The Problem Emerged
Receiving 30+ buyer emails daily created a nagging frustration. I could only guess who responds quickly and who always delays. With limited inventory, I wanted to prioritize decisive buyers, but had no objective way to measure this.
As a Korean seller especially, I needed to distinguish delays caused by time zones from actual procrastination.
Before: Manual Work
Manually logged email receipt times in Excel
Tracked reply times by hand and did manual calculations
Retained buyer patterns only from memory
Made rough decisions like, 「This buyer typically takes 3 days to reply」
After: Automated Tracking System
Connected Gmail API from Windows device to Google Sheets and used Claude Code to automate the following:
Stage 1: Auto-Collect Email Timestamps
Gmail receive time → Auto-logged to Google Sheets
Sender address → Normalized (duplicates removed)
Subject line analysis → Auto-classified as "Order-related" vs. "General inquiry"
Stage 2: Calculate Response Time
Claude Code automatically computes time gaps between previous and current emails. We set "time from initial inquiry to first response" as the base KPI.
Stage 3: Auto-Generate Buyer Statistics
For each buyer, automatically calculate:
• Average response time (in hours)
• Fastest response on record
• Slowest response on record
• Monthly response success rate (reply ratio)
Stage 4: Smart Alerts
Buyers responding 48+ hours slower than usual are flagged as "Attention"
Buyers with less than 50% monthly response rate sorted into "Relationship Check" category
Buyers averaging 2-hour replies highlighted as "Priority Targets"
Real Results
Finding 1: Time Zone Misconception
European Buyer A surprisingly responded faster than expected. After accounting for business hours, we added auto-filtering by time zone.
Finding 2: Priority Reordering
"Large order buyers" and "fast responders" don't overlap. Small-volume buyers who decide quickly actually proved more reliable.
Finding 3: Auto Recontact Alerts
Automatically generated a list of buyers unresponsive for 6+ months to identify accounts needing re-engagement.
Tech Stack
• Google Sheets (database)
• Gmail API (auto email collection)
• Claude Code (statistics & classification logic)
• Simple AppsScript triggers (auto-refresh every morning)
What We Learned
Patterns invisible in manual work became crystal clear after automation. Notably, "response speed" isn't linked to order volume but strongly correlates with reliability. Now when inventory is tight, we use data to set priorities.