Intent Score estimates the likelihood a visitor will purchase, scored 0-100 based on behavioral signals like product views, add-to-cart actions, and checkout progress.
Intent Score = Weighted Model ( Product Views + Cart Actions + Checkout Progress + Engagement Signals )
| Metric | Definition |
|---|
| Product Views | Number and depth of product page visits |
| Cart Actions | Add-to-cart, cart views, and cart modifications |
| Checkout Progress | Steps completed in checkout flow |
| Engagement Signals | Time on product pages, scroll depth, interactions |
| Metadata | |
|---|
| Type | Number |
| Data Source | Upstack Pixel |
| Aggregation | Average |
Example
Sessions averaged an intent score of 34/100, with high-intent sessions converting at 12%.
| Intent Score Range | Sessions | Conversion Rate |
|---|
| 80-100 (High Intent) | 1,200 | 28% |
| 50-79 (Medium Intent) | 3,500 | 8% |
| 20-49 (Low Intent) | 4,000 | 2% |
| 0-19 (Browsing) | 3,300 | 0.3% |
How It Works
The Intent Score model analyzes behavioral patterns correlated with purchase completion. Visitors who view multiple products, add items to cart, and progress through checkout receive higher scores. The score updates in real-time as visitors take actions.
When to Use
| Scenario | Action |
|---|
| Retargeting segmentation | Target high-intent non-converters |
| Real-time personalization | Show urgency to high-intent visitors |
| Traffic source quality | Compare intent scores across channels |
| Checkout optimization | Analyze why high-intent visitors drop off |
| Metric | Relationship |
|---|
| Peak Intent Score | Highest intent reached |
| Session Score | Overall engagement composite |
| Avg. Funnel Stage | Funnel progress |
