Amazon Customer Behavior Analysis
Exploratory Python analysis of Amazon purchasing behaviour, crossing demographic variables with browsing and decision patterns.
What it solves
- Crosses demographic variables with browsing habits to find where the purchase decision actually breaks down.
- Public Kaggle notebook, reproducible end to end.
- Organised around business questions rather than a catalogue of charts.
An analysis of Amazon customer behaviour based on demographic and platform-interaction data. The guiding question isn’t what people buy but when they decide and what makes them walk away.
What it examines
- Preferences by demographic segment: how categories of interest shift with age and profile.
- Browsing habits: how many searches and visits precede a purchase, and where the friction shows up.
- Decision process: which signals accompany a conversion and which precede an abandonment.
On the approach
The notebook is written to be read, not only executed. Every code block is preceded by the question it answers, and charts only appear when they show something the text can’t say better. An exploratory analysis that produces twenty charts and no conclusion isn’t analysis, it’s inventory.
Reproducibility
The notebook is published on Kaggle against the Amazon Consumer Behaviour Dataset, so anyone can fork it, run it, and disagree with the conclusions on the same data.