Natural Language Audience Builder
Describe an audience in plain English and SignalSmith generates the filter conditions automatically.
Overview
The NL Audience Builder uses AI to translate natural language descriptions into audience filter conditions. Instead of manually configuring filters in the visual builder, you can describe the audience you want and let the AI generate the conditions.
How It Works
- Describe your audience — Type a natural language description
- AI generates filters — The system analyzes available traits and attributes, then generates filter conditions
- Review and refine — Review the generated conditions in the visual filter builder
- Save — Save the audience once you’re satisfied
Examples
| Natural Language Input | Generated Conditions |
|---|---|
| ”Customers who spent over $500 in the last 90 days” | lifetime_value > 500 AND last_purchase_date within 90 days |
| ”Users who signed up this month but haven’t purchased” | created_at within 30 days AND total_orders = 0 |
| ”High-value customers in the US or Canada” | lifetime_value > 1000 AND country in [“US”, “CA”] |
| “Churning customers who were active last quarter” | days_since_last_activity > 60 AND last_activity_date within 90-180 days |
How to Use
- Navigate to Segment > Audiences
- Click Create Audience
- Click the AI button in the filter builder toolbar
- Type your audience description in natural language
- Click Generate
- Review the generated filter conditions
- Modify if needed using the visual filter builder
- Click Save
Tips for Best Results
- Be specific about metrics — “spent over $500” works better than “high spenders”
- Include time frames — “in the last 90 days” helps the AI set appropriate date filters
- Reference known traits — If you know the trait name, include it for more accurate results
- Iterate — Start with a simple description and refine the generated conditions