The two-bin system: kanban reordering that works in a stockroom, not just a factory

Two stackable tubs behind the counter of a stationery shop. The front one feeds the pegboard; the moment it empties, whoever emptied it flips the lid off the back one and drops a card in the order tray. No spreadsheet, no app check, no "I thought you reordered." That's the entire system — and for a meaningful slice of a small store's catalog, it beats anything you can install, because the trigger is physical and cannot be forgotten on a busy Saturday.

This is the two-bin system: split one item's stock into two labeled bins, sell only from the front, and treat the empty front bin as the order signal while the back bin covers demand until delivery. It's the simplest form of kanban — the pull-replenishment discipline developed at Toyota, where a card travels to trigger each order (Taiichi Ohno, Toyota Production System: Beyond Large-Scale Production, 1988). Here the empty bin is the card. What follows is the operating manual for that card: the mechanics, the bin-sizing arithmetic, and the places the system breaks.

The mechanics, precisely

Four rules run the whole thing:

1. Sell only from the front bin. The back bin is untouchable while the front holds stock.
2. The empty front bin is the order. Place a standard order for a fixed quantity the same day. Not tomorrow.
3. The back bin covers the wait. Every unit sold between order and delivery comes from it.
4. Receive, then rebuild the front. Delivery arrives, the front bin is refilled, the remainder goes behind it.

A bin card on each tub makes it survive staff turnover: SKU, quantity the bin should hold, supplier, and the order-by reminder. The card is the training manual.

Why it works on a shop floor: the trigger can't be missed the way a dashboard badge can, it fires exactly once per cycle (no double-ordering from two people who both "felt" stock was low), and the cost of running it is two labeled tubs. The trade-off is built in too: it reorders the same quantity every cycle regardless of what demand is doing, so it needs honest sizing and a periodic check.

Sizing the bins, with numbers

Ledger & Loom keeps gel pens at the register — a steady counter item from a regional stationery distributor.

• Sales velocity, from eight weeks of in-stock days: 5.2 pens/day
• Supplier cycle: order Monday, delivery Wednesday — a 2-day lead time, reliably
• Supplier slip scenario: a missed Monday turns the wait into 9 days
• Case pack: 24 pens

The back bin gets sized first, because it carries all the risk. It must cover demand during the wait plus a cushion:

5.2 × 2 days = 10.4, plus a cushion of 8 = 18.4 → round up to the case pack: back bin = 24 pens.

The front bin is the rest of the standard order quantity. Ledger & Loom orders two cases per cycle:

Order quantity = 48 (two cases of 24) → front bin = 48 − 24 = 24 pens.

Running the cycle: the front bin empties in about 24 ÷ 5.2 ≈ 4.6 days, the card drops in the tray, the order goes in, and the back bin's 24 pens cover a 2-day wait that needs only ~10. Now the honest part: if the supplier misses a Monday, the wait stretches to 9 days and demand during it is 5.2 × 9 = 46.8 pens — far more than one back bin holds. Ledger & Loom accepted that risk on purpose: pens are cheap, a dark pegboard for a few days costs little, and the distributor has missed one Monday in a year. A pricier item or a flakier supplier gets a bigger back bin or a different method. Measuring the supplier honestly first is what makes that call rational — your own PO history will tell you the real lead time.

The general sizing rule, then: back bin ≥ demand per day × lead time + a cushion sized by supplier reliability; front bin = standard order quantity − back bin. If the back bin's cushion has to be enormous to feel safe, that's the system telling you two-bin doesn't fit this supplier.

