Starbucks Recommendation Engine Data

What's inside

Four analysis-ready files for building a Starbucks personalized recommendation engine:

1. starbucks_menu.csv — 242 Starbucks US menu items with nutrition and pricing
2. fred_macro.csv — CPI, average hourly earnings, and real wage index (monthly, 2021-2026)
3. weather_daily.csv — Daily temperatures for 5 US cities (2024-2026)
4. synthetic_transactions.csv — 100,000 synthetic purchase transactions constrained by real data

Synthetic Data Transparency

The transaction data is synthetic — generated from probability distributions anchored to real-world constraints:

Every assumption is documented in the generation notebook. See Notebook 3 for the full generation pipeline and Notebook 6 for validation.

Use cases

• Content-based recommendation systems
• Product design optimization
• Price elasticity analysis
• Customer segmentation
• Seasonal demand modeling
• Synthetic data generation methodology

Data sources & licenses

Related notebooks (6-notebook series)