Dataset | Count |
---|---|
Orders | 3421083 |
Products in Orders | 33819106 |
Unique Products | 49688 |
Aisles | 134 |
Departments | 21 |
Dataset | Count |
---|---|
Orders | 3421083 |
Products in Orders | 33819106 |
Unique Products | 49688 |
Aisles | 134 |
Departments | 21 |
Our dataset contains shopping data from 206,209 unique customers with 3,421,083 total orders. The average customer placed 16.6 orders during the observation period.
Can we break this down to learn more about customer behavior? To understand the who, we look at the when and what:
total_orders | avg_days_between_orders | total_items | unique_products | items_per_order | department_diversity |
---|---|---|---|---|---|
11 | 19.00000 | 70 | 19 | 6.363636 | 7 |
15 | 16.28571 | 226 | 121 | 15.066667 | 13 |
13 | 12.00000 | 88 | 33 | 7.333333 | 9 |
6 | 17.00000 | 18 | 17 | 3.600000 | 9 |
5 | 11.50000 | 46 | 28 | 9.200000 | 10 |
4 | 13.33333 | 14 | 12 | 4.666667 | 5 |
We engineered features that capture: - Ordering frequency and timing patterns - Product diversity and basket characteristics - Reordering behavior and loyalty indicators
[1] "Using these columns for clustering: total_orders, avg_days_between_orders, total_items, unique_products, department_diversity, aisle_diversity"
Based on these evaluation methods, we’ll proceed with 5 clusters for our customer segmentation.
Our analysis identified five distinct customer segments, each with unique shopping patterns. Power Shoppers place orders most frequently (4.5 times per month), while Basic Necessities customers shop just once monthly. Frequent Samplers maintain regular biweekly shopping patterns, whereas Occasional Bulk Buyers and Specialty Shoppers fall in between with 1-2 monthly orders.
Power Shoppers go shopping 4-5 times a month, the highest frequency of the user categories and buy from several departments, and they get the highest number of products.
Frequent Samplers order consistently every 10-14 days with medium-sized baskets of 8 items on average. They explore a range of categories (12 different ones) but show a strong preference for convenience items.
Occasional Bulk Buyers place infrequent orders (1.2 per month) but purchase large quantities when they do shop (15 items per order). They have a very long interval between orders at 25 days.
Basic Necessities customers place the fewest orders (just 1 per month) with the smallest basket size (5 items) among all segments. These customers exhibit the least product diversity, shopping in only 5 categories and having the longest intervals between orders (30 days).
Specialty Shoppers shop mainly from premium categories despite moderate ordering frequency (1.5 orders per month) and average basket size (6 items). What makes them unique is their high category diversity (15 categories) despite the relatively small basket size.
With these insights and future directions, Instacart can create more personalized shopping experiences that drive customer satisfaction and business value.
It would be a good idea to keep track of how segments evolve over time and if they change seasonally. This could power some predictive modeling that may help tailor incentives.