Insights · Early access
Every conversation is telling you something.
Insights turns your conversations and forms into answers: what customers ask, where your agent falls short, who might cancel, and what to fix first.
Get early access- Analytics across conversations and forms
- Direct feedback kept apart from AI inference
- Signals with a suggested next step
The answer to “what should we fix?” is already in your chats.
The same shipping question keeps coming back. A customer says “we might switch” in passing. A feature request shows up in a refund chat. Insights surfaces these so your team acts on them instead of scrolling past.
Know where answers fall short
See conversations, form activity, content, deployments, and model and tool usage. Find the questions your agent struggles with and fix the knowledge behind them.
Feedback you can trust
Visitor feedback and thumbs on individual messages show what customers told you. AI-inferred sentiment and satisfaction are labeled separately, so you never confuse the two.
Signals worth a second look
AI flags support needs, onboarding blockers, cancellation risk, and product feedback, each with a suggested next step for your team to review.
Read it. Review it. Act on it.
Check the numbers
Review conversation and form analytics to see volume, topics, and where your deployments are busiest.
Review feedback and Signals
Read what customers rated, then work through Signals and decide which deserve follow-up.
Fix the cause
Update knowledge, adjust agent instructions, or open a Case for the customer who needs a person.
Turn conversations into decisions.
Support data is product data, retention data, and onboarding data. Use it that way.
Improve agent answers
Spot weak answers and low ratings, then update the Library or the agent’s instructions.
Catch cancellation risk
A Signal flags the customer who mentioned switching, so someone reaches out before the renewal date.
Find onboarding blockers
See where new customers get stuck and fix the step, not just the one conversation.
Measure form activity
See how customers use your AI Forms alongside the conversations around them.
The full picture, with clear labels.
Insights separates what customers said from what AI inferred, and what AI suggests from what your team decides.
- Analytics
- Conversations, forms, content, deployments, models and tools, and usage in one place.
- Visitor and message feedback
- Ratings from visitors and thumbs on individual replies, shown as direct feedback.
- Contacts
- The people behind your conversations and form responses, so you see a customer’s history rather than isolated chats.
- Signals
- AI-detected moments like a support need, an onboarding blocker, cancellation risk, or product feedback, each with a suggested next step.
Questions about Insights
What does Insights measure?
Conversations, forms, content, deployments, model and tool usage, plus feedback from visitors. It shows where your agents help and where they fall short.
How is direct feedback different from AI-inferred sentiment?
Direct feedback is what a customer rated. Inferred sentiment and satisfaction are AI estimates from the conversation, and Insights labels them separately.
What is a Signal?
An AI-detected moment worth your attention, such as cancellation risk or an onboarding blocker, with a suggested next step. Your team reviews it and decides what to do.
Do Signals take action automatically?
No. Signals are for review and don’t run actions. Follow up yourself, or open a Case to track the work.
What are Contacts?
The people Help.ai has observed across your conversations and forms. A Contact brings one customer’s interactions together.
Find out what your customers already told you.
Get early access to Insights, Signals, and the rest of Help.ai’s support tools.
Product updates by email. Privacy Policy