William Preddy

Instacart Grocery Basket Analysis

• Objective-
  • Derive insights and suggest strategies for better targeted marketing, using Python. (Project brief)

• Skills-
  • Data Cleaning: Wrangling and Subsetting
  • Data Consistency Checks
  • Combining and Exporting Data
  • Deriving New Variables
  • Grouping Data and Aggregating Variables
  • Python Visualization and Excel Report

• Data-
  • Customer data set– provided by CareerFoundry
  • Data dictionary– provided by Instacart
  • “The Instacart Online Grocery Shopping Dataset 2017”– provided by Instacart

• Tools–
  • Excel
  • Python
    • Numpy
    • Matplotlib
    • SciPy
    • Pandas
    • Seaborn
  • Tableau

Process Highlights

Problem: Data was spread out throughout multiple databases, some of which contained inconsistencies.

Solution: Merged and cleaned multiple databases.

Problem: Categories of interest were not available.

Solution: Created categories using “If” statements and “loc()” functions.

Data insights

(To view more Python code, click here)

Loyal customers go no longer than 7 days between orders.

The majority of orders take place between 9am-3pm.

Customers above the age of 40 have much more spending power than those below 40.

The majority of shoppers are middle-aged.

Recommendations

Excel reporting (Final report)

• Create a Loyalty Program
  • Convert new and regular customers into loyal customers by incentivizing their next order within in 7 days.
• Targeted Marketing
  • Higher priced items should be advertised towards an age demographic of 40 years of age and above, due to their higher spending power.
•  Advertising
  • Commercials/advertisements should take place during off-peak hours (3pm-9am)