Running twenty files at once without losing one
Capacity is not about working faster. It is about how much of the tracking lives in a system rather than in your head, and where that ceiling actually sits.
Every coordinator has a number, the point past which files start slipping. What is interesting is that the number has very little to do with how hard anyone is working, and almost everything to do with where the tracking lives.
The ceiling is a memory problem
A coordinator running files from a spreadsheet plus recall is doing something cognitively expensive: holding the state of every file, and knowing which of roughly a hundred and fifty live obligations is closest to hurting someone.
That works, and it works well, right up to a threshold that is usually somewhere between ten and fifteen active files. Then it does not fail gradually. It fails suddenly, because the failure mode is not slowness, it is a dropped item, and one dropped item is indistinguishable from excellent performance until the day it surfaces.
With the recall moved into a system that computes obligations from the contracts, the same person commonly runs twenty-five or more. Nothing about their skill changed. What changed is that being right no longer depends on remembering.
Sort by obligation, not by file
The instinct is to organize the day around transactions: work the Elm Street file, then the Oak Avenue file. It feels orderly and it is the wrong shape, because deadlines from different files compete for the same hour and a per-file view hides that competition.
The working view is a single list of obligations across every file, ordered by consequence, answering one question: what happens today, and what becomes impossible if it does not.
- Everything due today, across all files
- Everything at risk this week, meaning the lead time no longer fits the remaining runway
- Everything blocked, waiting on a third party, with the age of each wait visible
- Everything ratified in the last forty-eight hours and not yet fully extracted
That fourth bucket is the one people leave out, and it is the most expensive omission. A new ratification that has not had its dates pulled is not a quiet file. It is an unexploded one.
A file you have not extracted is not a file with no deadlines. It is a file whose deadlines are running without you.
Rank by irreversibility
When two things are due the same day, date order is the wrong tiebreak. Consequence order is right, and the useful question is what is actually lost if this slips twenty-four hours.
An expiring inspection or financing contingency is near the top, because letting it lapse can forfeit a right or the earnest money and cannot be undone by apologizing. A settlement-week task that everyone would happily move sits lower even if it is due sooner. A courtesy update to a party sits lower still.
Most coordinators do this instinctively on their best days and lose it on their worst, which is exactly when it matters. Writing it down as a rule, rather than relying on judgment under pressure, is most of the value.
Batch by counterparty
Chasing is the bulk of the middle of a file, and chasing is far more efficient batched by who you are chasing than by which file you are chasing for.
One call to a lender about four files beats four calls about one file each, and it is better for the relationship. Same for title, same for the association management companies, same for the cooperating agents you work with repeatedly. The person on the other end also has a queue, and being the one who batches makes you the one they answer first.
Make waiting visible
The single most useful column in a pipeline view is how long something has been waiting. Not whether it is outstanding, everything is always outstanding. How long.
A resale package ordered nine days ago and a resale package ordered yesterday look identical on a checklist and are entirely different situations. Without elapsed time on screen, escalation becomes a feeling rather than a schedule, and escalation-by-feeling is what produces both the nagging and the neglect.
What to do when you are already over the line
If files are slipping now, adding effort will not fix it, because effort is what is already failing. Three moves, in order:
- Re-extract every active file. Not the ones you are worried about. All of them. The dangerous file is the quiet one nobody has looked at in two weeks.
- Move the tracking out of your head. Any system that computes dates from the contract and recalculates when one moves. The specific tool matters much less than the property.
- Then, and only then, decide whether you need more people. Most teams that hire to fix slippage discover the new person inherits the same untracked state and slips too.
Capacity follows from the system. Hiring against a broken system just buys you two people guessing instead of one.