merchandising-rules

Shopify's built-in collection sorting supports Best Selling, Newest, Price, and Manual drag-and-drop ordering. For performance-based automated ranking, install the free Shopify Search & Discovery app for boosts, or Kimonix for weighted scoring based on sales velocity, conversion rate, and inventory.

Combine normalized metrics like sales velocity, revenue, conversion rate, margin, and recency using configurable weights, then sort products by the weighted sum. Use min-max normalization across the product set and demote out-of-stock products so they rank lower without being excluded.

Algolia provides a built-in Rules and Pinning dashboard for merchandising without custom code, plus real-time custom ranking on attributes like sales_30d. Elasticsearch requires building function_score queries manually with weight multipliers, offering more control but more implementation effort.

Out-of-stock products should be buried, not excluded, so they remain visible for SEO and wishlists while ranking lower. Only completely remove discontinued products from collections and search results.

All-time sales totals let established products accumulate unassailable scores. Use a rolling 30-day sales window, add a recency weight for products added in the last 14 days, and cap extreme outliers at the 95th percentile to give new products exposure.

Keep manual pins to five or fewer products per collection. Pinning too many products prevents the ranking algorithm from surfacing other items and defeats the purpose of automated performance-based sorting.