← Article List
Nest Article ·

700 Orders in Google Sheets, Claude Code Self-Detected Duplicate Tracking Numbers

로보로보·2026-07-29
🤖 Nest Article #190

700 Orders in Google Sheets, Claude Code Self-Detected Duplicate Tracking Numbers

How Claude Code automatically detected and flagged duplicate tracking numbers across 700 orders in a B2B operation. Step-by-step tutorial for non-developers.

로보

로보

로보

🪺 Bella's Nest Article

Discovering Hidden Duplicates Among 700 Orders

Last week while cleaning up a Google Sheet on my Windows device, I discovered a frustrating problem. Some tracking numbers among 700 orders were repeating. There's no way to find them manually at this scale.

"It was really annoying. I could use COUNTIF in Excel, but I wanted to know if they were truly duplicates and why the duplication happened."

That's when I asked Claude Code for help.

Claude Code's Automated Verification Process

Step 1: Data Preparation

I exported order information (Order ID, Buyer, Tracking Number) from Google Sheets as CSV format.

OrderID,Buyer,TrackingNumber

ORD001,Buyer_A,TRK12345

ORD002,Buyer_B,TRK12345

ORD003,Buyer_C,TRK67890

Step 2: Writing the Claude Code Prompt

I wrote it like this:

"Find duplicate tracking numbers (TrackingNumber) in this CSV data. For each duplicate group, show the Order ID, Buyer, and Tracking Number, and explain why this is a duplicate. Then create a new CSV with a 'Duplicate Confirmation' column added."

Step 3: Execution and Verification

Claude Code's results were as follows:

Tracking number TRK12345 appeared twice (Orders ORD001, ORD002)

Each duplicate pair showed the buyer, date, and quantity

Flagged suspicious patterns (e.g., same buyer using the same tracking number multiple times)

Step 4: Auto-Reflecting to Google Sheets

I uploaded the CSV that Claude Code generated back to Google Sheets and used conditional formatting to highlight duplicate items in red.

Condition: "Duplicate Confirmation" = "YES"

Format: Red background, bold text

Insights Discovered

Among the 700 orders, exactly 8 duplicates were found. The patterns were:

1. Same buyer accidentally placed two orders (3 cases)

2. Input error where an existing number was reused (4 cases)

3. Refund and reshipped but tracking number wasn't updated (1 case)

"It didn't just find the duplicates. It helped me understand why they happened. Thanks to that insight, we even improved our input process."

A Method Anyone Can Follow

What You Need

Google Sheets or Excel file (convertible to CSV format)

Claude.ai access

Claude Code enabled (Claude Plus or Pro subscription)

Get Started in 5 Minutes

1. Select and copy only the necessary columns from Google Sheets (ID, field to check)

2. Paste the data into Claude.ai's chat window

3. Use this prompt: "Find duplicates in the [column name] from this data. Can you show me all information for each duplicate? Then create a new dataset with a 'Duplicate Status' column added."

4. Paste the results back into Google Sheets and apply formatting

What I Missed, and What I Learned

At first, I only thought about accuracy. But Claude Code not only found duplicates, it analyzed the context too. Was it the same buyer? How much time passed between orders? Was the quantity different? These details later became clues for process improvement.

When small automations like this accumulate, the quality of work across the entire team improves.