Research question and scope

This protocol measures how often a closed item in a delegated queue is reopened. The unit is one closed record that returns to an open state for the same issue. A new request from the same customer is not a reopening unless it concerns the same record and outcome. The measure describes closure durability, not the quality of the original decision.

Set the scope to one queue, one work type, and a fixed observation window after closure. Record the closure definition and the reopening window, such as thirty days. If the window changes, the result is not comparable. A longer window will usually show more reopenings.

Keep the observation separate from a performance review. The purpose is to find process gaps, not to rank specialists. A neutral framing makes it easier for the team to report a reopening honestly. State the observation window in the report.

Record whether the customer or the team initiated the reopening. A customer reopening may indicate an unclear answer, while a team reopening may indicate a quality issue. The two causes need different follow-up. Separate them in the data.

Record whether the reopening happened within the same shift or after a handoff. A same-shift reopening may be a quick correction, while a later reopening may be a missed issue. The timing changes the interpretation. Report it with the share.

Population and unit of analysis

Define the population as every record closed during the window. The unit is the record, not the message, because one record can generate several follow-up messages. Count a record once even if it is reopened more than once, and note the number of repeat reopenings separately. A record with two reopenings is still one record.

Exclude records that were closed by the customer before the team acted, and count them separately. Include records that were reopened by the team for an internal correction. An internal reopening is still a closure that did not hold. Record who reopened the record and why.

Record the original closure date, the reopening date, and the issue. The time between the two is part of the measure. A reopening after one day is different from one after three weeks. Report the distribution, not only the share.

Include records that were closed by an automated rule as well as by a person. An automated closure can miss a condition that a person would have caught. Record the closure method. A rule-based closure needs its own review.

Include records that were reopened and closed again during the window. A second closure is part of the durability story. Count the record once but note the number of closures. A record that closes twice may need a different fix.

Definitions that make the measure reproducible

Define a reopening as a return to an open state for the same issue within the observation window. Define a new request as a different issue, even when it comes from the same customer. Apply the same rule to every record and publish two boundary examples. A second reviewer should reach the same classification.

Decide whether a record reopened for a missing attachment counts. If the attachment was part of the original closure, the closure did not hold. If the customer later sends a new document for a new purpose, it may be a new request. Write the choice into the method and apply it consistently.

Record the closure note and the reopening reason. A closure note that says done without a summary makes the cause hard to find. The reviewer should be able to see what the customer was told. A clear note is part of the evidence.

Write the reopening reason list before reading the sample. Use no more than six reasons and test them on ten records. A reason that does not fit should go to an other category with a short note. A consistent list makes the pattern visible.

Check whether the closure note names the customer's next step. A note that ends with done gives the next person nothing to verify. A clear next step is easier to test. Include note quality in a small subsample.

Sampling and data collection

Draw the sample from the queue export rather than a list of complaints. Record the export date, the filters, and the row count. Include records from every owner and every work type so the result is not dominated by one person. If the population is small, review every closed record.

Match the reopening to the original record using the record identifier, not the customer name. A name may be duplicated or misspelled. If the system does not link the two, use the customer and the issue with a manual check. Report the match rate and keep unmatched cases in an exception list.

Capture the closure note, the reopening message, the time between, and the person who handled each step. Add the customer's stated reason if one is available. Store the evidence with the record so the classification can be checked. A number without the records cannot be audited.

Record the time from closure to reopening in business hours and state the calendar. A reopening after two business days means something different from one after two weeks. The calendar should match the queue's operating hours. Publish it with the result.

Record the owner who closed the record and the owner who reopened it. A pattern across owners may reflect a shared rule rather than an individual habit. Report the pattern without naming people unless the team agrees. The process is the focus.

Measuring reopening frequency

Report the share of closed records reopened at least once, the share reopened more than once, and the median time to reopening. A high single-reopening share may be normal in a complex service, while a high repeat share suggests a closure process that is not resolving the issue. Show both. Report the count of records behind each share so the reader can judge the base.

Break the result down by work type, owner, and closure note type. A work type with a high share may need a clearer closing script. An owner with a high share may need a different review, but do not report names without the team's agreement. The pattern matters more than the individual.

Report the reasons for reopening using a short list, such as incomplete answer, wrong information, missing document, or customer changed their mind. A reason list that is too long will not be applied consistently. Test the list on ten records before the full sample. Keep the reason list stable across the comparison period.

Show the reopening share for records closed with a summary note and those closed without one. A summary may improve durability, but it may also reflect a more careful owner. The comparison is a useful starting point, not proof of cause. Report it with the counts.

Analysis, limitations, and next action

The record cannot show a customer who gave up instead of reopening. It also cannot show a reopening that happened in a private channel and was never logged. Treat the result as a lower bound on closure failure. Note the channels that were not included.

Give the result to the queue owner and the team with one change to test. The change might be a closing checklist, a required summary, or a second review for a high-risk work type. Record the effective date and repeat the measure with the same window. A controlled change is easier to evaluate.

Review the closure notes in the next sample and check whether the change improved the summary. If reopenings fell but the notes became longer without becoming clearer, the fix may be cosmetic. Look at both the rate and the note quality. A durable closure is the goal.

Set the next observation date before the review ends. A measure that is not repeated becomes a one-time story. Use the same window and definitions. Put the date in the operations calendar so it is not forgotten.

Record the next review date and the person who will run it. A measure that is repeated by a different analyst is a stronger test of the definition. Put the date in the operations calendar. The second observation should use the same closure and reopening rules, and it should note any process change that happened in between so the comparison stays honest.