# Ecommerce Search Performance > Keep onsite search fast and current with a dedicated search index, monitored updates and a safe recovery process. - Type: Merchant guide - Level: Intermediate - Reading time: About 6 minutes - Audience: Written for store owners, builders, and designers refining high-performance ecommerce stores. - Last updated: 2026-11-04 - Canonical page: https://academy.getzenbasket.com/merchant-questions/ecommerce-search-performance/ - Part of: [Merchant Questions](https://academy.getzenbasket.com/merchant-questions/) on ZenBasket Academy --- ## What you need to know ### Why is ecommerce search slow? Ecommerce search becomes slow when it scans the main catalog inefficiently, applies expensive filters or ranking, calls several services or runs on overloaded search capacity. Measure query, filter, and response-building time separately. ### Should product search query the main database directly? Usually not for substantial catalogs or search features. A dedicated search index is designed for text relevance, filtering and ranking, while the main database remains the authoritative product and inventory record. ### What is a search index? A search index is a prepared, searchable representation of product fields that helps the engine find and rank matches quickly. It is updated from catalog changes rather than replacing the source catalog. ### Why can search results become outdated? Search results become outdated when catalog events are delayed, lost or rejected, the indexer is behind, or an index switch fails. The search copy then no longer matches the authoritative product record. ### How quickly should product changes appear in search? Set a freshness target based on the change: stock and price usually need faster updates than a description edit. Measure the delay from accepted catalog change to searchable result and alert when it exceeds that target. ### How should search indexing failures be handled? Keep the last valid index available, retry temporary failures, isolate rejected records and support a controlled full rebuild. Monitor lag and compare sampled search documents with the authoritative catalog. ### A quick store check Change a test product's name, price and stock and measure how long each change takes to appear in search results. --- ## Frequently asked questions ### Should search read live stock from the search index? The index can support discovery, but final availability should be confirmed by the authoritative inventory system before purchase. ### Can reindexing interrupt search? It can if the process replaces live data unsafely. Build a new index, validate it, and switch traffic in a controlled way. --- ## Related learning path - **Ecommerce Database Performance** - **Ecommerce Background Jobs and Queues** - [ZenBasket Build Guide](https://help.getzenbasket.com/zenbasket-builder/)