Harvard Case Study

GoodBelly is a probiotic juice produced by the Colorado-based NextFoods Inc. By 2008, the global market for probiotic and probiotic food and beverages was substantial, at $15.4 billion, and still growing (Ahn & Dickerson, 2012). Goodbelly products were dairy-free, soy-free, and vegan while most probiotic products were dairy-based. Probiotics are live bacteria that are beneficial to the host organism. Since its first product launch in January of 2008, GoodBelly products have been sold in nationwide retailers such as Whole Foods Market and Safeway. It is designed to naturally renew individual digestive and immune health. GoodBelly set up in-store tables near their product for product demonstration by offering sample, information, and coupons. As a small start-up, GoodBelly lack sufficient budget to promote the products nationwide and faces the challenge of effectiveness of promotion for customer satisfaction and customer retention through in store demo. Management modified the program to focus only on Whole Foods stores, which represented approximately 75% of sales in 2009 (Ahn & Dickerson, 2012). In order to increase sales and product awareness at Whole Foods Market, the marketing team led by Marty Wellbeing developed a strategy in 2009 that included demonstrations, sales representatives and endcap displays. The company management had concerns on incur expenditure in the campaigns months after the program was launched. To evaluate the effectiveness of the promotion, GoodBelly marketing manager, Marty Wellbeing tasked her intern, Caroline Dickerson to assess GoodBelly’s sales spreadsheet based on a total of 1,386 observations collected from 126 Whole Foods stores over a 10- week period. Dickerson was responsible to identify the impact of the in-store demos and endcap promotions. A regression analysis is used to generate the most accurate regression. It is a reliable method of identifying which variables have impact on a topic of interest. It allows to confidently determine which factors matter most, which factors can be ignored, and how these factors influence each other. It is helpful statistical method that can be leveraged across an organization to determine the degree to which particular independent variables are influencing dependent variables. After analyzing the data, the management will be able to draw a recommendation for future campaigns and research backed with statistical data.

BISHWO PANDEY

GoodBelly is a start-up probiotic company that sells probiotics drink to the consumers through nationwide retailers such as Whole food markets and Safeway Inc. The main ingredient in its products is active cultures of a proprietary probiotic strain called Lactobacillus plantarum. Unlike the competition, its probiotic drinks are dairy-free, soy-free, and vegan. As a new start up Good belly has small marketing budget. They must optimize the allotment of its constrained marketing budget. Its low-cost promotional programs have yielded some results however management still do not believe the efficacy of the promotional programs to continue such programs. Some of the promotional programs includes: Endcaps, in-store demonstrations, and sales representatives. These promotional efforts are targeted to increase sales and product awareness.

Problem

GoodBelly has products that are beneficial and affordable to the probiotic consumers as compare with others probiotic available in the market. Probiotics are live bacteria that are beneficial to the host organism. Probiotics and other pathogens enter the bloodstream through the stomach lining (Ahn & Dickerson, 2012). GoodBelly has a limited marketing resources and faced the challenge of raising product awareness. As a start-up, GoodBelly did not have enough marketing budget to place nationwide advertisement. As a result, they decided to execute in-store demo and end display promotion. The management focused only on Whole Foods stores for in-store demo. But the problem with their promotion lies within the budget, worker training, and product displacement. Some executives questioned whether the increase in sales volume could justify the associated costs. To justify the effectiveness of in-store demo and endcap displays, the statistical analysis of 1,386 observations collected from 126 Whole Foods stores over a 10- week period is discussed below.

Descriptive Statistics

Descriptive statistics are broadly defined as quantitative measures meant to summarize and interpret properties of a dataset which can be either a representation of the entire or a sample of a population. Descriptive statistics provide simple summaries about the sample and the measures and are typically distinguished from inferential statistics. // Add

BISHWO PANDEY

Graphical analysis on the figure 1 shows the total amount of unit sales weekly across all regions. It shows trend that the weekly sales, increased 24% from May 11 to July 13 of 2010.

BISHWO PANDEY

Similarly, Table 2 and Figure 2 shows the sales result by region. The region with most sales units was NC with 51,540 units sold, that translates to 14.7% of the total sales.MW and MA region were second and third with 14.1% and 13.2% of total sales respectively. Likewise, SW region ranks the lowest contributing to only 3.9% of total sales.

BISHWO PANDEY

Data in Table 3 reveals top 10 stores by units sold volume. Store in Redwood city located in NC region was the no.1 store when compared to unit sales volume, contributing to 1.5% of the total, followed by Redmond, Stevens Creek, and so on.

