Shopify Into Mexico: Splitting Stock Across Two Warehouses Without Running Out

A Shopify store selling into Mexico from a single US warehouse has one lead time and one failure mode: every order pays a border crossing. A store with stock on both sides has two lead times, two cost profiles and a decision to make on every SKU. That decision is the subject of this article.

The mistake that costs the most is treating the Mexican warehouse as an overflow bin. Stock placed there has already paid duty, IVA and brokerage. It is the most expensive inventory in the business, and it should be the fastest moving.

The two lead times, measured rather than assumed

A shipment from a Laredo facility into Monterrey is a short drive and a customs event. The driving portion is roughly 3 hours. The crossing is what determines the day count: a shipment cleared on the green light arrives at the destination warehouse in 1 to 2 business days, and a physical inspection on the red light adds 1 to 2 days on top.

Local fulfillment from a Monterrey facility skips all of that. Same day pick and pack, domestic carrier, no pedimento. The lead time difference between the two paths is not marginal, and it is the reason the split exists.

What crossing the border does to a reorder point

A reorder point is demand during lead time plus safety stock. Both halves change when the replenishment path crosses a border.

Lead time stops being a single number and becomes a distribution with a tail. Most crossings clear on the green light, and the ones that do not add 1 to 2 days. Safety stock has to cover the tail, not the median, which is why a reorder point copied from a domestic operation runs the Mexican warehouse dry every time an inspection lands.

The second change is that replenishment is lumpy. Because the customs processing fee carries a fixed minimum of 258.91 pesos and brokerage starts at 3,500 MXN per shipment, small frequent replenishments are expensive. The economics push toward fewer, larger inbound movements, which raises the average holding and therefore the reorder point again.

Each transfer ends in a receiving event, and the standards that event is judged against are in Mexico Warehouse Receiving: Appointments, Labels and the ASN That Prevents a Rejection.

Which SKUs belong on the Mexican side

Four attributes decide it, and only one of them is sales volume.

• Velocity. Fast movers earn their place because the inventory turns before the holding cost accumulates.
• Return rate. High return SKUs benefit twice: faster delivery reduces returns, and a local return address recovers the ones that happen.
• Labeling status. A SKU whose NOM label file is complete can be received and made sellable immediately. One whose file is incomplete sits in the warehouse it cannot legally leave.
• Value density. High value low volume goods carry more capital across the border per pallet, which raises the cost of getting the forecast wrong.

Labeling is a receiving gate, not a marketing task

Product information has to be presented in Spanish, and the requirement is set in consumer protection law rather than in marketplace policy:

"Los datos que ostenten los productos o sus etiquetas, envases y empaques y la publicidad respectiva, tanto de manufactura nacional como de procedencia extranjera, se expresarán en idioma español y su precio en moneda nacional."

Ley Federal de Protección al Consumidor

The operational consequence is specific. A unit that arrives at the Mexican warehouse without a compliant label is not sellable stock, it is work in progress. If the facility can relabel, the gap is an hour of handling. If it cannot, the unit is stranded and the reorder point that assumed it was available was wrong.

Modeling the split, with an example

Take a SKU selling 400 units a month into Mexico, with a 6 percent return rate and a complete NOM file.

The line that usually decides the case is the second one. Consolidating 240 units into 1 or 2 crossings collapses the per shipment charges from a variable cost into a fixed one, and those charges do not scale with units.

Keeping the two piles honest

Two warehouses create one new failure mode: selling stock that is physically in the wrong country. Shopify will happily accept an order that your fulfillment logic cannot serve within its promise. Three controls prevent it.

• Publish a single availability number to the storefront, not a sum of two locations.
• Route on the buyer's postal code before the order is confirmed, not after.
• Reserve a buffer on the Mexican side that is never published, sized to the crossing tail rather than to the median.

What to review monthly

The split is not a one time decision. Four numbers tell you whether it is still right: units shipped from each side, stockout hours by SKU and location, average crossing time including inspections, and returns recovered locally as a share of returns received. When the third number moves, the reorder point has to move with it.

The inbound movement is a customs event with a name

Moving stock from the US facility to the Mexican one is not an internal transfer. It is an import, and the law is explicit that clearance is a defined set of acts and formalities rather than a shipping step:

"Para efectos de esta Ley, se entiende por despacho aduanero el conjunto de actos y formalidades relativos a la entrada de mercancías al territorio nacional y a su salida del mismo."

Ley Aduanera, Article 35

Three consequences follow for inventory planning. The declaration has to be filed by a licensed broker holding a patente, which means the paperwork has a queue and the queue has hours. The customs value is converted at the official rate for the entry date, so a replenishment budgeted at 18 pesos to the dollar can land at 19. And the goods are unavailable for sale during clearance, which is time that belongs in the lead time calculation rather than outside it.

