04 / Business analytics & ML
Blinkit Sales Performance
Evaluating relationships between marketing inputs, time patterns and sales using regression.
Overview
I investigated how marketing-related features and timing patterns relate to sales, creating a regression workflow intended to inform business decisions from data.
How it works
I cleaned the dataset, performed exploratory analysis, engineered time-based features to capture seasonality, and trained a Random Forest Regressor. I used GridSearchCV for hyperparameter tuning.
Model performance was evaluated using MAE, RMSE and R², with results interpreted in terms of the project’s marketing and sales analysis goals.
What mattered
Features and validation strategy matter: a model can easily overstate its generalization if temporal patterns or leakage are handled poorly. Model outputs should remain tied to the dataset and evaluation conditions.
Technologies
Public links will be added when available. The case study above describes the completed work.