BISHWO PANDEY

Units Sold		Average Retail Price
Mean	253.8207176		Mean	4.107093
Standard Error	2.981510884		Standard Error	0.012459
Median	236.7352663		Median	4.096667
Mode	#N/A		Mode	4.204286
Standard Deviation	110.9987311		Standard Deviation	0.463828
Sample Variance	12320.71832		Sample Variance	0.215137
Kurtosis	10.79476907		Kurtosis	0.349554
Skewness	2.597404809		Skewness	0.27252
Range	993.6404045		Range	3.362253
Minimum	47.5598519		Minimum	2.889286
Maximum	1041.200256		Maximum	6.251538
Sum	351795.5145		Sum	5692.431
Count	1386		Count	1386

The table indicates the minimum quantity of units sold per store per week is 48, maximum of 1041 units, and an average of 253 for the period of 05/11/2010-07/13/2010. The average retail price of GoodBelly was 4.10 with minimum of 2.8 and maximum of 6.25.

According to above table “units sold by Region”, the region with most units sold was NC with 51539 units (14.66%) of total unit. The region with least sales is SW with 13733 units (3.91%) of total unit sold. It shows that GoodBelly can focused in NC for promotion while it has to take into consideration when taking SW as for promotion.

Regression Analysis

We conducted Regression Analysis based on data available for GoodBelly on Excel to determine which variables have a positive relationship with the units sold. We have the Units Sold as the dependent variable, and other eight quantitative variables as explanatory to analyze the impact of the in-store demos and endcap promotions.

Regression table

The equation can be written as

Y= 298.48-28.53Price+77.43Salesrep+305.10Endcap+111.13Demo+73.51Demo1-3+67.56Demo4-5-1.59Natural-1.01Fitness

The positive coefficients indicate that when the value of one variable increases, the value of the other variable also tends to increase. The coefficients describe the mathematical relationship between each independent variable and the dependent variable. In this analysis, unit sold is the dependent variable and Average Retail price, sales Rep, Endcap, Demo, Demo1-3, Demo4-5, natural, and fitness are independent variables. The variables associated with demo are significant as their coefficients are positive. If GoodBelly has sales Rep, the number of sales increases by 77 units. If GoodBelly places an item on an endcap, the number of sales increases by 305 units. If the store has a demo during the current week, the number of sales increases by 111 units. If GoodBelly places the demo 1-3 weeks ago, the sales increase by 74 units. And if the demo was 4-5 weeks ago, the sales increase by 68 units. The regression analysis shows that R square is 0.67 (67%). It means 67% of the variation in units sold per store per week is explained by the variations of the average retail price, sales rep, demo, demo1-3, and demo4-5. Natural and Fitness have p-values of 0.34, and 0.23 respectively which are higher than 0.05 and is not statistically significant and indicates strong evidence for the null hypothesis. The marketing promotion used by GoodBelly is currently working and is beneficial to the company.

//Recommendation

-Analyzing the company/business, identifying problems and executing new plans

hiring an analytics team,

-developing a work place training program for new and existing managers, lower-level employees, and other positions

-incentive/reward for employee

-analyzing more than two months of data is recommended

-it should more focused on Endcap promotion rather than sales rep and demo

-

//conclusion

Statistical Analysis is beneficial when developing new products, processes, and changing target demographics. GoodBelly showed that their marketing tactics were good, but not as seamless as they could be. Due to the lack of worker training, uneven project distribution, and unnecessary marketing costs, the managers have valid concerns. Statistical Analysis is essential when GoodBelly needs to justify the marketing department expense. //ADD

Regression Statistics

Multiple R0.820143

R Square0.672635

Adjusted R Square0.670733

Standard Error63.69303

Observations1386

ANOVA

dfSSMSFSignificance F

Regression811477979.251434747353.66468350

Residual13775586215.6174056.801

Total138517064194.87

CoefficientsStandard Errort StatP-valueLower 95%Upper 95%Lower 95.0%Upper 95.0%

Intercept298.48813116.1830939418.444444.62343E-68266.7419452330.2343159266.7419452330.2343159

Average Retail Price-28.5353653.952152658-7.220218.55578E-13-36.28825657-20.78247365-36.28825657-20.78247365

Sales Rep77.43691363.86445275320.038261.39013E-7869.8560620685.0177652269.8560620685.01776522

Endcap305.1021239.05573741433.691585.4001E-182287.3375889322.8666564287.3375889322.8666564

Demo111.1328497.4036982715.010452.77853E-4796.60910129125.656597296.60910129125.6565972

Demo1-373.51717094.8953838515.017652.53098E-4763.9139539383.1203879563.9139539383.12038795

Demo4-567.56981066.54197318110.328663.87316E-2454.7364986580.4031225754.7364986580.40312257

Natural-1.59416821.776400688-0.897410.369654508-5.0789125821.890576155-5.0789125821.890576155

Fitness-1.01967121.0840229-0.940640.347056315-3.1461861471.106843839-3.1461861471.106843839

units sold by Region

RegionUnits soldPercentage

FL16586.763264.72%

MA46482.1241313.22%

MW49553.3226314.09%

NA25928.209137.37%

NC51539.5312114.66%

NE27981.173177.96%

PE34036.295199.68%

RM33262.919.46%

SO14796.494.21%

SP37724.0110.73%

SW13733.673.91%

Total351624.4987100.00%