Useful Tools

Reference

How to reduce picking errors: 8 causes and their fixes

Every mis-pick is one of eight failures — wrong item, wrong quantity, wrong order, skipped line and friends. What each one costs, what causes it, and the cheapest fix that actually works.

By Bastien HugonFounder & Engineer9 min readPublished July 9, 2026

TL;DR

Picking errors cluster into eight causes, and most stores only suffer from two or three of them. The highest-leverage fixes, in order: pick from a consolidated list in shelf order (kills revisits and line-skipping), name bins unambiguously (kills wrong-shelf picks), split bundles into components (kills the kit mis-pick), and separate picking from packing (turns packing into a free verification pass). Scanners help — after those four, not instead of them.

A mis-pick costs far more than the item: return shipping both ways, a replacement shipment, support time, and — repeated often enough — a review that mentions 'sent the wrong thing'. Industry benchmarks put manual-picking error rates around 1–3% of lines, and each error's fully-loaded cost at $20–60 for a small parcel operation once labor and shipping are counted. The good news: errors are not random. They are eight specific failures, each with a specific cheap fix.

The eight causes

#ErrorRoot causeCheapest effective fix
1Wrong item — look-alikeSimilar products adjacent on the shelfSeparate look-alikes physically; SKU on the list *and* the shelf
2Wrong item — wrong shelfAmbiguous locations ('third shelf on the left')Bin codes on every location, printed per line
3Wrong quantityQuantities recomputed in the picker's head across ordersConsolidated list with totals precomputed
4Skipped lineList not in walk order → backtracking → lost placeSort the list by bin; tick each line as picked
5Wrong order at packingPicked pile split by memoryPack against per-order packing slips, one order at a time
6Kit missing componentsBundle shown as one lineExplode bundles into components on the list
7Wrong variant (size/color)Variant not prominent on the listVariant in bold next to the title; barcode scan where stakes are high
8Stale listOrder edited/cancelled after printingPrint the batch list right before the run, not the night before

Fixes that pay for themselves the same week

  1. 01

    Consolidate and sort the pick list

    One list for the batch, every product once at its total quantity, lines in bin order. This single change addresses causes 3, 4 and half of 8 — [how to set it up](/guides/print-one-pick-list-for-multiple-orders).

  2. 02

    Put bin codes on paper and on shelves

    Zero-padded, prefix-consistent codes (A-04, not A4) printed on each line and labeled on each shelf edge. Cause 2 disappears; new staff become productive in a day.

  3. 03

    Make variants shout

    Size and color in bold at the line level. Most wrong-variant picks happen because the list whispered 'M' next to a title in the same type size.

  4. 04

    Split picking from packing

    Pick the batch, then pack per order against packing slips. Packing becomes a natural second count — cause 5 gone, and causes 1/7 get a second chance to be caught.

  5. 05

    Reprint on any batch change

    An edited or cancelled order invalidates the printed list. Reprint policy: list is generated after the batch is final, minutes before the walk.

Measure before and after: error rate = mis-picked orders ÷ shipped orders, weekly. Stores moving from per-order picking to a sorted consolidated list typically watch the rate drop by half or more in the first month — and the measurement itself (a tally at the packing bench) costs nothing.

When scanning becomes worth it

Barcode-scan verification at picking or packing catches wrong-item and wrong-variant errors mechanically — at the cost of hardware, station setup and slower lines. The honest threshold: after the list, bins and bundle fixes above, if wrong-item errors still appear weekly (look-alike-heavy catalogs, rotating temp staff, high volume), scanning is the next step; the scan-first apps exist for exactly that profile. Adding scanners *before* fixing the list just verifies a chaotic walk more precisely.

Doing this with Solvi Pick Lists

Solvi Pick Lists implements the paper-side fixes as defaults: one consolidated list per batch with totals precomputed, natural-sorted bin order, bold variants, bundles split into components, and checkbox lines for the tick-as-you-pick habit. The five-list free plan is enough to run the before/after measurement on your own error rate.

Frequently asked questions

What is a normal picking error rate?

Manual operations without process controls commonly sit between 1% and 3% of picked lines; well-run list-driven operations reach 0.1–0.5%, and scan-verified operations push below 0.1%. The spread between those tiers is process, not talent — the same team moves tiers when the list and the shelves stop fighting them.

What does one picking error actually cost?

For a typical small-parcel store: return label + reshipment postage, the replacement item's handling, 15–30 minutes of support and packing time — $20–60 fully loaded, before any customer-lifetime damage. At 30 orders/day and a 2% error rate, that is roughly 18 errors and $400–1,000 a month.

Do barcode scanners eliminate picking errors?

They eliminate wrong-item and wrong-variant picks at the verification point — and do nothing for skipped lines on a chaotic list, stale batches, or kit components a bundle line never showed. Scanning is the right second investment; the sorted consolidated list is the right first one.

How do I reduce errors without buying anything?

Three free changes: pick from one consolidated list in shelf order instead of order-by-order, write bin codes on shelf edges (masking tape works day one), and pack against packing slips as a separate step. Most stores see the majority of their errors disappear with exactly those three.

Related guides

Do this in minutes with Solvi Pick Lists

Turn selected orders into one consolidated pick list — sorted, bundle-aware, printed in order.