Logistics — shipment tracking and on-time rate for a 3PL

15 minutes. Two CSVs. You play the ops manager at a third-party logistics company and find the hub that's killing your SLA.

Updated 2026-04-17

Industry: Logistics / 3PL Difficulty: Beginner Time: 15 min Plan: Free works

The scenario

You’re the ops manager at Meridian 3PL, a mid-size third-party logistics provider operating 8 hubs across the US. Your customer contracts all specify a 92% on-time rate. Last month you missed it, and your biggest customer flagged it.

You need to know where you’re missing and why, fast. Four questions:

  1. What’s our on-time rate overall? — month-to-date, trending up or down?
  2. Which hub is the problem? — ranked by on-time rate.
  3. Is it the hub or the carrier? — decompose by both.
  4. Can we alert on breaches proactively — before customer complaints?

Download the sample data

1,207 rows · 60 days
shipments.csv
Every shipment with origin, destination, carrier, promised date, actual delivery date, and on-time flag.
8 rows
hubs.csv
8 hubs with target on-time rate (92% across the board).
01

What's our on-time rate?

Overall rate for delivered shipments, last 30 days vs prior 30.
Prompt
Using shipments.csv (status='delivered' only), compute on-time rate for the last 30 days and the 30 days before that. Also show the daily trend — is it stable or sliding?

Expected: overall ~89% last 30 days vs ~91% prior 30. Trend chart shows a clear decline starting ~14 days ago.

Keep it: ＋ Save as Report — shared weekly with customers.

02

Which hub is the problem?

On-time rate per origin hub, this month vs last. SLA breach map.
Prompt
On-time rate by origin hub, last 30 days vs prior 30. Flag hubs below the 92% target. Sort worst-first.
meridian · hub performance
step 2 · expected answer
Tablize
✓ sql.query · on-time rate grouped by origin × period 184 ms
Atlanta is the problem. ATL is at 62% on-time rate this month, down from 82% last month — a 20-point drop. Every other hub is stable or up. If ATL were just at its own baseline, blended on-time would be above target.
HubLast 30dPrior 30dTarget
ATL · Atlanta62%82%92%
DEN · Denver85%86%92%
ORD · Chicago89%88%92%
JFK · New York89%89%92%
DFW · Dallas92%90%92%
SFO · San Francisco93%92%92%
LAX · Los Angeles95%94%92%
BOS · Boston92%91%92%
· Step 2 — ATL hub has deteriorated sharply
03

Is it the hub or the carrier?

Decompose the ATL problem by carrier to figure out where to push.
Prompt
For the ATL hub, show on-time rate by carrier over the last 30 days. Is one carrier worse than another, or is the whole hub struggling?

The Agent returns: “All four carriers are down at ATL, not a carrier-specific problem. Likely a warehouse issue — consider a site visit or an op audit.” (The generator injected the signal on the hub side, not the carrier, which the Agent surfaces correctly.)

Keep it: ＋ Save as Script — reusable when any hub trips the threshold.

04

Proactive SLA breach alerting

Daily Watch, per hub, per customer.
Prompt
Create a Watch that runs every morning. For each origin hub, if the rolling 7-day on-time rate drops below 90%, send me an alert with the hub, the current rate, and the number of late shipments in the last 7 days. Slack, please.

The Agent creates the Watch, runs it as a test once to verify, and confirms it’ll fire to Slack daily.

What you built in 15 minutes

  • 1 Report — weekly on-time rate, customer-shareable.
  • 1 Dashboard — hub-level live on-time map.
  • 1 Script — hub deep-dive, reusable.
  • 1 Watch — SLA breach alarm.

Your customers see the Report. Your ops team sees the Dashboard. You hear about problems from the Watch, not from the customer.

Next steps in this industry

  • Connect real data — carrier APIs (UPS / FedEx / DHL all have tracking APIs), your WMS, your TMS.
  • Add cost analysis — per-lane cost, per-shipment margin.
  • Add customer SLA segmentation — Premium customers at 95%, standard at 92%, economy at 85%.

Nearby tutorials

  • E-commerce — if you’re the merchant shipping, not the carrier.
  • Manufacturing — if you own the production, not just the delivery.