When issues occur with 50-100+ rows, it's typically due to performance limitations or API rate limits.
For AI enrichments showing "Retrying," this often happens with lower API tier limits (like Anthropic tier 1) that require processing smaller batches of around 50 rows at a time.
Performance issues manifest as latency where cell data takes longer to load, especially with larger tables or complex workflows. Try a hard refresh: Mac (⌘ + Shift + R), Windows (Ctrl + Shift + R).
Best practice is always testing with 5-10 rows first, then scaling to 50, then 500 to avoid these issues.