ecommerce_ops · ecommerce · workflow
Supercharging Discovery in Search with LLMs at Instacart
Instacart's Related Items section returned irrelevant alternatives when exact matches were unavailable, and the search stack failed to surface complementary products that paired with a user's primary find, leaving follow-up purchase intent unaddressed.
How it works
Common implementation structure
How this type of workflow is generally built, generalized across documented cases — not tied to any one vendor's stack. Click any stage to read what happens there. Specific products that implement these stages appear in “Tools commonly seen” below.
Stage 1 · Batch query data preparation
A batch job extracts historical search queries from logs and enriches them with metadata including query understanding signals and consecutive search terms.
Tools used
LLMsQuery Understanding modelsWhole Page Ranker model
Outcome
LLM-powered discovery content delivered substantial improvements in user engagement and revenue, with the next-search-term extension producing an 18% improvement in engagement rate with inspirational content.
What failed first
An initial basic LLM generation approach misinterpreted brand queries and produced overly generic recommendations, such as suggesting raw protein foods when users consistently converted on protein bars and powders, resulting in poor engagement.
Results
Volume18%
Cost replacedsubstantial improvements
Grounding & classification
Source type: technical build writeup
22 fields verified against source quotes.
content generationpersonalizationrecommendation systemproduct catalogfailure mode describedmetric backednamed customerproduction runtime claimedsource backedtools describedworkflow describedecommercecustomer satisfactionrevenue increasethroughput increasetechnical build writeupecommerce opsdata sync enrichment