Identifying the cause of a repeated mis-pick: which step selects the wrong SKU?

Two near-identical products sit in adjacent bins. Over several shifts, orders calling for SKU A receive SKU B. A reminder to “be careful” may not solve the problem if the bin labels are swapped, the package design hides the distinguishing detail, or the scanner accepts the wrong code. Identifying the cause of a repeated mis-pick means following the error path before choosing retraining, relabeling, relocation, or a system fix.

This is a root-cause decision about a recurring wrong-item pattern. A picker’s career guide explains the retrieval task; an order-accuracy metric describes a broader result. Here the operator has a specific pair of SKUs, a repeated failure, and several plausible mechanisms. The investigation should determine which mechanism is supported by evidence and whether more than one condition contributes.

Define the repeated mis-pick pattern

A mis-pick pattern identifies what was expected, what was selected, where, and when. “Pickers keep making mistakes” is too broad. For the adjacent-bin example, collect the order SKU, selected SKU, bin location, shift, device event, and point where the error was found. Check that the cases really involve the same pair and not several unrelated item problems.

Start with a small set of verified examples. A packer may find SKU B in a tote assigned to an order for SKU A. A customer report may name the same mismatch after delivery. Those observations are not identical evidence: one is caught before shipping, the other later. The investigator should trace each case back to the pick event without assuming the final wrong item was caused at the bin.

Look for a denominator. Ten wrong picks from a bin handling hundreds of selections is a different pattern from ten errors in a dozen selections. Compare the error rate with similar product pairs or other locations where possible. A high total may reflect high volume rather than an unusual failure mechanism. The goal is to identify where the process is unusually fragile.

O*NET’s stocker and order-filler profile includes reading orders for catalog numbers, sizes, colors, and quantities. That broad occupational fact helps explain why similar variants can matter during picking. It does not establish how this hypothetical warehouse labels bins or configures scanners. Those details must be observed in the actual process.

Record changes during the period. Did the SKUs move to adjacent bins last week? Did a new package design arrive? Was a scanner workflow updated? A timing match is a lead, not proof, but it helps decide which causes to test first.

Compare product similarity with the pick instruction

Product similarity can be visual, textual, or structural. Two cartons may share color and brand while differing in a small size code. Two item names may differ by one character. The investigator should compare the actual products as workers see them at the bin, not only catalog images on a screen.

For SKU A and SKU B, take a controlled look at the label, package front, barcode placement, and any quantity marking. Does the instruction show the distinguishing field clearly? If it names only a shortened description, the worker may have to search for a hidden code. If the code is prominent on the device and product, similarity may be less likely to explain repeated errors on its own.

Ask workers what they use to tell the products apart. Do not begin with “Why did you ignore the label?” A worker may reveal that the printed size is on the back of the pack or that a glare obscures it under shelf lighting. Those observations can be checked. The investigation should treat staff explanations as evidence to examine, not as automatic blame or automatic proof.

Similarity may interact with pace. If the bin is reached during a high-volume wave and the distinction requires turning each unit, the process may encourage a visual shortcut. The fix might involve clearer presentation or a stronger verification step, not a general lecture on attention. The operator should test whether the proposed change makes the correct choice easier under normal conditions.

A product redesign may be outside the warehouse’s authority. The team can still separate the two SKUs, improve bin labels, or ask a supplier for clearer packaging through the approved route. First establish which feature is creating the confusion and what local controls can address it.

Verify bin labels and physical placement

A bin label links a storage location with an expected item or location identifier. If labels are swapped or placed between two adjacent bins, even a careful worker may follow the wrong cue. The investigator should compare the system’s location assignment with the physical labels and stock at both bins. A photo of the shelf can preserve what workers saw before anything is moved.

In the example, perhaps SKU A’s label is fixed to the divider closer to B’s stock. Workers may interpret it differently from the system’s intended boundary. Perhaps replenishment placed B in A’s bin. Those causes require different corrections. A new training reminder will not repair a swapped label or mixed physical stock.

Check whether the problem follows the location or the item. If the same SKU pair is stored elsewhere without errors, the local layout deserves attention. If the pair causes errors in several locations, product or instruction similarity may be stronger. This comparison does not prove a cause by itself, but it narrows the test.

Ask when the bins were last replenished and whether a stock move occurred before the first reported error. Replenishment records and physical counts may show whether goods were placed incorrectly. Avoid changing the shelf before documenting its original state, unless an immediate safety or customer-impact control requires a hold. The investigation needs to know what was present when the error occurred.

If the team temporarily separates the products to protect current orders, record that containment as a temporary action. Later error rates may improve because the exposure changed, but that does not by itself tell whether the root cause was label placement, package similarity, or scan logic. The permanent fix should match the evidence.

Test scanner behavior at the exception point

Scanner behavior includes what the device asks the worker to scan, which codes it accepts, and what happens when the code does not match the order. The investigator should use a controlled test environment or an authorized procedure. They should not create false live orders or change production settings simply to see what happens.

