Research question and scope

This protocol measures how often a replacement order is created without a recorded reason. The unit is one replacement order linked to one original order. A missing reason means the order record does not state the failure, the customer evidence, or the approval that justified the shipment. The measure describes the evidence around a remedy, not the honesty of the customer.

Set the scope to one store, one order system, and a fixed period. Record the replacement policy, the value limit, and the approval roles in force. If the policy changes during the period, split the sample at the effective date. A different limit creates a different population.

Keep the measure separate from a return or refund rate. A replacement can be appropriate even when returns are low. The purpose is to test whether the order record explains the decision. Report the three remedy rates separately.

Population and unit of analysis

Define the population as every replacement order created in the period. Include orders created by the desk, by a supervisor, and by an automated rule. The unit is the replacement order, not the support ticket. One ticket may create two replacement orders, and each should be judged separately.

Include replacements that were later cancelled before shipment, and mark the cancellation reason. Exclude exchanges that were processed as a return and a new purchase. The two paths have different records. Keep the boundary visible in the report.

Exclude replacements for internal stock adjustments that never touched a customer. The measure is about customer-facing remedies. State the exclusion count so the reader can see the base. A stock correction is a different record and should not dilute the rate.

Definitions that make the measure reproducible

Define a recorded reason as a stated failure category plus the evidence used. The category may be damaged, wrong item, missing, or not as described. The evidence may be a photo, a carrier record, or a customer statement. Publish one example of a complete and an incomplete reason.

Define an approval as a named person and a time, not a status label. A note that says approved by ops is incomplete unless the person can be identified. A replacement within the desk limit still needs the desk's name. The rule should match the policy in force.

Decide how to treat a reason recorded in a chat but not in the order. Classify it as missing from the order record and note where it exists. Do not merge the two sources for the main rate. A separate reconciliation can show the size of the gap.

Sampling and data collection

Draw the sample from the order system for the full period, not from the support queue. Record the export date, the filters, and the row count. Include every sales channel that can create a replacement. A sample built from escalated cases will overstate the problem.

Capture the replacement order, the original order, the failure category, the evidence type, the approver, the value, the ship date, and the close date. Add the support ticket reference where one exists. Store the export with the report. The reason and approver fields are the core evidence.

Check a small overlap sample with a second reviewer. If the reviewers disagree about whether a reason is complete, the category list is too broad. Rewrite the definitions and retest before the full sample. Agreement matters more than sample size.

Record the time spent matching replacement orders to tickets. A report that requires reading every thread will not be repeated. Note the fields the order system does not capture. The collection cost belongs beside the result.

Measuring missing replacement reasons

Report the share of replacement orders with a complete reason, the share with a partial reason, and the share with no reason. Show the counts behind each share. The three groups describe different evidence gaps and should not be collapsed into one rate. A partial reason still leaves the reviewer guessing.

Break the result down by product, failure category, channel, and approver group. A product with many missing reasons may need a better intake form. A channel with a high share may need a different evidence request. Compare only groups with similar volume and policy.

Report the share of replacements above the desk limit that lack a named owner. A high-value shipment without an approver is a control failure even when the reason is clear. Keep the value band beside the result. The two measures answer different questions.

Limitations and next action

The measure cannot show whether a replacement was justified or whether the customer told the truth. It also cannot show a reason that exists only in a phone call. Treat the result as a measure of the order record. Note the channels and systems that were not included.

Give the result to the support lead and the finance owner with one change to test. The change might be a required reason field, a photo prompt, or a weekly sample. Record the effective date and repeat the measure with the same definitions. A controlled change is easier to judge.

Review the boundary cases before changing the policy. A few product lines with a shared failure may explain a large share of the missing reasons. Fix the intake or the product record first if that is the pattern. The goal is a record a reviewer can follow, not a lower number by definition.