RETURNS OBSERVATION — BLANK DISCUSSION WORKSHEET

Use this original preparation template to discuss one returns workflow. It contains no observations, pilot results or KPI targets. Fill the blanks together and adapt the definitions to the actual site. An observation can describe a process; it does not establish why performance changed or guarantee an improvement.

1. AGREE THE BOUNDARY
Choose one receiving-to-assessment segment. Start with physical presentation at the agreed receiving handoff and end with the first recorded assessment decision. A carrier delivery, return request, refund or stock update is a different event. Define any product-level or multi-item observation separately.

Site and receiving point: ____________________
Product group and observation unit: ____________________
Included parcels, shifts and eligibility rule: ____________________
Exclusions, reason codes and separate count: ____________________
Baseline dates, coverage and timezone: ____________________
Later observation dates, coverage and timezone: ____________________
Selection rule and its limitations: ____________________
Follow-up cutoff for unfinished parcels: ____________________

2. ASSIGN THE WORK
Name the workflow owner, observers and the person maintaining the log. Agree who can clarify an exception or uncertain parcel match. Walk the actual route with the intended people and devices before collecting observations. Keep clearly labelled test records outside the observation counts.

Workflow owner: ____________________
Observer and device at each handoff: ____________________
Log and matching owner: ____________________
Definition reviewer and unresolved questions: ____________________

3. DEFINE EACH EVENT
Write a local definition and evidence source for each event. Use only events that can be recognised or evidenced. Do not infer active handling from an elapsed interval or a system timestamp alone.

Arrival presented — observable handoff: ____________________
Assessment ready — agreed readiness criteria, if observable: ____________________
Assessment started — observable first action, if needed: ____________________
First assessment recorded — decision categories and saved evidence: ____________________
Definition version and date: ____________________

4. KEEP EVENT AND RECORD TIMES SEPARATE
Use a project sample key rather than customer names, addresses or payment details. Record when the event occurred separately from when it was written down or became available. Include timezone, clock/device, precision and known alignment limits. Leave unknown times blank and explain why. Record availability delay is not automatically network latency.

One row per event:
Sample key | Event | Occurred at | Recorded/available at | Observer/device/clock | Evidence or uncertainty
__________ | _____ | ___________ | _____________________ | _____________________ | _______________________

Ambiguous or multiple parcel matches: ____________________
Missing-time reason and person reviewing it: ____________________

5. RECONCILE THE RECORDS
For each window, compare the eligible intake register, selected parcels and usable matched observations. If the intake count is unknown, say so. Calculate an interval only from compatible, matched start/end times; show its usable denominator. Preserve missing and pending records rather than treating them as zero-duration cases.

Enter baseline / later counts or mark unknown:
Eligible parcels and register source: ____________________
Selected eligible parcels: ____________________
Matched usable start/end observations: ____________________
Missing start only / missing end only / both missing: ____________________
Still unfinished at cutoff: ____________________
Duplicate rows / ambiguous matches / resolutions: ____________________
Excluded parcels by agreed reason: ____________________
Unreconciled intake or recording gaps: ____________________

6. REVIEW CONTEXT AND SHARE A BOUNDED OUTPUT
Record differences in staffing, shifts, workload, product mix, batching, layout, routes, assessment practice, devices, software and recording methods. Keep process changes separate from measurement changes. A later window alone does not prove a change caused an improvement.

Concurrent changes and supporting evidence: ____________________
Sharing scope, permitted access and retention: ____________________
Review participants and meeting: ____________________
Outputs: checked workflow map, event dictionary, bounded log, reconciliation and uncertainty notes.
Descriptive findings, alternative explanations and next questions: ____________________

Participants agree the project and any data access. This blank template establishes no research partnership, validated protocol, measured result or service guarantee.