Amazon Vendor vs. Seller Central: The Real Cost Difference

Move $2 million of Cinderhaven Provisions' volume onto Amazon and the model you choose changes what you keep by $145,000 a year. Seller Central retains about $1 million of that $2M. Vendor Central retains about $855,000. The gap is real. It is also the wrong number to decide on.

Amazon offers a specialty food brand two doors. Vendor Central (1P) makes Amazon the buyer: it sends purchase orders, you ship wholesale, Amazon owns the price and the fulfillment. Seller Central (3P) makes Amazon the landlord: you stay the seller of record, you set the price, and you pay rent in fees. The economics diverge from there, and so does the risk.

Cinderhaven Provisions is a fictional company and its figures are a synthetic dataset; Amazon's referral and co-op rates are real and public.

Vendor Central pays less per unit and calls it simpler

Under 1P, Amazon buys at a wholesale discount of 45 to 65% off retail. On $2M of retail-equivalent demand, that is roughly $1M in the door. Then the allowances start. Co-op marketing, damage allowance, freight allowance, and AVS run 3 to 8% of shipped revenue as a structural floor, with co-op climbing toward the top of that band. Call it $120K.

On top sit chargebacks. Amazon assesses a carton content accuracy chargeback when the units in a box do not match the Advance Ship Notice: shortage, overage, case pack defect. If that sounds familiar, it is the same fee schedule Walmart runs, now wearing Amazon's codes. Net the disputable share against a thin recovery rate and Cinderhaven carries about $25K. Retained before COGS: $855K.

What 1P buys is the absence of an operation. Amazon forecasts, warehouses, ships, and handles the returns. What it costs is the price. Cinderhaven no longer sets it, and Amazon reprices against competitors without asking.

Seller Central keeps the margin and hands you the operation

Under 3P, Cinderhaven stays the seller and collects the full $2M in retail, then pays rent.

| Line (on $2M retail-equivalent) | Vendor Central (1P) | Seller Central (3P) | |---|---|---| | What Amazon pays you | $1.00M wholesale | $2.00M retail | | Referral fee (~12%) | not charged | -$240K | | FBA fulfillment, storage, surcharge | Amazon ships | -$560K | | Advertising to stay visible | optional | -$200K | | Co-op, MDF, damage, freight | -$120K | not charged | | ASN, carton & shortage chargebacks | -$25K | listing-suppression risk | | Retained before COGS | ~$855K | ~$1.00M |

Grocery carries an 8% referral fee under $15 and 15% over; blended across a specialty catalog, call it 12%. FBA fulfillment, storage, and the 2026 fuel surcharge take the larger bite; on grocery, referral and FBA together run near 40% of retail before a dollar of advertising. Add the ad spend it takes to stay visible and Cinderhaven retains about $1M. The retained line is only the top of the margin waterfall; COGS, returns, and storage still come out below it.

That is $145K more than 1P returns on the same volume. The catch is in the word "operation." The extra margin is not found money. It is the wage for running demand planning, listings, advertising, returns, and the working capital tied up in FBA inventory. A brand that cannot run that operation well gives the $145K back in stockouts, suppressed listings, and ad waste.

Both models bill you for the same wrong fields

The pitch treats 1P and 3P as a margin choice. Underneath, they are the same data problem in two costumes.

Under 1P, the chargeback line is built from the item file: a case pack, a GTIN, an ASN quantity. Wrong field, recurring fee, exactly as it works at Walmart. Under 3P, the same item file drives a different failure. A GTIN that does not match the GS1 registry, a title that trips a category filter, a missing attribute, and Amazon suppresses the listing. A suppressed listing sells nothing, which is a 100% margin loss that never appears as a fee.

Neither model rewards a clean catalog. Both punish a dirty one. The brand that ships Amazon the same product master it ships Walmart pays on whichever side it chose.

The decision is which costs you can control

For a specialty food brand, the honest framing is not which model prints the higher number. It is which costs the brand can govern. Under 1P, the brand controls its data and little else; Amazon owns the price, the forecast, and the fulfillment. Under 3P, the brand controls the price, the assortment, and the ad spend, and owns every operating cost that comes with them.

Amazon has been narrowing the choice on its own, culling 1P invitations for vendors under roughly $5M to $10M in Amazon sales and steering the marketplace toward 3P. For most brands scaling into the channel, the door is chosen for them; the work is running 3P without handing back the margin that makes it worth more. That work starts where the chargebacks do, in the product master that feeds every channel. Clean that, and the door you walk through matters less than the catalog you carry in.

Send me your Amazon fee report

Send me one month of your Amazon settlement or vendor remittance, whichever door you are in. I will rebuild the retained margin line by line, show you which fees are structural and which are a data defect wearing a fee code, and tell you what the other door would have kept. Start here.