Scope: sample consecutive handoffs from one Philippines operations lane during a fixed month. Include completed, waiting, corrected, cancelled, and exceptional items. Before selection, define the five facts a reviewer must recover: requested outcome, current state, supporting source, unresolved uncertainty, and next owner with deadline.
Methodology: give authorized reviewers the approved shared record but no private chat or help from the preparer. Record whether each required fact is correct, time to reconstruction, extra systems opened, and clarification questions. Randomize order, double-score a subset, and retain failures instead of replacing them with easier examples.
Inference boundaries: ILO, GAO, and NIST provide context for remote work, records, and governance. They establish no universal recovery-time target and do not show that a record format, provider, worker, or location caused the outcome. The results apply only to the sampled lane and access conditions.
Limitations: reviewer familiarity, search skill, permissions, item complexity, redaction, and platform indexing influence retrieval. Silent guessing can appear fast, so score correctness and confidence alongside time. Use findings to test a specific record improvement and repeat the protocol.