Similar Products
The Similar Products feature allows the storefront to proactively suggest alternatives to a customer based on a hybrid similarity score. This score is calculated using product attributes (category, material, fit) and image similarity.
Core Logic & Scoring
- Hybrid Approach: SCAYLE Storefront API combines attribute matching with k-NN (k-nearest neighbors) vector search for images.
- Pre-defined Suggestions: Any products manually defined in the SCAYLE Panel as "suggested" will always appear at the beginning of the list, followed by calculated results.
- Availability Guardrails: Products that are sold out or inactive are automatically excluded from the slider, even if they were manually pre-defined.
.png)
Similar Products Recommendation on Product Detail Page
Frontend Integration Use Cases
By utilizing the where parameter in the useSimilarProducts composable, you can implement different "flavors" of similarity to match specific merchandising goals:
Default general discovery
Brand-loyal discovery
Cross-brand style discovery.
Narrow exploration within a specific category.
For more information on configuring and implementing recommendations in your Storefront Application, please check the SCAYLE Recommendation Step-by-Step Guide.
