How Vendor Central's relationship fundamentally differs from Seller Central
If you've only ever sold through Seller Central, the vendor relationship looks deceptively similar from the outside, you're still getting products onto Amazon, still dealing with Amazon's systems and processes. The underlying legal and financial relationship is a completely different animal. On Seller Central, you're a third-party seller, you own your inventory the entire time it sits in an FBA warehouse, and Amazon is simply providing the marketplace and fulfillment infrastructure you're paying to use.
On Vendor Central, Amazon is the direct buyer of record. You sell wholesale, Amazon issues a purchase order, and once that PO is fulfilled, Amazon owns the inventory outright, the same way any retailer owns the stock on its shelves after buying it from a supplier. That structure sounds like it should be safer for a vendor, you got paid, it's Amazon's inventory risk now. And under normal conditions, it largely is. The problem shows up specifically around cancellation, because a cancelled PO isn't just a missed sales opportunity the way an unsold FBA listing might be. If the PO is cancelled after you've already manufactured the goods against it, you're left holding fully produced, already-paid-for-in-labor-and-materials inventory with no buyer at all, and no marketplace of your own to sell it through instead.
This is worth sitting with for a second, because a lot of manufacturers and brands actively pursue a Vendor Central relationship because it looks more stable than selling directly. No advertising to run, no listing optimization to manage, no customer service queue, just fulfill purchase orders and get paid. That simplicity is real, and for a long stretch of a normal year it delivers exactly what it promises. The risk that simplicity is quietly trading away is control over what happens when Amazon's own economics on your product change unexpectedly, and a PO cancellation is where that trade becomes visible.
Why tariff shocks specifically expose vendors more than 3P sellers
A third-party seller absorbs a tariff increase as a landed cost problem, their margin gets squeezed, they adjust price, it's unpleasant but it's their own cost structure to manage. A vendor relationship puts Amazon itself on the hook for that same cost increase, because Amazon is the one that purchased the inventory wholesale and now owns the exposure of reselling it at a profit.
That changes Amazon's incentives in a way that matters enormously when trade policy shifts suddenly. When Amazon's own landed cost on wholesale-purchased inventory spikes unexpectedly due to a new tariff, Amazon has a direct and immediate financial reason to cancel outstanding purchase orders rather than honor them at the old economics, especially POs placed before the tariff took effect but not yet fulfilled. A 3P seller's inventory relationship doesn't create that same pressure point for Amazon, because Amazon never owned that inventory or its cost basis in the first place. Vendors are structurally more exposed to exactly this kind of shock, precisely because the vendor relationship that normally feels more stable, guaranteed wholesale orders, is the same relationship that puts Amazon's own margin directly at risk when costs move fast.
It also helps to understand the scale Amazon is operating at when it makes a decision like this. A retailer managing purchase commitments across an enormous, category-spanning catalog isn't evaluating your specific PO in isolation, it's running a portfolio-level response to a cost shock across thousands of vendor relationships simultaneously. From Amazon's side, cancelling a batch of POs tied to a newly unprofitable category is a single policy decision applied broadly. From the vendor's side, on the receiving end of just one of those cancellations, it's a specific, disruptive, and often financially serious event. That mismatch in scale is a big part of why these cancellations land with so little individualized warning or explanation.
A documented case: cancelled after the chairs were already made
During the April 2025 tariff escalation, following the announcement of expanded tariffs covering more than 180 countries and territories, including China, Vietnam, and Thailand, Amazon cancelled purchase orders across a range of merchandise sourced from China and other Asian countries, beach chairs, scooters, and air conditioners among them, affecting multiple 1P vendors.
One documented case involved a vendor that had been supplying beach chairs made in China to Amazon for more than a decade, a long-standing relationship, not a new or speculative one. That vendor received notice that Amazon was cancelling purchase orders it described as having been placed "in error," cancelling a $500,000 wholesale order after the chairs had already been fully manufactured. The vendor was left holding that entire run of finished inventory with no buyer, no advance warning before the cancellation notice arrived, and no meaningful recourse built into the relationship to recover the cost of goods already produced against a confirmed order.
A former Amazon vendor manager corroborated that this wasn't an isolated incident specific to one product category or one unlucky vendor. He reported that the same pattern, sudden, unexplained PO cancellations tied directly to tariff timing, affected several of his own former clients during the same period. That corroboration matters, it means the beach chair case wasn't a fluke or a one-off internal mistake, it reflects a broader operational response Amazon applied across its vendor base when its own cost exposure shifted fast.
Framing a cancellation as an error rather than a deliberate business decision affects how a vendor can reasonably respond. If a vendor accepts that framing without pushing back, they lose the ability to argue the PO was valid and binding when they made their production decision. Treat that language as a starting negotiating position from Amazon, not a neutral factual description of what happened.
Contract and PO practices that reduce your exposure
None of this is fully preventable, Amazon holds the structural leverage in a wholesale relationship of this size, but a few practices meaningfully reduce how exposed a vendor is when conditions shift suddenly.
- Actually read the cancellation clauses in your vendor agreement. Most vendors sign these agreements once at onboarding and never revisit them. Understand exactly what circumstances allow Amazon to cancel a confirmed PO, and whether there's any distinction in the contract language between cancellation before production starts versus after.
