Customer segmentation
Find meaningful customer groups—and inspect the evidence
Use Datumry for customer segment comparisons and guided clustering, with clear methods, supporting records, and limitations for responsible decisions.
A repeatable path from data to action
Explore which customer groups behave differently without confusing correlation or model-generated clusters with proven causes.
Define the customer outcome
Select the customer measure, attributes, and time period connected to the decision you want to support.
Compare or discover groups
Rank known segments or run guided clustering to explore groups with similar numeric characteristics.
Validate before targeting
Inspect underlying records, sample size, stability, business meaning, and potential harmful segment effects.
What the workflow includes
- Known-segment performance comparisons
- Guided k-means clustering
- Supporting metrics and record counts
- Explicit cautions against causal overreach
Frequently asked questions
Can Datumry discover customer segments automatically?
Datumry supports exploratory clustering when the data has suitable numeric fields, as well as comparisons across existing segment labels.
Should model-generated clusters be used immediately?
No. Clusters are exploratory patterns. Review the underlying records and business meaning before naming or targeting a group.