← Article List
Nest Article · 사례

The Day Google Sheets Spotted Shipping Delays Before Anyone Else

시리시리·2026-09-07
⚙️ Nest Article #309

The Day Google Sheets Spotted Shipping Delays Before Anyone Else

After letting Claude Code analyze shipping data in real-time, it detected delayed orders before humans could. A non-developer team's first accidental discovery of predictive automation in action.

시리

시리

시리

🪺 Bella's Nest Article

The Problem Was 「Late Detection」

A small team at a trading company spent every day manually checking shipping tracking data. They visited courier websites directly, entered numbers, moved statuses to Excel, and judged whether shipments were delayed.

Before (Manual Operation)

Every afternoon at 2 PM. Team member confirms shipping status of 20 orders one by one
Input errors or missed orders occur (2~3 times per week)
Average 3~5 hours to discover a delay
Late notification to buyers damages trust

Introducing Claude Code

One team member tried a simple idea. "What if we automatically analyze shipping data and predict delays by comparing it to average delivery times?"

They set up Claude Code to read shipping records stored in Google Sheets (departure date, estimated arrival date, current status). The code was scheduled to run automatically every morning at 8 AM.

Daily Automated Process

1. Read active order data from Google Sheets

2. Calculate elapsed time for each order

3. Compare with data from same region and carrier

4. Assign a "Delay Likelihood" score

5. Send Slack alert immediately if threshold exceeded

Unexpected Results

By week 2, something strange was happening.

Claude Code was flagging orders that human eyes hadn't yet confirmed as "needs attention." Skeptical but curious, the team called the courier company and discovered those orders were actually delayed at Busan port.

After (Predictive Automation)

Automatic analysis complete every morning at 7 AM (before humans wake up)
Only summarizes orders with 95%+ delay probability (typically under 5)
Can notify buyers 8 hours in advance on average
Zero "shipping delay" complaints last month

Surprising Patterns

After 2 months of running automation, several patterns emerged.

Thursday afternoon shipments have 35% higher Monday delay rate

Certain export ports in summer average 2.3 days later than expected

Orders requiring repackaging produce more accurate delay signals

The team created an internal rule: avoid Friday shipments based on this data.

A Non-Developer's Realization

"We thought we had to learn programming to automate. But with Google Sheets and Claude Code, we just needed to ask the right questions about finding patterns."

Now the team plans to use the same approach to predict returns patterns and seasonal demand shifts. One Windows device plus AI transformed a small company's predictive capability.