Suppose the device accepts SKU B’s barcode when the task calls for A. That would be a strong lead about configuration or item-master mapping. If the scanner rejects B and the error still occurs, perhaps workers bypass the exception, use a manual override, or the wrong item enters the tote later. Each path calls for a different next check. The system log should be read alongside the physical process.

Some scans confirm a location but not the individual unit. A successful location scan can therefore be mistaken for item verification. The investigator should find out exactly what each beep and screen message means in this workflow. Barcode identity basics are a separate topic; this decision is about whether the current scan control prevents this specific wrong-SKU selection.

Look for timing and overrides. Were mis-picks associated with a device update, a new manual-entry option, or intermittent connectivity? A log may show that the correct item was scanned even if the wrong unit later reached packing, which would shift attention to tote mixing or a post-pick handoff. Do not force every wrong parcel into a scanner explanation.

Ask systems support to verify configuration when the evidence points there. Operations can describe the observed error path and provide IDs and timestamps. A software team can determine whether a mapping or validation rule is wrong. The root-cause decision should preserve these distinct responsibilities.

Use an error-path test without risking orders

An error-path test uses a safe, authorized setup to see how a wrong choice could pass through the process. It might use training inventory, a test order, or a supervisor-observed demonstration under the employer’s rules. The purpose is to confirm a plausible mechanism, not to trick a worker or knowingly ship a bad order.

For the SKU pair, the team could ask a trained worker to walk through the normal instruction at the actual shelf while an observer notes where the distinction appears. A separate system test can check which barcode is accepted. If a label is ambiguous, the observer can record how different workers interpret it. The test should isolate one variable when possible so the result is interpretable.

ASQ’s problem-solving resources describe root-cause analysis as a way to uncover causes before corrective action. The principle applies here: a proposed fix should follow evidence. A repeated error may have several contributing causes, and a single neat “why” can oversimplify the process. The investigator should keep alternative explanations visible until testing rules them out.

If the test cannot reproduce the error, do not declare the problem imaginary. The live cases may depend on a busy wave, a replenishment event, or a screen state absent from the test. Compare the test conditions with the real incidents and look for the missing factor. A failed reproduction is evidence about the test’s limits as well as about the hypothesis.

Protect current orders while the investigation continues. A temporary second check or separation of the SKUs may be appropriate under the site’s rules. Record the containment and who owns it. The team should not quietly remove it later without confirming the permanent correction works.

Choose a cause-matched correction

A cause-matched correction changes the mechanism that produced the wrong selection. If the bin labels were swapped, correct and verify the labels and stock. If the products are too similar at the point of pick, separate them or improve the distinguishing cue. If the scanner accepts the wrong item, fix the mapping or validation through the systems owner. If the approved procedure is clear but workers have not learned it, targeted training may help.

Blanket retraining is weak when the environment remains misleading. Relabeling is weak when the device records the wrong SKU as correct. A software change is weak when replenishment regularly mixes the products in one bin. The investigation may require two coordinated fixes, such as cleaning up placement and strengthening scan verification.

After the change, monitor the same error definition used at the start. Did SKU B stop appearing on orders for A across comparable volume and shifts? Were wrong items caught earlier rather than reaching customers? A short run without an error is encouraging but may be too little to prove durability. Set a review period suited to the site’s volume.

Document the before-state, test, correction, and result. Future workers can then see why the bins were separated or the scanner rule changed. If the same pair starts failing again after a product refresh, the team has a usable history rather than another vague “be careful” message.

Keep the analysis distinct from picking and accuracy metrics

A picker’s immediate task is to select the correct item and quantity from the assigned location. This article begins after a repeated wrong-item pattern has been verified and asks why it recurs. The order-accuracy rate can tell a manager that errors rose, but it does not identify whether these adjacent bins, product labels, or scan controls caused them.

The two-SKU example is a good test of investigative discipline. If someone proposes moving the bins, ask what evidence supports layout as the cause. If someone proposes training, ask what skill was missing. If someone proposes software work, ask what the device actually accepted. The team can act quickly to contain errors while still reserving a permanent fix for the tested mechanism.

Delpack’s transferable skills for a delivery driver resume guide concerns how workers describe real experience. An operations applicant can cite a documented mis-pick investigation if they compared evidence and helped verify a correction. They should not claim root-cause ownership from merely noticing one wrong item.

Fix the mechanism behind the repeated selection

For a real pattern, define the expected and selected SKUs, compare affected cases and total pick volume, inspect the shelf and labels, test the scanner path safely, and trace where the wrong item first entered the order. Choose temporary containment where needed, then make a correction tied to the supported cause. Recheck the same error pattern after the change.

Identifying the cause of a repeated mis-pick protects orders by replacing assumption with a tested explanation. Adjacent near-identical SKUs may point to product presentation, bin labeling, scanner validation, or more than one condition. The right fix is the one that interrupts the actual wrong-item path.

Delpack’s Fort Wayne driver careers page lists driver opportunities, not a description of Delpack’s warehouse bins or scanning system. The SKU example is hypothetical. Apply the investigation only with the actual item records, shelf state, device behavior, and authorized correction process.