Prompt
Why is our asset search not finding the right files?
Latest observation
Nov 3, 2026
OpenAI APIWeb search: off
There are a few common reasons asset search returns the “wrong” files or misses the right ones:
-
Metadata is incomplete or inconsistent
- Missing tags, titles, descriptions, creators, dates, or categories
- Different teams using different naming conventions
- Assets uploaded without proper metadata mapping
-
Search index is stale or not fully updated
- New/edited assets haven’t been reindexed yet
- Some folders/buckets aren’t included in the index
- Sync jobs are failing silently
-
The search algorithm is too literal or too broad
- Exact keyword matching misses synonyms or abbreviations
- Fuzzy matching pulls in loosely related files
- Ranking favors popular/old assets over the most relevant ones
-
File contents aren’t being extracted correctly
- PDFs, images, scans, or Office docs may not be OCR’d or parsed well
- Embedded text or attachments may be ignored
- Corrupted files or unsupported formats can’t be indexed
-
Permissions are affecting results
- Search may hide assets the user can’t access
- Shared collections and inherited permissions can create gaps
- Different users may see different result sets
-
Duplicates or similar versions confuse ranking
- Multiple versions of the same asset crowd out the intended one
- Old drafts and final versions are both returned
- Duplicate filenames reduce precision
-
Query formulation is mismatched to how assets are stored
- Searching by filename when the system indexes title/description
- Using plural/singular or brand/product variants
- Expecting full-text search when only metadata search is enabled
Fast ways to diagnose
- Search for a known exact filename to see if indexing is working
- Check whether a recently added asset appears after reindexing
- Compare results for two users with different permissions
- Inspect one “wrong” result and verify its metadata/ranking signals
- Review index logs for failed parsing/OCR/reindex jobs
Most likely root causes
If this started recently, it’s often:
- indexing delays/failures
- missing metadata
- permissions filtering
- OCR/text extraction problems
If you want, I can help you troubleshoot this systematically. Share:
- what search tool/platform you’re using,
- an example query,
- one expected file that’s missing,
- and one wrong result you’re getting.