Resource
Classify Support Tickets with AI
Billing, Technical, Account, How-To, Feature Request — ticket triage with definitions that decide the borderline cases for the model.
Categorize Customer Feedback with AI
Praise, Complaint, Feature Request, Bug Report, Question — multi-label feedback categorization where one message can carry three signals.
Overview
Customer feedback refuses to stay in one box: "love the app, but exports crash, and could you add CSV?" is Praise, Bug Report, and Feature Request in one message. This setup runs Multiple Labels mode, whose edge-case rules change accordingly: return every label that genuinely applies, strongest first — but a merely mentioned topic doesn't earn its label. The definitions split the classic confusions: Complaint is dissatisfaction without a requested remedy, Bug Report is something behaving incorrectly, Feature Request is a wish for new capability.
How to use this resource
-
One message per call
Multi-label works per message — the labels list comes back ordered by strength.
-
Trust the mention threshold
"A merely mentioned topic does not earn its label" keeps drive-by mentions from inflating your counts.
-
Watch the Complaint/Bug border
Dissatisfaction without a remedy is Complaint; something behaving incorrectly is Bug Report. The definitions decide, not the model's mood.
Why This Works
- Multi-label mode matches how feedback actually arrives — bundled
- Strongest-first ordering gives downstream consumers a primary signal anyway
- Exclusion-shaped definitions stop the same sentence from earning two labels
Best for
- Product teams processing feedback at volume
- Mixed-signal messages that single-label classification mangles
- Pipelines that route bugs and wishes to different boards
Not for
- Extracting the quotes and topics themselves — that's the Extraction Prompt Generator
- Overall sentiment scoring — that's a single-label problem with its own setup
Use cases
- Feeding a feedback board where one message can hit three categories
- Separating bug reports from complaints automatically
- Counting feature demand without losing it inside mixed messages
FAQ
How does the prompt keep a Complaint from being labeled a Bug Report?
It leans on exclusion-shaped definitions, not tone. Complaint is 'dissatisfaction that does not request a specific remedy,' while Bug Report is 'something behaving incorrectly or not working.' The Ambiguity Policy tells the model to re-read both definitions and pick the closer one when torn, so a frustrated 'this is terrible' stays Complaint unless the text describes broken behavior.
Why does a message get several labels back, and which one is the primary signal?
It runs in Multiple Labels mode, so 'love the app, but exports crash, add CSV?' returns Praise, Bug Report, and Feature Request together in the labels array. The Edge Case Rules order them strongest-first, so the first element is your primary signal — routing can take labels[0] while the rest stay available for secondary use.
When a review mentions a feature only in passing, will it still get the Feature Request label?
Not necessarily — the Edge Case Rules state 'a merely mentioned topic does not earn its label; it must be a real subject of the text.' A comment that name-drops CSV export while praising something else won't earn Feature Request. The model applies that threshold as a judgment call, so spot-check borderline mentions before trusting the labels array.
More resources from Data Classification Prompt
Resource
Text Classification Prompt — the Anatomy
The blocks a reliable classification prompt needs: defined labels, classification rules, edge-case rules, an ambiguity policy, and a confidence contract.
Resource
CRM Activity Classification with AI
Discovery, Demo, Negotiation, Follow-Up, Closed Won, Closed Lost — classify activity notes by sales stage so pipeline reports stop lying.
Resources that pair well
Resource
Extract Data From Text with AI
Free text in, named fields out. The extraction prompt pattern that turns any unstructured text into consistent, parseable records.
Resource
Force JSON Output from AI
Stop getting 'Sure, here is the JSON…' — the output-contract pattern that forces models to return only parseable JSON: schema, example, and a strict rule block.
Resource
Fix Invalid JSON from AI
The JSON won't parse and you can't see why. Deterministic cause-sniffing — trailing commas, single quotes, unclosed brackets — and the repair prompt that fixes it.
Related tools
Tool
Data Classification Prompt
Build classification prompts that assign labels from a closed set — with label definitions and edge-case rules.
Projects that use this resource
Project
Build an AI Support Agent with AI
The full path to a support agent you can put in front of customers — write its instructions, ground it in your docs, route and handle tickets, then evaluate and cost-control it before it goes live.
Project
Build a Customer Support System with AI
The full path to a support operation, not just a bot — stand up the knowledge base, route the tickets, add the AI agent, integrate your stack, close the feedback loop, evaluate, and deploy.
Workflows that use this resource
Workflow
AI Product Validation Workflow
Find out whether the thing you shipped actually worked — define the success metric, plan the measurement, classify the real evidence, then render a verdict and an iterate / pivot / scale decision.
Guides for this resource
Guide
Classify and Tag Data With AI Consistently
AI tags one batch pricing, billing, bug — then tags the same items cost, plan confusion, error next week, so every count measures a moving target. Here's how to classify and tag data consistently: fix a taxonomy, define every label, and hold the rules steady across runs.