Accept a NetSuite WMS picking process only after testing what happens when a bin is short, an operator stops, an order changes, or fulfillment generation fails. A wave released successfully is the beginning of warehouse execution. It does not prove that picked stock, staged goods, and fulfillment records will agree at the end.
Use a small wave with deliberately difficult cases and record expected outcomes before operators begin. The test should establish who owns the shortage, whether work can resume safely, and how the remaining order quantity stays visible.
For each test, capture the order and line, wave, pick task, item, required quantity, chosen inventory detail, picker, staging destination, and resulting fulfillment. Add the actual physical state after every exception.
Use one test sheet per scenario rather than a screenshot collection. Record the action, expected result, observed result, transaction references, and pass or fail reason. This makes it possible to distinguish a product configuration issue from an operator following the wrong recovery procedure.
Confirm the installed WMS and SCM Mobile components, account release, roles, and warehouse preferences. Third-party WMS products and older NetSuite warehouse products can use different task and fulfillment behavior. Test the actual application rather than relying on familiar terminology.
Choose an item whose system quantity exceeds the physical quantity in the recommended bin. Do this safely in a sandbox or controlled test location. Have the picker report the actual quantity through the supported mobile flow instead of entering the expected quantity to finish the task.
Observe the available alternatives, such as another eligible bin or an appropriate no-stock action in the configured picking mode. Record whether the remaining requirement stays open, whether another task is needed, and who investigates the bin discrepancy.
Do not treat the short pick itself as proof that inventory should be adjusted. The goods may be in another bin, already staged, or affected by a transaction timing issue. Keep warehouse execution recovery and inventory correction as separate decisions.
A wave includes Order A for 12 units and Order B for eight units of the same item. The recommended bin is recorded as holding 20, but the picker finds only 15. Five additional units are visible in another bin, but their status is under inspection.
The picker records 15 actual eligible units through the approved flow. The supervisor then determines how those units are assigned to the orders under the business priority policy. The inspection units remain held; they are not a convenient way to make the wave complete.
If A receives 12 and B receives three, B still has five outstanding units. The test must prove the resulting staged quantities, fulfillment quantities, and remaining demand. A “done” task status alone cannot answer all three questions.
Next, release the inspection stock through a separate authorized test decision and demonstrate the approved process for the remaining five. Confirm that the second execution does not recreate A's shipment or duplicate B's first three units.
NetSuite WMS provides a Mark Open Picks Done process for eligible partially picked orders. Fulfillment generation follows the account's Per Order or Per Order Line preference. Tasks with no picked quantity do not contribute quantity to a fulfillment simply because they are marked done.
Test both partial and untouched tasks in the same order. Under a per-order process, unresolved task state can affect when a fulfillment is generated. Use the current supported menu and account procedure rather than assuming an older partial-fulfillment menu remains appropriate.
Before marking work done, require a clear physical-state check and an approved shipment decision. Closing a task can end the current warehouse work while customer demand remains. Assign ownership of the remaining quantity explicitly.
Have one operator begin a task and stop at a defined point. Record goods already picked and their staging location. Then demonstrate the supported reassignment or resumption process with another operator.
The second operator should be able to distinguish completed physical work from remaining work. If the process relies on a supervisor handoff, document it. Do not assume a task's assignee field alone communicates where a half-picked carton was left.
Include a device interruption or session timeout during a test transaction. Before repeating the action, inspect whether the first attempt created a record. A slow response can lead to duplicate scanning or movement if staff assume nothing happened.
A pick task can show a fulfillment-generation failure caused by concurrent inventory changes, order edits, or fulfillment activity outside the wave. Inspect the recorded failure reason and relevant transaction history before retrying.
Supported recovery options can include retrying updates after resolving the cause, overriding fulfillment creation, or reversing picks. These represent different outcomes. Choose one based on the physical goods and approved transaction plan, not whichever button is easiest to reach.
Test the expected record count after recovery. If one fulfillment should exist, verify that exactly one does and that it contains the intended quantities. Reconcile order remaining quantity and staged stock before allowing shipment to proceed.
Change a test order's quantity, location, or cancellation state after release using the approved process. Observe how the warehouse task responds. Establish which edits are permitted and which require withdrawal or other controlled handling of released work.
A customer-service change that does not reach the floor can produce a correct pick against an obsolete instruction. Conversely, a manual fulfillment created directly from the order can conflict with WMS generation. Restrict process bypasses through appropriate permissions and training, validated in the target account.
Include customer shipment rules. A partially picked order may be technically fulfillable while the customer requires a complete delivery. Keep the commercial approval visible to the warehouse supervisor before releasing staged goods.
Require evidence for a short bin, unavailable alternative stock, interrupted picker, partial task completion, untouched task, concurrent order change, generation failure, retry, and pick reversal. Add lot or serial substitutions where relevant.
For every scenario, verify four outcomes: physical custody, pick-task state, fulfillment record, and remaining demand. Include the responsible operator and supervisor roles in the test. A result that works only as administrator is not ready for warehouse use.
Document what remains manual. A manual shortage review can be acceptable if its owner, timing, and evidence are clear. An invisible manual step is a recurring failure waiting to happen.
Write the runbook around symptoms: bin short, no eligible alternate, task interrupted, fulfillment missing, and order changed. For each, state the initial checks, supported recovery route, approval required, and final reconciliation.
Keep configuration versions and test evidence with the runbook so release changes can be assessed. Repeat the high-risk scenarios after relevant WMS, mobile, script, or integration changes.
CuriousRubik's NetSuite support services can support an account-specific exception review. Bring one failed task with its physical and transaction history; that is more useful than a general report that wave picking is unreliable.
No. Wave, task, fulfillment, and remaining-order states answer different questions. Reconcile actual picked and staged goods with fulfillment records and outstanding demand before considering the order operationally complete.
Marking an unstarted task done does not give it picked quantity to include in a fulfillment. Verify the account's partial-pick process and the resulting fulfillment rather than interpreting Done as shipped.
Not automatically. Investigate location, staging, recent transactions, and physical stock first. The picking exception and the inventory correction need separate evidence and appropriate approval.
Use the supported retry after the recorded fulfillment-generation cause is resolved and existing records have been checked. Confirm the expected fulfillment count and quantities afterward to avoid duplicate or inconsistent recovery.
The new operator can identify remaining work and the location of already picked goods, continue through the supported process, and finish with matching task, staging, fulfillment, and remaining-demand records.