AI Dashboards: Scoring Models
The Scoring Models page lets you configure scoring rules and multipliers used by Tango.ad to evaluate products, brands, and customer behavior.
By adjusting these settings, you can influence product scoring, personalize campaigns, and prioritize visitors based on their shopping behavior.
For an overview of AI configuration modules, see AI Dashboards Overview.
To open this page:
- Navigate to AI Dashboards.
- Click Configure on Scoring Models.
Overview
Scoring Models includes four configuration areas:
- Brand Booster — Increase the importance of specific brands.
- Category Booster — Apply score multipliers based on configured matching rules.
- Shopping Insights — Adjust scoring based on customer purchase history and cart value.
- Metadata Manager — Reserved for custom metadata configuration and not documented here because the current editor contains no usable configuration.
After changing the available scoring settings, click Save to save your configuration.
For related AI configuration options, see AI Dashboards: Advanced AI Settings and AI Dashboards: Traits Mapping.
Brand Booster
The Brand Booster lets you increase the AI score for products belonging to specific brands.
Each brand is assigned a multiplier.
Example
{ "Whyskas": 1.25 }
In this example:
- Products from the Whyskas brand receive a 1.25× score multiplier.
- The multiplier increases the influence of that brand on AI scoring.
When to Use
Use Brand Booster when you want to:
- Promote strategic brands.
- Prioritize premium product lines.
- Increase visibility for partner or sponsored brands.
Category Booster
The Category Booster applies score multipliers based on configured matching rules, such as page URLs or keywords.
The documented matching methods are:
Exact Match
Matches specific URLs.
Example:
{ "tags": [ "lipedem", "komprese" ], "match": [ "https://example.com/article" ] }
Regex Match
Uses regular expressions to match multiple URL patterns.
Example:
{ "tags": [ "lipedem", "komprese" ], "match": [ "(lipedem|lymfat|celulit|komprese)" ] }
Partial Match
Matches pages containing selected keywords or text fragments.
This option is useful when multiple URLs share common naming patterns.
When to Use
Category Booster can help you:
- Highlight seasonal collections.
- Promote selected product categories.
- Increase interest scores for educational content.
- Assign tags based on product or content topics.
For more information about how visitor interests can be used in analytics and personalization, see Analytics: Interests.
Shopping Insights
The Shopping Insights configuration adjusts AI scoring based on customer purchasing behavior.
It uses configurable multiplier buckets for:
- Revenue
- Number of completed orders
Revenue Buckets
Example:
{ "revenueBuckets": [ { "min": 0, "mult": 1 }, { "min": 500, "mult": 1.25 }, { "min": 1000, "mult": 1.45 }, { "min": 2500, "mult": 1.8 } ] }
In this example, visitors with higher lifetime spending receive larger score multipliers.
Order Count Buckets
Example:
{ "orderCountBuckets": [ { "min": 2, "mult": 1.08 }, { "min": 5, "mult": 1.18 } ] }
In this example, customers with more completed purchases receive additional weighting.
Benefits
Shopping Insights helps Tango.ad:
- Recognize loyal customers.
- Prioritize high-value shoppers.
- Personalize campaigns based on purchasing history.
- Incorporate customer purchase history into AI scoring.
For additional information about customer and purchase data, see Analytics Overview.
Saving Changes
After updating a scoring configuration:
- Review your JSON configuration.
- Click Save in the upper-right corner.
- Confirm that your changes have been saved.
Best Practices
- Use modest multipliers to avoid overly biasing AI recommendations.
- Reserve Brand Boosters for brands that align with your business priorities.
- Use Category Boosters to highlight important product groups or content.
- Configure Shopping Insights to reflect customer purchase history and lifetime value.
- Review scoring rules regularly to ensure they reflect current merchandising and marketing strategies.