Quick and easy marketing mix model. The model analyzes sales and marketing channel spend to indicate which channels or subchannels it deems to be most important.
- Generate a CSV with a column for dates, a column for sales, and a column for each marketing channel
- Populate the marketing channel columns with spend data
- Load the CSV and then click 'Analyze' to submit
- A chart shows up that compares channel feature importance vs. channel share of spend
URL
Details
- Python / FastAPI for endpoints (upload, results)
- Relies on Gradient Boosted Regression to do MMM modeling
- Redis for in-session data & caching
- Currently runs on AWS