Scoring Dimensions Overview
Vantage's 12 baseline scoring dimensions and conditional Reviews & Trust Signals score
Last updated: 7th September 2026
Scoring Dimensions Overview
Vantage evaluates audited products across 12 baseline dimensions. Each dimension measures signals in the synced product data and contributes to a weighted readiness score. Reviews & Trust Signals contributes an additional dimension when supported review data is available.
These are product-data checks, not a measurement of external AI rankings or a guarantee of recommendations, visibility or sales. Your plan controls how many products are audited.
The baseline dimensions
| Dimension | Baseline weight | What it measures |
|---|---|---|
| Title Quality | 15% | Title length, specificity and formatting signals |
| Description Quality | 15% | Description length, structure and factual-detail signals |
| Attribute Completeness | 15% | Expected category attributes in product data |
| Product Identifiers | 12% | SKU, barcode and manufacturer identifier signals |
| Category & Taxonomy | 10% | Product categorization and taxonomy detail |
| Image Quality | 8% | Image count and alt-text coverage in synced data |
| Variant Structure | 5% | Variant options and consistency |
| Pricing & Availability | 5% | Pricing and inventory data signals |
| Policy & Compliance | 5% | Policy, warranty and compliance-related product-data signals |
| Structured Data | 4% | Structured attributes, explicit JSON-LD metafield signals, brand and category |
| Shipping & Returns | 3.3% | Shipping-weight, delivery, returns and available shop-policy signals |
| Product Highlights | 2.7% | Concise highlights in metafields, description bullets and tags |
The weights above total 100%. The final three baseline dimensions are summarized here; separate detail pages are not yet available. Structured-data checks use synced data and do not inspect rendered storefront markup. Baseline image scoring is separate from the optional visual image-analysis feature.
How the composite score works
Each baseline dimension produces a score from 0 to 100. Multiply each score by its baseline weight, add the results, then round to the nearest whole number:
Baseline = Title × 0.15 + Description × 0.15 + Attributes × 0.15
+ Identifiers × 0.12 + Category × 0.10 + Images × 0.08
+ Variants × 0.05 + Pricing × 0.05 + Policy × 0.05
+ StructuredData × 0.04 + ShippingReturns × 0.033
+ ProductHighlights × 0.027
Without applicable review data, the displayed composite is the rounded baseline. Title, description and attributes account for 45% of that baseline; improving their factual content can improve those dimension scores, without promising external shopping-service outcomes.
Reviews & Trust Signals
This conditional dimension applies when the product has supported review-app metafield data or processed review signals. It is omitted when neither is available; missing reviews do not add a zero-score dimension.
Supported raw review data without processed signals receives a neutral review score of 50. When processed signals exist, the review score reflects confidence-weighted sentiment and attribute coverage. This is separate from the Growth-or-higher review-intelligence generation feature.
When applicable, the review score has an additional weight of 0.07 and the combined score is normalized:
Composite = round((Baseline + ReviewScore × 0.07) / 1.07)
The review contribution can raise or lower the result relative to the baseline. Its weight is not added as seven percentage points to the displayed score.
Where to start
Open a product's score breakdown and prioritize missing or inaccurate source data. Review proposed content changes before approving them. Use the current weights in this overview when comparing dimensions, and check the app's visible audit results after changes.