Research question and scope

This protocol measures how long a refund request waits for an approval after the evidence packet is complete. The unit is one refund case, identified by the order, the customer, and the date the packet reached the approver. Elapsed time is split into evidence collection, approver review, and customer notification so the owner can see which stage creates the delay. The study measures waiting time, not whether the refund decision was correct.

Set the scope to one client, one refund queue, and one approval tier for a fixed window. Record the policy version, the approver group, and the systems used when collection starts. If the tier or policy changes during the window, keep the earlier definition for the full sample or treat the change as a separate period. A policy edit can shorten apparent waiting without improving the work.

Population and unit of analysis

Define the population as every refund case that entered the queue during the window. The unit is the case, not the message, because one case can produce several customer messages and internal notes. A case that is reopened after a denial remains one unit and receives a second approval segment. Record both segments so the analysis does not hide the repeat.

Exclude cases withdrawn by the customer before the packet was complete, and record the exclusion count. Include cases denied at the first tier and sent to a second approver, because escalation is part of the waiting time. Include small goodwill refunds if they follow the same queue. Test the unit definition on five ordinary cases before collection begins.

Definitions that make the measure reproducible

Define packet completion as the moment every required field and attachment is present. Define approver review as the time from packet completion to the first decision, including requests for more evidence. Define notification as the time from the decision to the message the customer can see. A case that waits for a missing receipt is an evidence delay, not an approval delay.

Do not count the customer's own response time as queue time. Do not count a weekend or holiday unless the approver group was scheduled to work. Each definition should include one positive case and one boundary case so a second reviewer can apply it the same way. Publish the definitions with the sample so the numbers can be checked.

Evidence standard and data dictionary

For each case, capture the order identifier, amount, reason, required attachments, timestamps, approver, tier, and final decision. Add the message sent to the customer and the system events that show when the packet changed. Store the packet with the case record so the evidence survives a later review. A timestamp without the record cannot show what the approver actually saw.

Record whether the approver asked for more evidence and how long the specialist took to supply it. Record the approver's availability during the window, including leave and time-zone gaps. If the reason code is other, require a short note. A reviewer should be able to classify the case from the packet without interviewing the specialist.

Sampling and collection

Draw the sample from the refund system rather than a list of complaints. Record the export date, filters, row count, and selection rule. Include cases from every weekday, every approver, and every tier so the result is not dominated by one person. If the population is ordered by date, a fixed interval spreads the sample across the window.

Collect the packet as it existed at each timestamp, not as it looks after later edits. If the approver added a note after the decision, record when it was added. Do not reconstruct a missing field from the final message. Note the collection date separately so the reader can see how much of the record has aged.

Measuring elapsed time in stages

Report the median and the range for evidence collection, approver review, and notification. Use the same clock and time zone for every case, and state whether the clock includes weekends. A median alone hides the slow cases that frustrate customers, so show the slowest decile as well. The reader should be able to see both the typical wait and the worst wait.

Break the stages down by tier, approver, and reason. A tier that looks slow may be waiting for evidence rather than a decision. An approver who looks slow may be carrying an unreasonable share of cases. Keep the comparison within the same tier so a complex case is not compared with a simple one.

Analysis and reporting

Report the number of cases, the share that missed the target, and the median wait for each stage. Show the reasons for delay in a small table using the same categories as the definitions. Use the same denominator for every rate and state the exclusions. A percentage without its base is not actionable for the operations owner.

Compare the stages within the sample rather than against an external benchmark that may use different rules. Report the cases that never received a decision separately, because an open case has no final wait. If the sample is small, show counts instead of percentages. Do not present a difference as causation without a controlled comparison.

Limitations and sources of error

The record cannot show how often a customer abandoned the request before the packet was complete. It also cannot show the cost of a delay that the specialist absorbed by working around the queue. A slow approval may be invisible if the customer was told a different timeline. The study measures recorded waiting, not the full customer experience.

Packets can be edited, deleted, or split across systems with different clocks. A verbal approval may never appear in the system even though the work moved. Record those cases and their treatment rather than dropping them. If the approver and the specialist report to the same manager, note the independence limitation.

Review cadence and next action

Give the result to the approver group and the operations owner and ask for one observable change. The change might be a shorter packet, a named backup, or a rule that small refunds do not need a second tier. Record the owner, the effective date, and the measure that will show whether the change worked. Avoid changing the packet and the approval tier at the same time.

Repeat the protocol with the same definitions after the change and compare the same stages. If approval time fell but evidence time rose, the new field may be shifting work without reducing the wait. If only one approver improved, the change may not be the cause. Close the review when the owner accepts the finding and the next sample is scheduled.

Separating evidence delay from decision delay

Evidence delay and decision delay need different fixes. If most of the wait happens before the packet is complete, the queue needs clearer requirements or better source access. If most of the wait happens after completion, the approval tier or the approver schedule is the problem. Reporting one combined number hides which fix will help.

Add a field that records the first missing item for every incomplete packet. Count those items across the sample and rank them. A single missing document that appears in half the cases is a strong candidate for a template change. A decision delay that concentrates on one approver is a staffing question, not a training question.

Reporting to the client

Report the result in a form the client can act on: which stage, which tier, and which missing item caused the most waiting. Avoid presenting the numbers as a grade for the team. Use examples with identifying details removed. The client should leave with one change it can approve and one measure it can watch.

Show the denominator and the exclusions so the client can judge the sample. If the sample is too small to support a conclusion, say so. A cautious report is more useful than a confident one built on six cases. Offer to repeat the measurement after the change and name the date.

Replication and version control

Keep the definition, the policy version, and the export query with the report. A second analyst should be able to repeat the measurement on a later window and compare the same stages. If the definition changes, publish a new baseline rather than mixing periods. Replication is what turns a one-time review into an operating measure.

Schedule the next measurement before the first report is filed. The date should be far enough away for the change to take effect but close enough to keep attention. Assign the owner who will run it. A protocol that is never repeated becomes a story about the past. Put the date in the same calendar as the operations review so it is not forgotten.