- Time production starts relative to PO confirmation, not just PO receipt. Receiving a PO and having it fully confirmed with no contingencies aren't always the same event depending on how your specific agreement is structured. Understand precisely what triggers your production commitment, and don't start manufacturing runs against anything less firm than that trigger, especially during periods of known trade policy volatility.
- Diversify between 1P and 3P models for the same catalog where you can. Running a portion of your catalog through Seller Central alongside your Vendor Central relationship means a single channel disruption, whether a PO cancellation wave or a vendor negotiation breakdown, doesn't strand all of your inventory and all of your Amazon revenue simultaneously.
- Avoid over-concentrating production timing around known policy risk windows. When trade policy changes are actively being discussed or are pending implementation, that's a bad time to be starting large production runs against POs that haven't shipped or been formally locked in yet.
When you actually sit down and read the cancellation clause, look for a few specific things rather than skimming past the boilerplate. Check whether the agreement uses "cancellation for convenience" language, which typically means Amazon can cancel for any reason with minimal obligation to you, versus a "cancellation for cause" structure, which limits cancellation to specific breaches and generally gives you more standing to dispute an improper one. Look for a defined notice period, whether Amazon owes you a set number of days before a cancellation takes effect, and whether that period is actually honored in practice. And look specifically for language addressing goods already in production, some agreements draw a real line entitling the vendor to reimbursement once production has started against a confirmed order, while others are silent on it entirely, which in practice tends to favor Amazon's position in a dispute rather than yours.
The exposure gap between production timing decisions is bigger than it looks on paper. A vendor who starts cutting materials or booking a production line the moment a PO lands in their inbox, before it's actually confirmed with a firm ship window and no outstanding contingencies, is carrying open-ended risk with essentially nothing to point to if that PO gets pulled, because nothing was ever finalized on Amazon's side either. A vendor who waits until the PO shows as fully confirmed before committing factory capacity is trading a few days or weeks of lead time for a meaningfully stronger position if a cancellation happens anyway, because at that point they're disputing a cancellation of a confirmed order rather than something that was still provisional when they acted on it. That gap in timing is often the difference between having a real case to escalate and simply absorbing the loss.
Purchase orders, factory production schedules, deposit payments, and shipping bookings tied to a specific PO number are exactly the evidence that supports an escalation if a cancellation happens after the fact. Don't wait until you're disputing a cancellation to assemble this, keep it organized as production happens.
How to respond when a PO is cancelled after production has started
If you're in this situation, move fast and document everything rather than assuming a phone call will resolve it. Compile the full paper trail, the original PO with its terms, proof of when production started relative to that PO's confirmation, factory invoices or deposit receipts tied specifically to that order, and any correspondence confirming the order before the cancellation notice arrived.
Escalate through your actual vendor manager relationship rather than a generic support channel, vendor accounts of any real size have a named point of contact, and that relationship is where a case like this gets a real hearing rather than a form response. Frame the escalation around the documented financial commitment you made in direct reliance on the confirmed PO, not just frustration about the outcome. If the vendor manager relationship doesn't produce movement, escalate further within Amazon's vendor management structure, and don't be afraid to make clear that continued sourcing commitment on your end is contingent on some resolution, that's real leverage if you're a vendor Amazon has genuine reasons to want to keep.
It's also worth looping in legal counsel early rather than treating this purely as a relationship problem to smooth over. Whether a cancelled PO creates an enforceable claim depends entirely on the specific language in your vendor agreement, what constitutes a binding order, what cancellation rights Amazon reserved, and whether "placed in error" has any defined meaning in that agreement or is simply a phrase being used informally. A short conversation with counsel who's reviewed the actual contract terms tells you within a day or two whether you have real recourse worth pursuing formally, or whether your best path really is relationship-based escalation and absorbing the loss while you rebuild the relationship going forward.
The broader question this raises about vendor leverage
Step back from any single cancellation and there's a harder strategic question every vendor relying heavily on Amazon as a primary wholesale buyer eventually has to sit with. Amazon is simultaneously your largest customer and the counterparty that can unilaterally cancel a confirmed order with limited practical recourse available to you. That's a genuinely uncomfortable position to be in for any business, and it's structurally different from a typical buyer-supplier relationship where both sides have roughly comparable ability to walk away or enforce terms.
The honest answer is that an individual vendor, especially a mid-sized one, has limited leverage in the moment a cancellation happens. The leverage that does exist gets built well before that moment, through diversified sales channels so Amazon isn't your only outlet, through contract terms negotiated with real attention rather than accepted as boilerplate, and through relationships with vendor managers who know your business well enough to advocate for you internally when something goes wrong. Vendors who treat the Amazon wholesale relationship as inherently stable because it's been reliable for years are the ones most exposed when a shock like a sudden tariff shift reveals how much of that stability was actually just favorable conditions, not a guarantee.
None of this is an argument against the Vendor Central model itself, plenty of brands build genuinely healthy, profitable businesses on it for years without incident. It's an argument for going into that relationship with clear eyes about what kind of counterparty Amazon actually is in a wholesale arrangement, a buyer large enough that its own cost pressures can, and periodically will, override the assumption that a confirmed order is a guaranteed one. Build your production timing, your contract review, and your channel mix around that reality rather than around how the relationship has behaved in the calmer years, and a shock like the 2025 tariff cancellations becomes a serious problem to manage instead of an existential one.
