The Duplicate Companies Detection Agent is most useful when it's part of a routine, not a one-off cleanup. A few habits that get the most out of it:
1. Run it right after big imports or sync pushes
Bulk candidate imports and LinkedIn sync campaigns are the biggest source of new duplicate companies. Running a scan right after one of these events catches the mess while it's fresh, before duplicated records accumulate their own contacts and jobs.
2. Clear High confidence clusters first
High confidence merges (greater than 98% match) are safe to act on quickly — start with the global Merge all high confidence option to clear the bulk of duplicates in one pass, then slow down for Medium and Low bands.
3. Open the profile before merging Low confidence pairs
Low confidence pairs (90–92%) are more likely to be genuinely separate entities — regional offices, sister companies, or similarly named but unrelated firms. Click through to the company profile before merging these.
4. Use Dismiss for real non-duplicates, not to skip a decision
Dismissing tells Recruiterflow this exact pair isn't a duplicate, permanently. Reserve it for pairs you're confident are separate companies — if you're just not sure yet, leave the row unresolved and revisit it later instead.
5. Build it into a recurring admin routine
Since runs are manual and capped at once per 24 hours, pick a cadence that fits your import volume — weekly for high-volume sourcing teams, monthly for smaller or steadier accounts — and assign it to whichever admin owns data hygiene.