Where two-bin fits — and where it doesn't

Two-bin earns its keep on items that are cheap per unit, steady in demand, and quick to replace:

• Counter and pegboard goods: pens, enamel pins, small accessories
• Stockroom consumables and pick-face backups for fast movers
• Items where the staff turnover rate is higher than the reorder cycle — the system survives people who were never trained on the spreadsheet

It fails predictably on the opposite profile:

• Long or unreliable lead times. An overseas supplier at 38 days turns the back bin into a warehouse: 5.2 × 38 = 198 pens of cash sitting in a tub.
• Big-ticket, slow items. The empty-bin signal is too crude to protect a $200 order.
• Lumpy, unpredictable demand. Spare parts and repair goods sell in bursts the bins can't see coming — why min-max fits spare parts and lumpy demand covers that family.
• A fast-growing catalog. The system's cost is shelf discipline, and discipline scales worse than software.

One method per SKU family is normal. Two-bin for the counter goods, reorder points for the core bestsellers, something else for the tail — the full menu of replenishment methods and where each fits is the map.

Two-bin vs par level vs min-max, in two sentences each

Par level is the same trigger written as a number instead of a tub: when stock counted at or below par, order up — the mechanics live in par level inventory explained. Min-max formalizes it further, with a floor that triggers and a ceiling that sizes the order — min-max inventory explained covers where it breaks. Two-bin is what you choose when the count itself is the friction: the bin does the counting.

Making two-bin survive inside Shopify

The system is physical; Shopify's job is keeping the records true. When the card drops, place the purchase order natively and record the receipt when goods land — native POs, receiving, and counts live in the admin (verified October 2026). The two failure modes that corrupt counts are both procedural: the order placed but never recorded, so incoming stock is invisible; and the receipt recorded twice, so book stock drifts above the shelf. Two-bin makes ordering frequent and predictable, which makes those habits easier — ten small POs a month beat one forgotten one.

The failure modes, so you can spot them early

• The raid. Someone sells straight from the back bin because the front ran out mid-shift and the card "hasn't been processed yet." Now the signal fired but the cover is gone. The card drop and the order must be the same action.
• The quiet skip. "Delivery's coming Thursday anyway, I'll order next week." The back bin absorbs one skip — that's what it's for — and the second skip is a stockout risk waiting for a busy week.
• The unmarked bin. No card, no quantity, no supplier. Six months later nobody remembers which tub is which and the system quietly reverts to memory.

All three are one-person-fixable with a bin card and a standing rule: empty bin means order placed, same day, no exceptions.

Common questions

What is a two-bin system in inventory? You split one item's stock into two labeled bins. You sell only from the front bin; the moment it empties, that empty bin is the signal to place a standard order, and the back bin covers demand until delivery arrives. No counts and no reminders — the trigger is physical, which is why it survives busy weeks and staff turnover.

How do I size the two bins? Size the back bin first: demand per day × replenishment lead time + a safety cushion, rounded up to your supplier's case pack. The front bin is the rest of your standard order quantity. If the back bin needs an enormous cushion to feel safe against one slipped delivery, that's the signal the supplier cycle is too unreliable for two-bin.

Is a two-bin system the same as kanban? Two-bin is kanban in its simplest form. In factory kanban, a card travels to trigger each order; here, the empty bin is the card. Same pull principle — replenish what was actually consumed rather than what a forecast predicted — stripped down until a two-person stockroom can run it with zero software.

What are the disadvantages of a two-bin system? Three real ones. Cash sits in the back bin, so long or unreliable lead times inflate it badly. It reorders the same quantity regardless of demand changes, so a trending SKU drifts out of balance until you re-size. And it lives or dies on shelf discipline — one person raiding the back bin breaks the signal for everyone.

Can a two-bin system work with Shopify? Yes — the trigger stays physical while Shopify keeps the records true: place the PO when the bin empties and record the receipt when goods land, or book stock will drift from the shelf. Shopify's native purchase orders and receiving handle both steps, as of October 2026, and frequent small POs suit them well.

Next step

Pick one counter item this week — one SKU, one supplier, a lead time you can measure in days — and size the bins with the arithmetic above. Run it for a month beside your usual method and see which one you trust on a busy Saturday. When you're ready to place two-bin among the alternatives for the rest of the catalog, the full menu of replenishment methods is the map.