What customers ask, and how they feel
Turning chat logs into decisions
Patterns across many chats matter more than any single transcript.
Sentiment breakdown
Positive, neutral, negative and urgent chats, tracked over time. You see how customers feel and when that starts to change.
Unresolved themes
The questions the bot keeps failing to answer. Treat this list as your backlog of help pages to write next.
Lead extraction
Buying signals in conversations become CRM contacts, so a person who asked about a plan is not forgotten after the chat.
Resolution rate
How much of the chat volume the bot handled on its own, and the hours of support work it saved your team.
Knowledge gap alerts
When the bot is unsure, the topic is flagged as a gap. You see which help page is missing before more customers ask.
Resolution by page and hour
See how often chats end solved, by topic, by the page they started on and by hour. Fix the weak spots first.
Why chatbot insights matter
Every chat holds a small piece of product research. Insights add those pieces up, so you see what customers ask, how they feel and what the bot could not answer.
The figures come from the same tickets and transcripts your staff already work in. So the numbers match what the support queue shows.
How customers feel
Each chat is marked positive, neutral, negative or urgent. This is a simple form of sentiment analysis, tracked over time.
A rise in negative or urgent chats is an early warning. You can act before small problems turn into lost customers.
What the bot could not answer
Unresolved themes list the questions the bot keeps failing. Each one points to a help page or a product detail you have not written yet.
When you add that page, the bot can answer from it next time, and the list gets shorter. See how the AI chatbot answers from your own content. Short on time? A web design and SEO agency can write the missing help pages for you.
Leads and results
Buying signals in chats become CRM contacts, and you can export them to your contact list. The resolution rate shows how much work the bot took off your team. The busiest chat hours show when it matters most.
Turning chatbot insights into work
Start with the unresolved themes, because each one is a help page you have not written yet. Write that page, and the AI chatbot can answer the same question from it the next time a customer asks.
Then look at resolution by page and by hour. A page where chats often end unsolved needs clearer copy, and the busiest hours tell you when your team should keep an eye on the queue.
Key takeaways
- Chatbot insights show what customers ask and how they feel.
- Open questions point to the help pages you should write.
- Buying signals become CRM contacts.
- Resolution rates show how much work the bot saves.
Frequently asked questions
Is this a separate analytics product?
No. It is part of the support module, so there is nothing extra to install.
Can I export the leads?
Yes. Leads go straight into the CRM contact list.
How far back does it look?
It looks across all the conversations you keep.
How hard is the setup?
There is no separate setup. Insights come with the support module and fill in as chats arrive.
What support is available?
Open a ticket in the Support Center and a person will reply.
Find out what you are missing
Your chat log already holds the roadmap, and chatbot insights help you read it.