AI Knowledge & LogicAI Analytics & Insights

Data Collection

Data Collection allows you to automatically extract structured information from conversations that you can report on. Track common product issues, return reasons, customer types — or any other details that matter to your business.

You choose the data fields. The AI classifies every conversation for you.

What you can collect

You can capture almost any detail important to your business. For example:

Example Use CaseWhat the AI Extracts
Product issuesThe specific defect or complaint (e.g., not charging, broken part, poor fit)
Packaging issuesMentions of broken seals, damaged containers, missing parts, or leaks
Delivery problemsDelay, lost shipment, damaged on arrival
Order cancellationWhy the customer wants to cancel (e.g., changed mind, delivery taking too long, ordered by mistake)
Customer product interestWhat type of product the customer is looking for (e.g., protein powder, multivitamin)
Customer typeThe customer group defined by your business (e.g., male, female, younger, older, preferred language)

The possibilities aren’t limited to a list. Define the details that matter most to your team, and the AI will turn those conversations into structured data you can act on.


1. Define what to collect

Create and manage data fields in:

Settings → AI & Automation → Data Collection

Absolutely — here’s the table format with the improved wording for Prompt, exactly as requested:


Field settings include:

NameHow the field will appear in analytics.
PromptInstructions that tell the AI when to classify this data, and (optionally) how to choose between the possible values.
Possible valuesThe specific values the AI can assign — this can be detailed categories or a simple Yes / No.
Optional fieldWhen enabled, the AI only assigns a value if it finds relevant information in the conversation. (Recommended ON — many fields won’t apply to most conversations.)


2. How it works

After a conversation is closed/resolved or handed off, the AI:

  1. Reviews the full conversation
  2. Identifies which data fields should apply
    1. Required fields are always evaluated
    2. Optional fields are only applied when relevant
  3. Assigns the most appropriate value for each field
  4. Saves the results for reporting and filtering in analytics

3. Analytics

Insights are available in:

Metrics → Data Collection

You can:

  • See how often each value appears
  • Compare trends by channel, product, or time period
  • Spot rising issues early (e.g., “packaging leaks” increasing)
  • Filter conversations by any collected field for deeper analysis

This turns raw conversations into clear, structured data you can use to improve products, operations, and customer experience.


Examples

On this page