Game score
0
Quality plus decisions
Silhouette
—
Higher is clearer
Davies–Bouldin
—
Lower is better
Inertia
—
Compactness cost
Outliers
0
Unusual customers
Best score
0
Saved locally
Experiment controls
Create customer behaviour, select features, and test clustering choices.
120
4
6
18%
6%
0.27
5
Mission
Complete the current learning objective.
Build useful clusters
Untimed
Generate customers, run clustering, and aim for a silhouette score above 0.45.
Mission not started.
Interactive customer map
Drag centroids to test assignments. Select customers to inspect their baskets.
Start by generating a dataset. Then compare algorithms or drag cluster centres.
Selected customer
Behaviour and basket evidence.
Select a point on the map.
Cluster profiles
Interpret discovered groups and assign business names.
Run clustering to create profiles.
Visual analytics laboratory
Inspect separation, compactness, product mixes, and customer behaviour from several views.
Campaign decision stage
Match each segment with a suitable marketing action.
Cluster customers before selecting campaigns.
Game performance
Balance model quality, interpretation, and business value.
0
Clustering quality0
Feature choices0
Campaign decisions0
Mission bonus0
Quality points appear after clustering.
Customer records
Search, inspect, and export fictional shopping behaviour.
| ID | Cluster | Spend | Frequency | Basket | Diversity | Discount | Top products |
|---|---|---|---|---|---|---|---|
| Generate a dataset to populate records. | |||||||
Learning guide
Understand why each choice changes the result.
Achievements and replay
Unlock badges by exploring more reliable segmentation strategies.