AutoMerging Duplicates
You can select records from multiple sets and AutoMerge them all at once.
Everything on this page happens directly in your CRM's list views, one page of duplicate sets at a time. For a bulk operation with no page-by-page ceiling, submitted from the Management App instead, see AutoMerge requests in the administration guide.
You may select any (or all) records from any (or all) sets on the page and AutoMerge every set with a single click. The behavior is identical no matter which records you select from each duplicate set.
The record with a Rank of 1 is the winner-designate. It remains active but disappears from this list view of duplicates after you AutoMerge the set.
Merging in bulk
Once you understand how matching and ranking are being performed against your requirements, you can start to AutoMerge large numbers of duplicates quickly.
- Select the whole page of duplicate sets.
- Click AutoMerge.
The AutoMerge process is synchronous. Be prepared to wait several seconds for it to complete, especially when merging a full page of sets.
You can open the duplicates list in more than one browser tab or window and run an AutoMerge in each at the same time, which speeds through a large backlog faster than waiting for one page to finish before starting the next. If you lose track of which tab is on which page, there's no harm: running AutoMerge again against a set that's already been merged simply does nothing.
How many sets at once
A single manual AutoMerge operation can target all the visible records in a list view which will represent the records of perhaps 100s of duplicate sets. In practice you'll select a smaller more manageable number at a time, well under that ceiling.
If you need to merge more than a page's worth without clicking through them, that's what an AutoMerge request from the Management App is for.
The order merges run in
Clicking AutoMerge on a set fires a single AutoMerge operation, but behind the scenes that operation can comprise several sub-merges, one per losing record, run in ascending rank order.
For a set of 3 records, that's 2 sub-merges: Rank 2 into Rank 1, then Rank 3 into Rank 1. For a set of 4, it's 3 sub-merges: Rank 2 into Rank 1, then Rank 3 into Rank 1, then Rank 4 into Rank 1. The pattern continues for larger sets, always ascending, until every loser has been folded in.
That sequence is why field preservation resolves the way it does: the first loser's values reach the winner before the second loser's are considered. Whichever losing record ranks best gets first claim on any empty field.
What happens to the losing records
Losing records (Rank ≥ 2) are deactivated, not deleted. Any un-populated field on the winner is filled in from the losers in increasing rank order. See Viewing AutoMerged history to find them afterward.