Research question and scope
This protocol measures the share of open support records that have no named owner at a fixed observation time. The unit is one open record. An owner is one person who is accountable for the next action, not a team name or a shared inbox. The measure describes the ownership gap, not the workload of the people who do have records.
Set the scope to one queue, one client, and one observation window. Record the queue definition and the field used to store the owner. If the field changes during the window, keep the earlier definition for the full sample. A new assignment rule can change the share without changing the work.
Keep the observation separate from a staffing review. The purpose is to find where ownership disappears. A neutral description makes it easier for the team to accept the finding. State the observation time and the system used.
Record the queue's assignment rule at the time of the observation. A rule that assigns to a team rather than a person will produce a high unowned share. The rule is part of the context, not an excuse. State it in the report.
Record whether the queue has a default owner for new records. A default owner can hide a gap by assigning everything to one person. Check the load as well as the share. An owner who is assigned everything is not a real control.
Population and unit of analysis
Define the population as every record that was open at the observation time, including records that arrived before the window. The unit is the record, not the message. A record with several messages and no owner is one unowned record. A record with a team name in the owner field is also unowned under this definition.
Exclude records that are closed but awaiting an administrative step, and count them separately. Include records that are waiting on a customer if a named person is still responsible. A waiting record without an owner is more serious than a waiting record with one. Record the state and the owner together.
Record the intake source, the work type, and the date the record was created. These fields help explain where ownership is lost. A record created by a web form may enter without an owner, while a record created by phone may be assigned at intake. The difference is a process clue.
Include records that were created by an integration or an automated form. An integration may not set an owner even when a human queue would. Record the source and compare it with manual intake. The difference often points to a configuration fix.
Include records that are assigned to a person who is on leave or has changed roles. Record the availability separately. A name without availability is only partial coverage. The distinction matters for the next action.
Definitions that make the measure reproducible
Define an owner as a single person who can state the next action and the date. A team name, a role, or a shared inbox does not count. A record with two names counts as owned only if one is clearly primary. Publish two examples so a second reviewer can apply the rule the same way.
Decide how to treat a record that is assigned to a person on leave. It is owned, but the owner cannot act. Record a separate flag for availability so the reader can see the difference. A record with an unavailable owner may need a backup, not a new assignment.
Record the last ownership change and the person who made it. A record that lost its owner during a handoff is a different problem from one that never had an owner. The history helps the owner choose the fix. Capture it when the system allows.
Decide whether a record assigned to a person who has left the company counts as owned. It has a name but no available owner. Record it as a separate category. The distinction prevents a false sense of coverage.
Check whether the owner field is free text or a fixed list. A free-text field can contain a team name, a misspelling, or an old employee. A fixed list is easier to measure. Recommend a list if the field is free text.
Sampling and data collection
Draw the sample from the queue export at a fixed time, not from a list of escalations. Record the export time, the filters, and the row count. Include every owner group and work type so the result is not dominated by one lane. If the population is small, review every open record.
Capture the record identifier, state, owner field, intake source, work type, and age. Add the last message and the last assignment event. Store the export with the report so the observation can be repeated. A share without the records cannot be checked.
Check a small overlap sample with a second reviewer. If the two reviewers disagree on whether a record is owned, the definition is too vague. Rewrite the definition and retest before the full sample. A clear rule is more important than a large sample.
Record the last assignment event and the time. A record that was never assigned is different from one that lost its owner at a handoff. The history helps the owner choose the fix. Capture it when the system keeps an audit trail.
Record the time since the last ownership change. A record that has been unowned for a long time is a different risk from one that just arrived. Report the age distribution. The oldest records are the first to review.
Measuring the ownership gap
Report the share of open records with no named owner, then break it down by intake source, work type, and age. A gap concentrated in one intake source may be a form or routing problem. A gap that grows with age may be a handoff problem. Compare like with like and state the denominator.
Also report the share of records assigned to an unavailable owner. A record with a name but no available person is not truly covered. Show the two shares side by side. The owner can then see whether the problem is assignment or coverage.
Report the oldest unowned record and the reason it has no owner. One example is easier to discuss than a long list. If the same reason appears several times, name the pattern. A single case can point to a missing default owner.
Report the share of unowned records that are also past their lane threshold. An unowned record that is still new is less urgent than one that is already late. The combination shows where the risk is concentrated. Show the two measures together.
Limitations and next action
The record cannot show work that was done without an assignment, such as a quick reply from a shared inbox. It also cannot show a record that was owned in a private chat but not in the system. Treat the result as a measure of the system record. Note the channels that were not included.
Give the result to the queue owner and the intake owner together. Ask for one change, such as a default owner at intake or a required assignment before a record can leave the queue. Record the date and repeat the observation with the same definition. A small change can close a large gap.
Review the ownership field itself. If it allows a team name, the gap will return. A field that requires a person is easier to measure. A short validation rule at intake is often more effective than a weekly cleanup.
Ask the intake team which step is most likely to skip the owner field. A form that allows a blank field will be left blank under pressure. A required field or a default owner can close the gap. Test the fix on one intake channel first.
Check the owner field after the fix to confirm that it now requires a person. A rule that is written but not enforced by the system will drift back. Test a few new records from each intake source. Record the result and the date of the check so the next review starts from evidence rather than a promise.
