The Problem: Response Time Varies by Language
Running a small B2B trade team, the most frustrating moment comes when buyer inquiries arrive. English questions get answered quickly, but Chinese or Portuguese queries mean opening a translator and falling behind schedule. With only a couple of team members, it's easy to lose track of who should respond in which language.
「New Google Form responses come in, but every day we have to check what language it's in and reset our brains first」
It gets even messier when buyers mix languages. A sentence like 「Hello, 我想要30pcs...」 confuses even automated language detection.
Implementation: Let Claude Code "Decide" the Language First
We connected Google Form to Google Sheets on a Windows device, then set up a Claude Code agent with one simple job:
When a Google Form response arrives,
1. Detect the "primary language" from the inquiry text (English/Chinese/Spanish/Portuguese, etc.)
2. Record the detected language in a separate sheet column
3. Auto-recommend the corresponding language "response template row"
4. Send a notification email to the team ("Chinese inquiry - template ready")
Our pre-built response templates in Google Sheets look like this:
• 「English: Introduction / Inventory / Sample Orders」
• 「Chinese: Bulk Orders / Price Negotiation / Delivery Timeline」
• 「Spanish: First-time Buyers / Certifications / Payment Methods」
Claude Code doesn't just detect language - it briefly analyzes the buyer's actual question and recommends the closest matching template.
Results: Responses Arrive "Pre-prepared"
Now when we see a Google Form alert in the morning,
Before: 「New response (1) - text only」
After: 「New response (1) / Language: Chinese / Recommended template: Bulk Order Response / Link included」
The team member opens the template and just fills in the bracketed sections. Work that took 3 minutes now takes 45 seconds. No translator needed. No language confusion.
What's interesting is Claude Code sometimes catches details we missed. For a mixed sentence like 「Hello, 请问有现货吗」, it recognizes 「primary context is Chinese, but the greeting style is casual English-level」, then marks it as 「English/Chinese mixed」 and only assigns it to bilingual team members.
Small Insight
Automation doesn't always mean 「fully automatic.」 We automated only the 「preparation stage」 before decision-making. Having AI fully generate buyer responses was too risky - language nuance and business relationships matter. Instead, we eliminated the time cost of 「situation assessment」 before responding.
For any multi-language team, this approach should be more than sufficient.