Research question and scope
This protocol measures the share of delegated work items that miss a published same-day cutoff. The unit is one work item that was expected to finish before the cutoff. A miss means the item was not complete at the cutoff time, even if it finished later that day. The measure describes the promise and the queue, not the effort of the specialist.
Set the scope to one desk, one client promise, and one time zone. Record the cutoff, the work types covered, and the handoff point that defines completion. If the promise changes during the period, split the sample at the change date. A different promise creates a different denominator.
Keep the measure separate from a performance rating. A miss may come from a late request, an approval delay, or a capacity shortfall. The purpose is to find where the promise breaks. A neutral framing makes the team more willing to report the real cause.
Population and unit of analysis
Define the population as every work item that carried a same-day promise during the period. The unit is the item, not the message or the customer. One customer may send several messages about one item, and each message should not count as a separate miss. Use the record identifier to group them.
Include items that were completed after the cutoff and items that were still open at the end of the day. Exclude items that were cancelled by the requester before the cutoff. Report cancellations separately because they change the denominator. A cancelled item is not a late item.
Include items that were reassigned during the day. A reassignment does not remove the promise, and the item still counts. Record the owner at the cutoff and the owner at completion. The handoff may be the cause of the miss, so the history matters.
Definitions that make the measure reproducible
Define completion as the point when the result reaches the agreed handoff and meets the quality check. A draft that is waiting for review is not complete under this definition. Publish one example of a complete item and one example of an item that was finished but not accepted. The boundary should be the same for every reviewer.
Define the cutoff with a date, a time, and a time zone. Convert the cutoff into the desk's local time and the client's local time in the report. A time without a zone is not a usable definition. Keep the conversion rule in the method so the result can be repeated.
Define a late input as a request that arrived after the intake cutoff for the same-day window. Define a slow decision as an item that was ready but waited for an approval. Define a capacity miss as an item that had no available owner before the cutoff. Use these causes consistently.
Sampling and data collection
Draw the sample from the queue export at the end of each day. Record the export time, the filters, and the row count. If the volume is large, select every item in a fixed set of days rather than the easiest items. A sample of only completed items cannot measure a miss.
Capture the item identifier, the arrival time, the promised time, the completion time, the owner, the work type, and the blocking reason. Add the handoff time when the result left the desk. Store the export with the report. The timestamps are the core evidence.
Check a small sample against the client-facing record. If the desk marked an item complete but the client did not receive it, the definition may be wrong. Record the discrepancy and correct the method before the full sample. A measure that stops at the desk can overstate on-time performance.
Record the reason for each miss in the specialist's own words, then map it to one of the defined causes. Keep the original note for review. A coded cause without the note can hide a problem that does not fit the list. That note should stay short enough to read with the code.
Measuring cutoff misses
Report the share of items completed by the cutoff, the share completed later the same day, and the share still open at the end of the day. Show the counts behind each share. The three groups describe different levels of failure and should not be collapsed into one rate. A rate that mixes them will hide the boundary problem.
Break the result down by work type, owner group, and cause. A work type with a high miss rate may need a longer promise or a smaller scope. A cause concentrated in approvals points to a decision path rather than the queue. Compare only groups with similar volume and complexity.
Report the median and oldest delay for the missed items. A miss of ten minutes is different from a miss of two days. The delay distribution shows whether the problem is at the boundary or earlier in the day. Include the cutoff time in the report so the reader can interpret it.
Limitations and next action
The measure cannot show work that was promised verbally and never entered the queue. It also cannot show a client who did not check the result until the next day. Treat the result as a measure of the recorded promise. Note the channels and work types that were not included.
Give the result to the desk owner and the client contact with one change to test. The change might be an earlier intake cutoff, a faster approval, or a different promise for one work type. Record the effective date and repeat the measure with the same definitions. A controlled change is easier to evaluate than a broad push.
Review the boundary cases before changing the whole promise. A few items that arrived in the last ten minutes may explain a large share of the misses. Fix the intake rule first if that is the pattern. The goal is a promise the desk can keep, not a lower number by definition.
