ecommerce_ops · ecommerce · workflow
Whatnot enhances search with GPT-powered query expansion and spell correction
Misspelled search queries like 'jewlery' produced nearly empty results pages, causing users to falsely conclude Whatnot lacked relevant content. Acronym and abbreviation queries such as 'lv' or 'nyfw' also returned low result counts and lower downstream engagement.
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 · Search query data collection
Search queries are collected from backend logging, capturing query text, applied filters, and SERP tab.
Tools used
GPT
Outcome
The GPT-based query expansion approach reduced irrelevant content by more than 50% for misspelling and abbreviation queries compared to the previous method, while also yielding substantial improvements in query expansion accuracy and streamlining the generation and serving process.
Results
Volumemore than 50%
Source
https://medium.com/whatnot-engineering/enhancing-search-using-large-language-models-f9dcb988bdb9
Grounding & classification
Source type: technical build writeup
16 fields verified against source quotes.
content generationmetric backednamed customerproduction runtime claimedtools describedworkflow describedecommerceaccuracy improvementerror reductiontechnical build writeupecommerce opsdata sync enrichmentextract classify route