FGG Sales Analysis
FGG Sales Analysis
Embedded Files
2021 Xiaochen Chen
Introduction
FGG (FreshGoGo) is a New York-based online grocery startup. FGG specializes on delivering fresh Asian groceries to suburbs and towns that lack Asian produce and Asian grocery stores. FGG has adopted the business model of "cut off per order + cold chain delivery," which can cover a larger service area than traditional grocery delivery services and minimize operational costs through economies of scale. Therefore, FGG can profitably supply fresh groceries to customers more than 500 miles distant.
FGG opened its Chicago warehouse in 2020 with the intention of serving Midwestern communities more quickly and effectively. After nearly a year of expansion, FGG's services now reach six states and twenty cities or towns in the Midwest. Every week, FGG has 1-2 deliveries.
Project Motivation
As FGG's business grew to new places, FGG's operations team encountered significant difficulties. Order preparation, logistics routing, warehouse inventory, and truck capacity were all under a tremendous amount of stress. Simultaneously, the decision-making team at FGG's headquarters discovered that numerous Midwestern cities were unable to reach the desired goal. Therefore, FGG needs a quantitative model to examine the key factors influencing its expansion and to provide data-driven guidance for its future business development.
Method
This project used Multiple Linear Regression to determine the key factors affecting FGG business and used regression weight to evaluate the degree of influence of these key factors.
Step 1. Literature Review
I conducted the literature review to discover how other researchers chose impact factors and which dataset should be utilized. At this point, I incorporated as many impact factors as possible in preparation for the modeling stage.
Step 2. Build Initial Model
I established a regression model with the June 2021 sales volume of FGG in 20 cities or towns as the dependent variable and the six impact factors of these cities or towns as the independent variable.
Step 3. Model Optimization
I checked the multicollinearity, p-value, and std. error to see if there is a significant association between the impact factors and FGG’s sales. Then, I removed the impact factors with a p-value more than 0.05 and rebuilt the model until all remaining impact factors had p-values less than 0.05.
Step 4. Results
Through optimization, I determined that the level of household income, the proportion of Asians in the population, and the presence of competitors are the most significant impact factors on FGG's business development.
FGG’s sales = -7.644e+01 + (1.821e-03* Median Household Income) + (1.945e+03* Asian Population Share) + (-2.060e+02* Access to Asian Supermarket)
I utilized R square and Standard Error to test the model's precision. On 16 degrees of freedom, the resulting model has a R square of 0.72 and a Residual standard error of 48.48. The model as a whole is sufficiently credible.
To use these three variables and the regression model, I created a forecast for the whole Midwest and a hot spot map indicating the locations with the most potential for FGG's business development as well as the areas where FGG must discontinue service. Here FGG modified the business accordingly and achieved 18% expansion.
Data Analysis: Excel, QGIS
Modeling: R
Web Development: HTML, ArcGIS Online
Google Sites
Report abuse