How does Cobalt determine which record is the best match?
What’s a practical NY “waterfall” we can implement on our side?
What “waterfall” name-matching strategy works well for tight state searches (e.g., NY)?
How should we handle generic names?
If we have an entity suffix (e.g., “LLC”), should we include it in the query?
How does string length affect confidence scores?
Why did I see confidenceLevel = 1.0 when the names didn’t match?
How does the API handle messy or inconsistent entity names?
Do you treat “&” the same as “and”?
Why did shortening the name to a core form help?
How should we combine confidenceLevel with address signals?
We got a low-confidence result that shared the ZIP but not the name—why?
Fields (name/score/URL)? Can we get more?
Can we see more than just the top match when several are close? How should we use the alternatives list?
Will Cobalt’s API detect entity name changes (e.g., Pizza Inc → Pizza LLC)?
How does the API handle name disambiguation (e.g., punctuation, suffix differences)?
Do you provide DBA (fictitious name) information?
Can Cobalt return multiple possible matches with confidence scores?
How does Cobalt’s confidence score affect underwriting?
What confidence thresholds should we use?
Close