I walked into Mercado Libre with my Amazon playbook under my arm, and South America handed me the bill in year one. I could run ads. I could write listings. What I could not do, apparently, was run the warehouse side of the business: overstock tied up my cash, understock killed my momentum, and returns piled up in a corner until I cleared them at a loss months later. It took me a while to admit the obvious — if the warehouse work is wrong, every bit of front-end selling is wasted effort.

Here is the mistake that taught me. Before Black Friday 2024, I pushed more than 1,200 units of home goods into a warehouse near São Paulo, stocked for 45 days in one shot, Amazon-style. The logic felt safe: never stock out in peak season. The result was not safe at all. One of the three SKUs took off. The other two moved at a snail's pace, and after peak season the leftovers sat in that warehouse for over four months. Storage fees, tied-up capital, and a clearance markdown later, the whole profit from that push was gone, plus some. Platform rules and fee schedules change regularly, so always check the platform's official pages for the latest terms — I will not quote numbers I cannot stand behind. But one thing does not change: in South America, the cost of sitting inventory is bigger than you think.

Inbound: Don't Copy Your Amazon Playbook to Mercado Libre

The Amazon way is to stock deep and let a big warehouse absorb volatility. South America punishes that. First, your replenishment cycle from China is already long, so one wrong bet locks you in for months. Second, buying rhythms and return habits across South American markets do not behave like North America, so your bestseller forecast is shakier than you think. Third, the storage cost structure means the longer goods sit, the more they bleed.

My rule now is small batches, more often. The first batch covers about 20 days. I watch one week of real sell-through, then replenish against actual movement, not against my forecast. Yes, it means more inbound shipments and more receiving work. The math still wins:

Approach Cash tied up Stockout risk Cost of slow movers
One 45-day stock-up High, one bet sits for months Low High, clearance plus storage
Rolling 20-day batches Low, cash keeps moving Medium, covered by safety stock Low, a bad bet is only a small batch

One more detail, and it is boring until it bites you: check packaging and labels before the goods leave China. If a barcode does not scan or a carton label does not match the system record, your stock can sit in the warehouse unreceivable. I lost a full promotional week that way once — 300 units stuck because the outer labels did not match the listing data, and the promo ran with nothing to sell.

Delivery Speed on Mercado Libre Is Won Inside the Warehouse

Most sellers treat delivery speed as a carrier problem: faster linehaul, faster delivery. My experience says the opposite. For orders already sitting in a South American warehouse, the bottleneck is usually inside the building. Picking, checking, packing, and dispatch drag, and the buyer still sees a slow order.

In March 2025, a cluster of orders bound for the Buenos Aires area doubled our daily volume for a few days. When we reviewed it, the slow part was not transport at all. Our hottest SKUs were buried deep in the racking, and pickers were walking extra distance on every single order. We moved the top 20 selling SKUs next to the packing benches and switched final checks from inspecting every order to sampling plus full checks on high-risk SKUs. Dispatch speed stabilized visibly, and it cost nothing — just a layout change and a rule change. Delivery promises shown on the platform are the platform's latest wording, so I go by that; what a seller actually controls is always these few steps inside the warehouse.

So my speed checklist is embarrassingly simple, and it works: bestsellers live next to packing; every morning we look at the day's order waves before sequencing picks; a week before any promotion, likely hot SKUs get moved to easy-pick locations in advance.

Returns: Where the Quiet Losses Live

Every South America seller complains about returns, but most only watch the return rate and ignore what happens after the parcel comes back. Here is my counterintuitive take: processing returns fast matters less than judging them right. A returned item needs three decisions as quickly as possible: can it be resold, does it need repair or repackaging, or is it scrap? Delay the decision by a week and you pay another week of storage. Get it wrong the other way — reshipping a defective unit — and you buy a bad review plus a second return, which costs far more.

In June 2025 I worked through a batch of more than 40 returned small appliances. The old habit was to stack everything in the returns corner and deal with it when someone had time. The new rule: every return gets inspected and sorted within 48 hours of arrival. Resellable units go back to good stock, units needing new packaging go to a rework zone, scrap goes straight through the disposal process. That one rule lifted the share of returns we could resell in that month, and the returns corner stopped growing like a landfill. I will not quote a return rate — it varies wildly by category and season, so trust your own store data — but the sorting step itself always pays.

One more trap: do not judge returns with the same checklist you use for new stock. A new-goods standard applied to returns will condemn plenty of units that could still sell; too loose a standard ships problems to the next buyer. I run two checklists now. The returns one grades on function and appearance, A or B grade. B grade gets cleared at a discount. A grade goes back to good stock.

The Checklist I Actually Use

Before inbound: verify labels and barcodes at origin, first batch sized for about 20 days, safety stock yes, gambling stock no. Daily: bestsellers by the packing bench, order waves reviewed before picking, pre-promotion moves done a week early. Returns: inspect and sort within 48 hours, A/B grading, resellable units never left sleeping in the returns corner. Platform terms and fees are whatever the official pages say today — this checklist only covers the warehouse side.

Ads can buy traffic in South America. Only a warehouse that does not leak turns that traffic into profit. This one I tested myself, and it works.

By Ah K, E-commerce Seller (5 yrs)