Retail Data Extraction Exercise
Extraction and analysis of retail inventory data to compute baseline ABV-to-dollar efficiencies across 150+ SKUs.
Due: 26 April 2027 ยท Weight: 30%
The brief#
Economic analysis should be supported by empirical data. In this exercise, you will use Python to collect live pricing and volume data from the digital storefront of a suburban supermarket or liquor retailer.
Your script must collect at least 150 distinct Stock Keeping Units (SKUs) and calculate the ABV-to-dollar ratio for each product. The dataset must use a stratified sample rather than random selection and include three beverage categories: cask wine, bottled spirits, and ready-to-drink (RTD) products. This ensures that the dataset captures differences in pricing, taxation, and promotional activity relevant to the Capstone Portfolio.
What you submit#
You must upload a single .zip archive containing three files:
- Your raw, unedited CSV dataset.
- A Python script (
.pyor.ipynb) used to execute the extraction. - A 500-word document outlining your methodology.
Your methodology must describe your data-parsing approach, justify the sample stratification, and document each step along the way.
The spec
The brief poses the problem and leaves room for a response. The spec is the fixed part: what a marker checks. Some lines can be checked mechanically; the rest call for judgement.
- extraction of a minimum of 150 distinct SKUs from a nominated regional liquor retailer
- dataset must represent a stratified sample encompassing cask wine, ready-to-drink (RTD) premixes, and distilled clear spirits
- individual rows must strictly isolate product name, raw volume (mL), stated ABV, retail price, and the computed standard-drink yield
- the ABV-to-dollar index must correctly apply the 10g baseline metric established in Week 2
- the methodological documentation must allow for exact replication of your scraping parameters by a third party
How it is marked
| Criterion | Weight |
|---|---|
| Programmatic functionality and script efficiency | 40% |
| Data analysis comprehension | 30% |
| Methodological transparency and rate-limiting ethics | 30% |