354	105507.39
Total	38858507.93	365

A one-way between subjects ANOVA was conducted to compare the effect of difference months of the year on the sales. There was no significant effect of month of the year on the total sales at the p>.05 level for the twelve conditions [F(11, 354) = 1.30, p = 0.222].

Is there a difference in gross profits between different months of the year?

The next research question we sought to answer was whether there is a difference in gross profits between different months of the year (Leigh, 2008). Using ANOVA, we compared the mean gross profits between the different months of the year. Results are given below;

Table 9: ANOVA

Profit Total
	Sum of Squares	df	Mean Square	F	Sig.
Between Groups	35370.95	11	3215.541	3.867	.000
Within Groups	294370.01	354	831.554
Total	329740.95	365

A one-way between subjects ANOVA was conducted to compare the effect of difference months of the year on gross profits. There was a significant effect of month of the year on the gross profit at the p<.05 level for the twelve conditions [F(11, 354) = 1.30, p = 0.000].

Post hoc comparisons using the Games-Howell test indicated that the mean total gross profits from March (M = 19.34, SD = 16.23) was significantly different from the October (M = 46.26, SD = 38.94).

The last research question we sought to answer was whether differences in sales performance exists between seasons. Again this was tested using ANOVA test. Results are given below;

Table 10: ANOVA

Gross_Sales
	Sum of Squares	df	Mean Square	F	Sig.
Between Groups	560240.41	3	186746.80	1.765	.153
Within Groups	38298267.52	362	105796.32
Total	38858507.93	365

A one-way between subjects ANOVA was conducted to compare the effect of difference seasons on sales performance. There was an insignificant effect of seasons on the sales performance at the p > .05 level for the four conditions [F(3, 362) = 1.77, p = 0.153]. 

The main aim of this study was to analyse the business performance of the Good Harvest Company and present recommendations to the CEO on the way forward of how the company could maximize on sales and eventual profit. A number of research questions were formulated and eventually answered. The best and worst selling products were analysed where it was established that Water, Fruit, Vegetable, Dairy and Drinks were among the top performing products while on contrary Juicing, Herbal Teas, Spices, Snacks, Salad greens were the worst performing products in terms of sales.

In terms of payment methods, cash received from credit was found to be much higher compared to all other payment methods with cash payment method resulting to the lowest total cash received.

Location of the product in the shop was established to have a very significant role in determining the total sales. In particular, total sales from outside front was much higher compared to all other locations with left location resulting to the lowest total sales.

There was no significant effect of month of the year nor the season on the total sales of the products. However, month of the year played a critical role in the total gross profits made by the company.

Having looked at the above results, we recommend the following actions to be undertaken by the CEO in order to improve on the company sales

• The company should do away with non-performing products such as Juicing, Herbal Teas, Spices, Snacks, and Salad greensbut instead focus more on products such as Water, Fruit, Vegetable, Dairy and Drinks which have a great potential of making the company great.
• Location of the product was found to play a critical role on the sales; the management should take keen note on how they display out the products in order to attract the customers.
• Payments methods varied in terms of cash received; the management should for instance ensure that credit services are working well at the same time try to find out from the customers which payment channels fits them best so that they can improve
• The management should focus their promotions throughout the year since there was no significant difference in sales throughout the year however, there seems to be some cost that are either lower or higher at some periods of the year that makes the total gross profits to significantly differ. The management need to find out which costs are these and make appropriate measures e.g. purchasing such materials when their prices are lower in order to maximize on the profits.

References

Gelman, A. (2005). Analysis of variance? Why it is more important than ever. The Annals of Statistics, 33, 1–53.

Hinkelmann, K., & Kempthorne, O. (2010). Design and Analysis of Experiments.

Howell, D. C. (2007). Statistical methods for psychology.

Leigh, E. S. (2008). Consumer rites: the buying & selling of American holidays. 106–191.