Route dealership SMS complaints and close simple replies

A UK dealership group used Make to sort incoming SMS, close routine messages, and send unhappy customers to Aftersales.

Done with Makeยท Intermediate

What they did
Stellantis&You UK set up a two-way SMS support flow in Make that linked Knack, ClickSend, and OpenAI sentiment analysis. Each customer text was checked to decide whether it needed action. Messages with no issue were closed automatically, while messages showing real frustration were passed to the Aftersales team, even if they were written politely.
How it went
Across 12 months, the system reviewed more than 47,000 texts and automatically closed over 18,000 that did not need follow-up. Make says that cut about 151 hours of manual review.
Tips
Leave room for a human handoff when the model is uncertain or when the customer sounds unhappy. It is safer to review a sample of past messages before turning on automatic closure.

Try it yourself with Make

Set up a Make scenario for [company] that watches incoming SMS messages from [SMS platform], checks each one for sentiment and whether it needs a human reply, and then labels it as "no action" or "send to [team/person]". If the message is a simple confirmation, thank-you, or other low-risk note, mark it closed automatically; if it sounds upset, confused, or unclear, forward it with the customer details and message text. Before anything is auto-closed, show me a test run on [number] past messages with the classification for each one, and stop for my approval before any message is sent, closed, deleted, or submitted.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on make.com โ†—

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.