Best way to segment last-mile delivery customers by pain point?

K-means can work but honestly start simpler — run a factor analysis first to see which pain points cluster naturally, then use that to inform your segmentation. For last-mile specifically, I'd bet you'll find 3-4 clear groups around delivery speed, visibility/tracking, flexibility, and cost sensitivity. Once you validate those factors, THEN k-means (or even just cross-tabs) gets you actionable segments without overcomplicating it.