Auto-handle dealership customer SMS support
A car dealership group built a Make flow analyzing customer SMS sentiment, auto-closing over 18,000 of 47,000 messages and flagging unhappy customers to Aftersales.
Done withMake
- What they did
- Stellantis&You UK, a UK car dealership group, built a two-way SMS support flow in Make. It connected Knack, ClickSend and OpenAI sentiment analysis. The flow read each customer message and closed those needing no action. It flagged genuine dissatisfaction, even when politely worded, to the Aftersales team. The source gives only this summary, not the build steps.
- How it went
- Over 12 months it analyzed more than 47,000 messages and auto-closed over 18,000 that needed no action. Make estimates this saved 151 hours of manual review.
- Worth knowing
- The guide gives only a short summary of this case. Its generic build advice says to include an escalation path for anything the AI is unsure about.
Try it yourself with Make
Build a Make scenario that receives incoming customer text messages from [SMS platform] and uses AI to classify sentiment and whether the message needs a human reply. Auto-close messages that are simple and positive, such as thanks or confirmations, and send anything negative or unclear to [team or person] with the message and customer details. Test it on [20] past messages and show me how each was classified before enabling any auto-close or live alerts.
Want help getting this working? Join the Vibe Coding Academy ↗ Run by Alex Finn, who also runs this site.
Source: make.com · Undated
Related use cases
Manage email, calendar and tasks from Slack
An agent built on Make.com with Gmail, Google Calendar and Notion triages emails, checks the calendar and adds tasks on Slack commands.
Summarize a YouTuber's latest video into a Google Doc daily
A Make scenario runs at 9 AM, finds a YouTuber's newest video, has Hugging Face summarize it, and appends the summary to a Google Doc.
Research, score and draft outreach for leads
A Make-based AI agent researches leads, scores them, and drafts personalized follow-up outreach automatically.
Sell services through an agent that takes payment and delivers
An agent runs a transaction flow, creates Stripe checkout links, verifies payment before delivering a service, and asks the owner on Telegram to approve refunds or disputes.
Custom agent
Research and draft cold outreach with Claude Code
Claude Code handles research and drafting for cold outreach while a separate platform sends at scale.
Build lead lists and multichannel outbound sequences
Grok Bot built a lead list from the author's website and ICP, then drafted a Salesforge sequence for review before launch; author reports a 38% reply rate.
Get 5 like this every morning
The best things people got an AI agent to do, each with the prompt to try it.