Udit P. Clay doesn’t “sort” companies by default, but when you import large lists (like 25k), the enrichment jobs usually run top-down in batches. That’s why you see more hits in the first chunk and fewer later if the data source rate-limits or times out.
Fix: add a “Randomize Rows” step after import, then apply your job filter. This evens it out across the whole list instead of front-loading the matches.