Make.com automation · 2026

AI Feedback
Sentiment Analyzer

A Make.com workflow that turns raw customer feedback into instant action. When a form is submitted, AI classifies the sentiment, routes positive reviews to social media, escalates negative ones to the support team, logs everything to Google Sheets, and sends personalized follow-up emails. A companion scenario runs every 2 hours to re-alert Slack on any negative ticket still open past 24 hours.

◆ Built in Make.com 2 scenarios OpenAI sentiment 3-way routing SLA escalation

Linked scenarios

Sentiment routes

24h

SLA watchdog

Main scenario — feedback intake

🔗

Webhook

↩️

Respond

🤖

OpenAI

📝

Set variable

🔀

Router

😊

Positive → Sheets → Gmail → Slack → Buffer

😐

Neutral → Sheets → Gmail → Slack

😠

Negative → Sheets → Gmail → Slack (urgent)

SLA watchdog — every 2 hours

⏰

Schedule 2h

📊

Sheets · Search Rows

❓

If-else (open > 24h)

🚨

Escalate → Slack "RESOLVE NOW"

💬

Else → Slack digest

How it works

Trigger

Feedback webhook

Customer submits the feedback form. A Make custom webhook receives the payload and immediately responds so the form never hangs.

AI Analysis

OpenAI sentiment

An HTTP call to /v1/chat/completions asks OpenAI to classify the feedback as Positive, Neutral, or Negative and return a short reasoning string.

Route

3-way router

A Tools "Set variable" normalizes the label, then Make's Router fans out into three deterministic branches — no manual triage required.

Positive path

Amplify to social

Row appended to Google Sheets, thank-you email sent via Gmail, team pinged on Slack, and a status update is queued to Buffer for social publishing.

Negative path

Escalate to support

Row logged with an "open" status, apology + owner assignment sent by Gmail, and an urgent Slack alert lands in the support channel so the team can act in minutes.

SLA watchdog

24h re-alert every 2 hours

A companion scenario runs on a 2-hour schedule, searches Sheets for negative tickets still open past 24 hours, and pushes a "RESOLVE NOW" Slack alert so nothing slips through the cracks.

Proof of build

◆

Main scenario — webhook → AI → router → 3 delivery branches

Webhook receives the feedback → responds immediately → HTTP call to OpenAI chat completions classifies sentiment → Set variable normalizes the label → Router splits into Positive, Neutral, and Negative branches → each branch logs to Google Sheets, sends a Gmail reply, and posts to Slack. The positive branch also queues a Buffer social update.

One webhook trigger — customers get instant confirmation, back-office runs asynchronously

OpenAI sentiment classification via a single HTTP module — swap models without touching the rest of the scenario

Router branches are deterministic (Positive / Neutral / Negative) — every submission ends up in exactly one path

Google Sheets is the source of truth so the SLA watchdog can reason over historical tickets

Make.com main sentiment analysis scenario

Main scenario: Webhook → OpenAI → Router → Positive / Neutral / Negative delivery branches

⏰

SLA watchdog scenario — every 2 hours

Scheduled scenario searches Google Sheets for negative feedback rows still marked "open". An If-else module checks whether the ticket has been open for more than 24 hours; if yes, it fires a "RESOLVE NOW" alert to Slack; otherwise it posts a lightweight status message so the team knows the queue was checked.

Runs on a fixed 2-hour cadence — predictable load, no polling storms

Single source of truth: the same Sheet the main scenario writes to

Escalation lives in one place — change the SLA window without touching the intake scenario

Watchdog scenario: Sheets → If-else → Slack escalation or Slack digest

AI FeedbackSentiment Analyzer

Want one like this for your support team?

AI Feedback
Sentiment Analyzer