Operational & Customer Analytics | Dashboards That Drive Decisions
Olist: Why Delivery Promises Run Long and What Would Fix It
Business question: Are Olist's long delivery estimates necessary or excessive, and what would actually make customers happier?
Dashboard design (use the grey boxes at the top to navigate): A guided analysis of ~100K orders across Brazil, plus an interactive tool where anyone can compare two states on six customer-experience scores.
Key Findings
-
Customers reward reliability, not speed. Late orders average 2.2★ versus 4.0–4.3★ for the rest, and arriving extra early barely moves ratings.
-
The estimates are long for a reason. A small share of deliveries take more than twice as long as usual, and the promise has to cover them. In the Northeast it doesn't, so 13% of orders there arrive late.
-
Sellers ship within about 3 days everywhere. The delay is the courier trip: 7.5 days to the Southeast vs. 19 to the North, where shipping also costs about twice as much per kg.
Why Does It Matter?
It turns a vague complaint ("deliveries take too long") into two specific actions: a quick fix (adjust Northeast delivery estimates) and a bigger one (more reliable couriers on northern routes), plus a tool teams can keep using to track progress.