Where the other four days went
On Monday morning an analyst is asked to reconcile two files and identify the discrepancies that require a decision. The files are ready, the matching rules are known, and the analyst can do the work in a day. Give the task until Friday and it can still take the whole week. The analyst takes on other requests, the draft waits for a reviewer, and a discussion about how to present the findings produces another round of formatting. Friday arrives and the manager gets the same answer.
C. Northcote Parkinson captured that tendency in a 1955 essay in The Economist. As he put it, “work expands so as to fill the time available for its completion.” Timing the work is how you establish whether that happened. A week can also reflect a missing file, a reviewer with another priority, or a mismatch that genuinely needs investigation. Those facts call for different responses.
Measure the work and the wait
Take one recently completed task and write down the time it came in, when someone started it, when it was ready for review, when the review took place, and when the recipient accepted it. Ask the people involved for a rough account of time spent doing the work. The difference between elapsed time and hands-on time shows where to look. The National Institute of Standards and Technology describes value stream mapping as a way to see how work and information move and find opportunities to reduce lead time. You can borrow that approach with a few timestamps and no new software.
A waiting period has a cause. If the manager only reviews work on Thursdays, reserve a review slot when assigning the task. If the analyst manages four priorities at once, agree which can wait while this job is done. If the recipient changes the matching rules halfway through, agree on the new scope and update the estimate. A believable deadline includes these dependencies. It also makes clear which parts of the schedule people can change.
Try one day with a clear endpoint
For the reconciliation, agree by 9 a.m. on the files, the matching rules, and what the decision maker needs. Ask for the checked totals, an exception list, and the few questions that need a decision. Reserve a late-afternoon review. Set one early checkpoint to surface an input problem or a mismatch that could change the scope. Put optional formatting improvements on a separate list so the analyst can finish when the answer is ready.
Leave room for real uncertainty. If the analyst finds that the files count returns differently, the team may need the data owners to agree on a definition before the numbers can be signed off. The analyst should report what is known, what is blocked, and when an answer is realistic. The decision maker can then clear the dependency or revise the delivery date. Keep the required accuracy checks in place.
Protect the time you recover
On the next few comparable tasks, record elapsed time, rework, and whether the recipient could act on the result. If the task arrives sooner and holds up at review, keep what worked. If defects or repeated overtime increase, revisit the scope, staffing, or review window. If the work still takes a week, pull the timestamps and find the longest wait.
When the result meets the agreed standard, close the task. Give the remaining time to the next worthwhile job. The gain is the ability to make another decision sooner with the same people and the same quality of work.