CommerceLens
Customer Segmentation

Customer Segmentation

Rule-based, not ML

Every segment on this page is assigned by an explicit, documented rule applied to historical customer fields at dataset-generation time — not by an unsupervised clustering model. The rule for each scheme is stated in full below, so every downstream number can be traced back to a definition, not a black box.

Behavioral Segment — customer_segment

Based on each customer’s historical order count (previous_orders) at the time of the session.

Assignment rule
previous_orders == 0 → "New" previous_orders 1–2 → "Occasional" previous_orders 3–7 → "Regular" previous_orders 8+ → "Loyal"
Distribution of sessions by behavioral segment
Share of the 60,000-session dataset in each segment

Value Segment — customer_value_segment

Based on each customer’s total spend in the dataset window (customer_total_spend), by quartile.

Assignment rule
customer_total_spend ≥ 75th percentile → "High Value" customer_total_spend ≥ 40th percentile (< 75th) → "Mid Value" 0 < customer_total_spend < 40th percentile → "Low Value" customer_total_spend == 0 (no purchases) → "No Purchase Yet"
Distribution of sessions by value segment
Share of sessions whose customer falls in each spend tier

Segment Profile Comparison

How the behavioral segments differ on value and conversion metrics, computed directly from sessions in each segment.

Conversion rate by behavioral segment
P(Purchase | Segment) — an observational relationship, not a causal one
Full metric comparison
All value metrics side by side, by behavioral segment
SegmentSessionsConversion RateRevenueAOVMedian Order ValueReturn RateRepeat Rate
New6,62320.5%₹61,32,542₹4,523₹1,24512.2%0.0%
Occasional26,96626.4%₹3,35,18,987₹4,710₹1,34111.3%100.0%
Regular25,36130.1%₹3,61,45,015₹4,730₹1,37211.9%100.0%
Loyal1,05039.0%₹16,13,027₹3,944₹1,39513.7%100.0%
Loyal-segment sessions convert at a higher rate (39.0%) than New-segment sessions (20.5%) in this dataset. Because the segment itself is defined by past purchase history, this observational difference partly reflects existing tenure and trust rather than proof that any given intervention would move a New customer to Loyal-level conversion.

Want to slice further?

These segment definitions are fixed rules, but you can explore how they interact with age, income, device, acquisition channel, category, purchase status, and return status ad hoc in the Data Explorer.

Open Data Explorer