Warehouse Operations
Warehouse bin numbering and sequencing
Designing bin labels and pick sequence so pickers move through the warehouse efficiently.
Last updated September 16, 2026
The label on the bin and the number that routes the picker are not the same thing. Mix them up and the walk is wrong even when every barcode scans.
Humans need aisle-rack-level-position. Something like B-29-1-3, left to right, bottom to top, fixed width, no I/O/0/1 if you can help it. That string is for finding the hole. It is not the pick path.
Pick sequence is a separate integer
The path is a separate sequence integer. Allocation walks lowest sequence with stock, then name if two bins share a number. Number in gaps: 10, 20, 30, or thousand-step blocks per zone so you can insert a rack without a full renumber. A new rack between 20 and 30 gets 25. Zone A can own 1000-1999, Zone D 3000-3999. Never leave sequences blank or duplicated. Ties fall back to alphabetical order, which almost never matches the floor. B-04-1-1 and C-12-2-2 both set to 10 will send the picker to B first, even if C is the next hole on the walk.
Choose one dominant sort order
Pick the dominant sort on purpose. Aisle-first (serpentine) keeps the cart in the aisle. Level-first (golden zone) empties waist-high bins before ladders. The textbook failure is mixing them. The same SKU sits in B-04-4-2 at sequence 1280 (level 4), B-04-1-2 at 2100 (level 2), and D-12-1-1 at 3320 (level 1). Lowest sequence wins. The picker climbs to 1280 while 3320 sits at waist in another zone. Keep zone ranges non-overlapping. Put overstock, reserve, and sales-floor bins in a high band so they are consumed last. Separate picking zones from overstock zones.
Sequence also drives replenishment
The same sequence drives replenishment transfers. If you want putaway to walk B then C then D, those bins get the low numbers. You cannot split pick path and transfer path. Preferred bin is a putaway default, not an override of which bin gets picked. When the preferred bin is empty, putaway still aims there. Pick still walks 1280. A SKU-level sequence that does not match the bin sequence will lie the same way.
Maintain sequencing over time
Maintain it like a map, not a one-time setup. At 20,000 bins you will not click through this. Export locations, rewrite sequences in the CSV, re-import, then test on a two-location shipment before you touch the whole building. Build the test on the SKU that lives at 1280 and 3320, two units, and walk it. After a bulk receipt that dumped 400 units into high racks, resequence is not enough. Move the units down or the path will still send people up.
Connect Claude or ChatGPT to your Fulfil data with the Fulfil MCP, then run this prompt on your own numbers.
You are a warehouse lead for a DTC brand. Here is the location master: location_id, name, zone, type (pick/overstock/sales-floor), aisle, rack, level, position, sequence, preferred_bin, on_hand. [paste] Produce: 1. Whether the current scheme is aisle-first or level-first, and where the two conflict. 2. Overlapping sequence ranges across zones. 3. Duplicate or blank sequences (these will sort alphabetically). 4. A proposed gap numbering (10/20/30 or zone blocks) that puts pick-face levels 1-2 first, overstock last. 5. A two-location test shipment to validate the new walk before a full import. Flag any SKU that would pick a high-level bin while a lower-level bin of the same SKU has stock.
Explore this in Fulfil
More on Warehouse Operations
Zone picking vs batch picking vs multi-order picking
Choosing the picking method that fits your order profile and warehouse layout.
FEFO vs FIFO and lot tracking
How FIFO and FEFO differ and when expiry dates should drive what your team picks.
Ecommerce returns: restock, quarantine, or write-off
Deciding whether each returned item goes back to stock, into quarantine, or gets written off.
See it run on
your data.
Fulfil runs inventory, fulfillment, purchasing, and accounting for scaling DTC brands in one system.