AI Dashboards: Advanced AI Settings
The Advanced AI Settings page allows you to fine-tune how Tango.ad evaluates visitor behavior. By assigning scores to customer actions and defining funnel stages, you can customize how the AI measures engagement and identifies visitors who are most likely to convert.
For an overview of the available AI configuration modules, see AI Dashboards Overview.
Overview
Advanced AI Settings helps you configure the intelligence behind Tango.ad's AI-powered customer insights. You can:
- Assign point values to customer actions.
- Define engagement stages based on accumulated scores.
- Prepare future AI model configurations.
- Access global AI configuration options (coming in a future update).
To open this page:
- Navigate to AI Dashboards.
- Click Configure on Advanced AI Settings.
Action Scores
The Action Scores section defines how much each visitor action contributes to a customer's engagement score.
Positive actions increase the score, while negative actions reduce it.
Available Actions
| Action | Description |
|---|---|
| hasCart | Visitor currently has items in their cart. |
| sign_up | Visitor signs up or creates an account. |
| purchase | Visitor completes a purchase. Usually assigned the highest score. |
| page_view | Visitor views a page. |
| view_item | Visitor views an individual product. |
| add_to_cart | Visitor adds a product to the shopping cart. |
| begin_checkout | Visitor starts the checkout process. |
| view_item_list | Visitor browses a product listing or category page. |
| remove_from_cart | Visitor removes an item from the cart. This is typically assigned a negative score. |
Example Configuration
{
"hasCart": 1,
"sign_up": 1.5,
"purchase": 4,
"page_view": 0.1,
"view_item": 1,
"add_to_cart": 2,
"begin_checkout": 3,
"view_item_list": 0.15,
"remove_from_cart": -0.8
}
How It Works
Every visitor action adds (or subtracts) points from the customer's overall engagement score.
For example:
- Viewing a product may add 1 point.
- Adding the product to the cart may add 2 more points.
- Completing checkout may add 4 additional points.
- Removing an item from the cart may subtract 0.8 points.
The total score is then used to determine the visitor's funnel stage.
Tip: Higher scores should generally be assigned to actions that indicate stronger buying intent.
📷 Screenshot: Insert the Action Scores page here.
Funnel Stages
The Funnel Stages section defines how engagement scores are translated into customer journey stages.
Each stage has:
- name – The label shown throughout the AI system.
- minScore – The minimum engagement score required to enter that stage.
Example Configuration
[
{
"name": "browsing",
"minScore": 0
},
{
"name": "interested",
"minScore": 4
},
{
"name": "hotLead",
"minScore": 10
}
]
Example
Using the example above:
| Score | Visitor Stage |
|---|---|
| 0–3.9 | Browsing |
| 4–9.9 | Interested |
| 10+ | Hot Lead |
These stages can be used by AI-powered campaigns to personalize messaging based on visitor engagement.
For more information about mapping customer events and attributes into AI-readable traits, see AI Dashboards: Traits Mapping.
ML Models
The ML Models section is reserved for future machine learning configuration.
Currently, the page displays:
Machine learning model configuration will be available in a future update.
No configuration is required at this time.
Configuration
The Configuration section is also reserved for future releases.
It will eventually contain global AI configuration settings for the workspace.
Currently, the page displays:
Global AI configuration settings will be available in a future update.
No configuration is available at this time.
Saving Changes
After modifying any available settings:
- Review your configuration.
- Click Save in the upper-right corner.
- The updated AI settings will be applied to the selected website.
Best Practices
- Assign higher scores to actions that strongly indicate purchase intent.
- Keep low-value browsing events weighted lightly to avoid inflating engagement scores.
- Use negative scores for actions that indicate decreasing purchase intent, such as removing items from the cart.
- Define funnel stages that clearly separate casual visitors from highly engaged shoppers.
- Review and adjust scoring periodically as customer behavior changes.