How to Forecast Inventory for Costco: Pallets, Regions and the Exit Scenario

Forecasting inventory for Costco is a pallet problem, not a shelf problem. Costco carries a deliberately small assortment, sells by the pallet from the warehouse floor, buys by region, and decides quickly whether an item stays or rotates out. Your forecast therefore has to answer three questions: how many pallets per warehouse per week will move, how large the initial fill is across the region you were awarded, and what happens to cash and inventory if the item exits after one rotation. This guide works through each with a fictional example.

How Costco orders from a supplier

Costco operates roughly 600 US warehouses and about 900 worldwide, organized into buying regions. Most new items start as a regional buy with a single regional buyer, not a national one, so your first program may be 60 to 120 warehouses rather than the whole chain. National status comes later, if velocity earns it.

Product typically moves one of two ways: to a Costco depot, which cross-docks pallets to warehouses in its region, or directly to warehouses for some categories and regions. Either way the unit of commerce is the pallet. Costco's warehouses have no backroom to speak of; what is on the floor is the inventory, so replenishment is driven by how many pallets each building has left and how fast they are moving. POs typically arrive by EDI with firm delivery appointments, and late or short deliveries draw chargebacks.

Two features of Costco's model matter more for forecasting than at any other retailer:

• Continuous versus in-and-out. Some items are continuous (always stocked), many are in-and-out programs with a defined window. Ask the buyer which yours is; the forecast shape is completely different.
• Fast exit decisions. Costco does not nurse slow items. If units per warehouse per week fall below what the buyer needs to justify the pallet position, the item is deleted and the remaining inventory is marked down. Model the exit scenario before you model the expansion.

Costco's economics are also unusually transparent. It publicly caps its markup on branded goods at roughly 14%, does not charge slotting fees, and is known for paying suppliers faster than mass retailers, often in the neighborhood of 30 days. Confirm your terms, but the cash cycle is typically kinder than at grocery or drug, which partly offsets the size of the orders.

Velocity and warehouse-count math

Replace "store" with "warehouse" and "facings" with "pallets" and the math is familiar:

Weekly regional demand = units per warehouse per week x active warehouses

Pallets per week = weekly regional demand / units per pallet

Weeks of supply at Costco = (pallets at depot + pallets on the floor) x units per pallet / weekly regional demand

Velocity benchmarks are the hard part before launch. Costco shoppers buy in multiples and a pallet position sells far more units per door than a grocery shelf, so DTC or grocery velocity badly understates it. Use the buyer's expectation for the category, a comparable club item, and if you can get one, a roadshow or demo result from the same region. After launch, Costco provides item-level sales by warehouse, and that replaces every assumption within a few weeks.

Promotions and the MVM

Costco's main promotional vehicle is the Multi-Vendor Mailer (MVM), a coupon book that runs for a defined window of a few weeks, with the supplier funding the discount. Being in an MVM typically lifts velocity sharply, and the buyer will expect you to have pallets in position before it starts. Model three effects: a build order two to four weeks ahead of the window, elevated sell-through during it, and a softer few weeks afterward as members work through what they bought. Demos run by Costco's demonstration partner are the other lever, usually funded by the supplier on a per-day basis; they lift that warehouse that week, which helps with trial but is not a run-rate signal.

Our post on how to forecast Costco sales like a finance pro walks through the MVM lift math in more detail, and the Costco forecasting template has the promo window and build order as inputs.

A worked example: Harbor Street Coffee in one region

Consider a fictional coffee brand we will call Harbor Street Coffee, awarded a regional program for a 2-pound club pack at 80 warehouses. Assumptions: retail $14.99, Costco's cost $13.15 (consistent with its markup cap), landed cost $8.50 per unit, 288 units per pallet (48 cases of 6), initial fill of 2 pallets per warehouse, steady velocity of 150 units per warehouse per week, roaster lead time 8 weeks with payment on delivery, Costco paying 30 days after receipt, and a 3-week MVM in week 12 that doubles velocity.

• Initial fill: 2 pallets x 80 warehouses = 160 pallets = 46,080 units. At $13.15 that is a $605,952 opening PO.
• Steady weekly demand: 150 x 80 = 12,000 units, or about 42 pallets a week. Each warehouse turns its 2-pallet position in roughly 3.8 weeks, so replenishment of roughly 1 pallet per warehouse every other week.
• Production for fill plus 8 weeks: 46,080 + (8 x 12,000) = 142,080 units, about 493 pallets. At $8.50 landed, $1,207,680 of inventory paid to the roaster before the first delivery.
• MVM build: three weeks at 24,000 units per week means an extra 36,000 units (125 pallets) in position by week 11, so the production plan for weeks 9 through 11 roughly doubles.
• Cash: the opening PO pays out in week 12 (delivery week 8 plus 30 days). Weekly replenishment of about $158,000 gross follows on the same lag. The business is cash negative by roughly $1.2 million at its deepest point, in week 8, and cash positive on this program by about week 20 if velocity holds.
• Exit scenario: if velocity comes in at 90 units per warehouse per week instead of 150, the 2-pallet position lasts over 6 weeks, replenishment slows to a trickle, and the buyer may decide not to continue. Harbor Street would then hold roughly 40,000 units of finished club packs with no Costco home, about $340,000 of inventory to move through DTC, Amazon or a markdown.

That exit number, not the opening PO, is the figure to put in front of the founder before signing. The open-to-buy and PO cash commitment template is built to show it, and our guide to purchase order forecasting covers the PO-to-cash mechanics. For how Costco compares with other accounts, see the retail inventory forecasting hub.

What to confirm with the buyer

• Continuous or in-and-out, and if in-and-out, the exact window.
• Depot or direct-to-warehouse delivery, and the pallet configuration Costco wants.
• The velocity the buyer considers a success and the threshold for deletion.
• MVM and demo plans for the first six months, with funding expectations.
• Payment terms, delivery appointment rules and the chargeback schedule.

Where Drivepoint fits

Drivepoint is the AI finance platform for consumer brands, and Costco programs are the kind of single decision that can define a year. It connects Costco sales and PO data alongside Shopify, Amazon and your accounting system through 100+ integrations, turns warehouse-level movement into a pallet-aware forecast inside your Excel model, and carries every PO through the P&L, balance sheet and cash flow. AI scenario planning lets you run the 150-unit case, the 90-unit exit case and the MVM build side by side in minutes, so the conversation with the buyer starts from numbers you have already stress-tested.