We chase the small variations and let the big ones sit
The easy ones close because they're easy; the ones that decide the account stay open.
What's happening?
Closing variations feels like progress, so the easy ones get closed: tidy items with clean paperwork and agreeable numbers. The three that will actually decide the account - complicated, contested, poorly papered - stay open, because each needs a week nobody has. The register looks healthier every month while the account's real risk doesn't move.
Triage is the fix: the open items ranked by value and risk rather than ease, the big ones given owners and deadlines, and each one's file - instruction, records, build-up - assembled while the evidence is still warm. A big variation left to the final account gets negotiated under time pressure against a party who knows you want to close; the same variation built now goes in as a position rather than a plea.
The solution
The account's deciders worked first, not last.