Reconcile manual dispatches before the automated queue resumes

Inventory what physically left during the outage

While the system was unavailable, staff may have picked, packed and handed parcels to a carrier using paper pick lists, a spare manifest or a handwritten handoff log. Those physical acts are real even though no system recorded them. Collect every genuine dispatch artifact created outside the system: the carrier manifest or acceptance scan, the packing note, the signed internal handoff entry. For each one, record the order reference, the items and quantities, and the dispatch time as actually documented.

Resist two shortcuts. A manually marked status alone does not prove dispatch, because the mark may have been made to keep a list tidy rather than to record a handoff. A missing status does not prove nothing shipped, because the system was the thing that was down. The reconciliation list must be grounded in the physical evidence, and it should use internal references only, not customer details.

Match each manual dispatch to a waiting queue task

Place the dispatch list beside the queued order tasks and connect them row by row. Three outcomes are possible. A queued task matched by solid dispatch evidence is completed manual work: it must be marked or held so the queue does not release it again. A queued task with no matching dispatch evidence is not automatically work to release, because the outage is exactly the condition under which evidence goes missing; it stays unresolved and held until a further check — packing records, carrier acceptance data, staff accounts — positively establishes the goods were not dispatched. A dispatch record with no matching task is unresolved: the order may have been created manually during the outage, or the reference may be wrong, and it needs investigation rather than an assumption.

Rows that cannot be matched stay open and are assigned to the authorized owner. Do not force a match because two references look similar, and do not create replacement fulfillment work to make the list balance. The open rows are the actual answer to how much risk remains in the queue.

Read the queue in the system you actually run

Interpret the queue through your own platform's documented behavior. On Shopify, bulk fulfillment reports failures for review, so a failed bulk action leaves you a list to inspect rather than a clean rollback of everything attempted. Shopify also distinguishes fulfillment controlled by a third-party fulfillment service: where a service owns the fulfillment, its status is the operational truth, and an administrative mark in the store does not override it.

The same Shopify documentation notes that a cancellation request can fail when processing is too far advanced. Neither an administrative cancellation nor a success response proves that physical picking stopped or that dispatch occurred. Treat every bulk result, cancellation result and service-controlled status as a prompt for order-level reconciliation. These are Shopify behaviors; if you run another system, its own documentation governs what its queue states mean.

Authorize retries one order at a time

The fulfillment owner classifies each row before the queue resumes: completed manually, so suppress it from release; positively established as still unfulfilled, so permit the retry; or unresolved, so hold it. Never replay the entire queue on the assumption that the outage blocked everything, and never treat the absence of dispatch evidence as permission to release. The realistic failure mode is a duplicate release, where a buyer receives a second parcel because the manual dispatch was never connected back to the task that produced it.

Where a third-party fulfillment service is involved, confirm with that service what it already processed during the outage window before releasing anything on your side. Where a cancellation or hold failed during the outage, recheck the order's current state rather than assuming the request eventually took effect.

Resume, verify, and record the handoff

Resume the queue in a controlled sequence. After each authorized release, verify the resulting status and confirm no duplicate dispatch occurred, then record who authorized the release and what evidence closed the row. Completed manual dispatches should be annotated so support staff can see the real dispatch date and method when a buyer asks, instead of reading a queue state that predates the reconciliation.

Keep the reconciliation sheet and the physical dispatch artifacts in authorized internal storage. If the outage exposed a wider concern with how the checkout and order workflow pass work to fulfillment, describe that concern through a Prism consultation and confirm scope, responsibilities, fees and terms before any work begins.

Manual dispatch reconciliation sheet

Complete one row per order task or dispatch record from the outage window. Use internal references and genuine evidence. Unmatched rows stay open with a named owner; the sheet does not authorize a queue-wide retry.

Worksheet entries are not submitted by Prism’s worksheet and are not saved by the site. Use record types, availability, anonymized observations, or match/mismatch results. Do not enter government identifiers, customer names or addresses, customer messages, receipt-access links, card or bank details, passwords, or keys. Send sensitive documents only through the provider’s verified secure channel.

These are temporary notes. Leaving or reloading this page may clear them. Worksheet entries are not sent automatically. If you copy notes into the consultation message and submit the form, Prism receives them as part of your request.