What the fixed charges do to replenishment size

The customs processing fee is 8 per thousand of customs value with a floor of 258.91 pesos, and that floor applies below a customs value of roughly 32,364 pesos. Brokerage runs 3,500 to 15,000 MXN per shipment. Neither charge scales with units.

Replenishing 240 units in 4 shipments of 60 pays those charges 4 times. Replenishing the same 240 units in 1 shipment pays them once. On a 3,500 MXN minimum brokerage alone that is 10,500 MXN of avoidable cost per cycle, which on a 400 unit a month SKU is roughly 26 pesos per unit that buys nothing.

This is the reason the classical economic order quantity understates the right batch size on a cross border replenishment. The ordering cost is not a clerk's time, it is a customs entry.

Sizing the buffer to the tail, with numbers

Take the same 400 unit a month SKU, which is roughly 13 units a day. A green light crossing puts stock on the shelf in 1 to 2 business days. A red light adds 1 to 2 days on top, so the worst realistic case is 4 days.

A buffer sized to the median consumes 26 units and runs dry every time an inspection lands. A buffer sized to the tail holds 52. The difference is 26 units of holding cost against the certainty of not going out of stock on the days when the crossing is slow, and out of stock days on a fast mover cost more than 26 units of storage almost every time.

Why the tail is not rare enough to ignore

Inspection outcomes are assigned per entry. A store replenishing once a month faces the draw 12 times a year. A store replenishing weekly faces it 52 times. Consolidating replenishment reduces the fixed charges and it also reduces the number of times the tail can hurt you, which is a second argument for the same decision.

Routing rules that survive a bad day

Order routing between two warehouses is usually written as a preference: serve from Mexico when stock exists, otherwise serve from the US. That rule is correct on a good day and expensive on a bad one, because it sends the order to whichever pile has a unit rather than to whichever pile can meet the promise.

A better rule evaluates three conditions in order. Is there sellable, correctly labeled stock on the Mexican side. Does the buyer's postal code fall inside the domestic carrier's committed coverage. Does the remaining Mexican quantity stay above the reserved buffer after this order. Only when all 3 are true does the order route locally. Otherwise it routes to the US warehouse and the storefront shows the longer promise before checkout, not after.

The third condition is the one that gets skipped. Without it, a single large order can drain the buffer that exists precisely to absorb a slow crossing, and the store discovers it 4 days later.

What breaks when the catalog grows

At 20 SKUs the split can be maintained by judgment. At 200 it cannot, because the four attributes that decide placement move independently: velocity shifts with season, return rates shift with a listing change, NOM files get completed one at a time, and value density changes when a supplier moves.

The practical answer is to score placement rather than to decide it. Each SKU carries a placement score built from those 4 attributes, the score is recalculated monthly against the previous 90 days of orders, and the warehouse team works a change list rather than a full re plan. Most months the list is short. The months it is long are the months a season turned, and that is exactly when a fixed placement would have been wrong.

The number that tells you the split is working

One metric summarizes the whole system: the share of Mexican orders fulfilled domestically within the promised window. It moves when placement is wrong, when the buffer is too thin, when labeling stalls at receiving, and when a crossing runs slow. Because all 4 failure modes push the same number in the same direction, it is the right thing to put on a dashboard, and the 4 diagnostics behind it are the right things to look at when it falls.

The provider side of this decision, meaning what a single view across both warehouses has to reconcile, is covered in Binational Inventory: What One Dashboard Across Two Warehouses Has to Reconcile.

How BringGo Ship Handles This

BringGo Ship operates both sides of the split: a Laredo facility for consolidation and prep, and Monterrey for domestic fulfillment inside Mexico. Our team sizes the buffer to the crossing tail rather than the median, publishes a single availability figure to the storefront, and holds a reserve on the Mexican side that never reaches the sales channel. Placement is reviewed monthly against the previous ninety days, so a seasonal shift moves stock before it becomes a stockout.

Frequently asked questions

Do I need a Mexican entity to hold stock in Mexico? Holding stock and acting as importer of record are separate questions. The second one has requirements worth settling before the first shipment moves.

Can I move stock back to the US if the split is wrong? Yes, but it is a fresh import into the United States with its own declaration. Getting the split wrong is more expensive to undo than to prevent.

How much should sit on the Mexican side? Start with the fast movers whose NOM files are complete, size the buffer to the crossing tail, and review monthly. A fixed percentage applied to the whole catalog ignores the four attributes that actually decide it.