Claude
130 hours back on WhatsApp monthly
A laundry business took more than 200 WhatsApp messages a day. The team spent over 32 hours a week on routine replies. Builders first mapped two weeks of messages. About 80% fell into 12 clear groups, such as pickup times and order status. They built three agents in n8n with Claude. One sorts each message. One drafts and checks a reply against a knowledge base. One sends hard cases to staff in Slack, with the order history attached.
How it wentAfter 60 days the system gave back more than 130 hours a month. It handled 78% of queries with no human help. Reply time fell from 2 to 4 hours to under 3 minutes. The human queue fell from over 200 items a day to about 45. Cost stayed under $80 a month. The build cost was back in 6 weeks, per the author.
Try it · based on the source's workflow
Sort each new WhatsApp message for [business name] into one of these groups: [your message groups]. For a known group, draft a reply from [knowledge base] and the order record in [order system]. Check each fact in the draft against the knowledge base. If a fact does not match, or the message is a complaint, a legal point, or a refund over [amount], send it to [staff channel] with the order history and your draft. Show me that queue item before any reply goes to the customer.
A starting template based on the workflow the source describes. Review it before you use it; you are responsible for what it does.
130 hours back on WhatsApp monthly — from The SmallOffice Brief, a daily brief of real AI cases for small businesses: https://smalloffice.space/case-130-hours-back-on-whatsapp-monthly