Two Hundred Flips Across Amazon and eBay, The Honest Numbers

Retail arbitrage is the simplest flipping model on paper. Buy from a physical store at retail, sell online at a higher price, pocket the gap. In practice the gap is eaten by platform fees, return fraud, category gates, and your own time. I ran 200 flips across fourteen months to measure the real spread between Amazon FBA and eBay on exactly the same pool of items, and the result changed how I route every unit I buy.

The method, so the numbers are comparable

I sourced from the same five big box retailers in my city, scanning clearance endcaps and seasonal close outs. For each item I checked both the Amazon sell price and the eBay sold comps, and I split routing based on a coin flip when both channels looked viable. That kept the sample balanced enough to compare net margin honestly. Items that were only viable on one channel (gated brands, used condition, incomplete box) were logged separately and are excluded from the head to head.

Amazon FBA, seventeen percent net

Amazon wins on sell through. My average FBA listing cleared in 11 days against 34 days on eBay. That speed collapses the cash conversion cycle, which matters when you are rotating 400 euros weekly. It loses on fee load. Referral fee, FBA pick and pack, inbound shipping, and occasional long term storage on dogs combine to around 28 percent of gross on my average basket. That left me at 17 percent net margin on Amazon, with low variance because fees are predictable and returns are absorbed by the platform.

eBay, twenty four percent net but slower

eBay takes a lower fee bite, roughly 13 percent gross once Managed Payments and ad spend are included. I handle shipping myself, which adds labour but saves the FBA pick and pack cut. Returns are a bigger deal here because the platform sides with the buyer aggressively, and I ate three bad return claims in my first fifty flips. Even so, the net margin averaged 24 percent, seven points above Amazon. The trade off is sell through; 34 day average means capital is locked longer, which caps how fast you can scale. The same cash conversion dynamic shows up in the thrift to Vinted sourcing article.

The routing rule I use now

  1. Branded, new, boxed, ungated: Amazon FBA. The speed and the buyer trust in the Buy Box overcome the fee hit.
  2. Used, open box, incomplete box, or gated brand: eBay. The platform accepts the condition nuance without drama.
  3. High ticket (over 150 euros) with thin Amazon comps: eBay. The auction format surfaces buyers that FBA search does not.
  4. Low ticket (under 12 euros): neither. The fee floor kills the margin on both venues. I pass.

This rule has held across another hundred flips since I finalised it and is the closest thing I have to a repeatable process in physical flipping. It is also consistent with what I see from operators in adjacent corners like digital products, where matching the channel to the product format is the single biggest lever.

The traps I paid for

I got an Amazon account warning once for listing a brand that turned out to be on their silent review list. No suspension, but the listing was pulled and inventory had to be removed for disposal (35 euros disposal fee). I got a negative eBay feedback once for a unit I shipped to the wrong buyer in a batch packing session; tracking proved innocence but the feedback stayed up. Both were operational mistakes, not strategy mistakes, and both are why I now keep a written block list and double check packing slips.

John's rare tip

Price eBay listings at the Amazon Used Good price, not the New price

When the same unit exists on Amazon as a used offer at a lower price than the new offer, buyers searching across both platforms benchmark against the used price. I tested pricing ten identical units at Amazon new minus 5 percent on eBay; they sold in 48 days average. The next ten priced at Amazon used plus 10 percent sold in 19 days average and the net margin ended up nearly identical because of the capital rotation.

What I'd avoid

Do not scale into gated categories unless you have ungated approval in writing. I spent a weekend sourcing cosmetics at an 80 percent clearance, assumed first sale doctrine would cover me, and ended up with 430 euros of inventory I could not list on either channel because Amazon required brand authorisation and eBay flagged the batch during a category sweep. I sold it locally at break even. Test a single unit first, always.

Frequently asked

Which platform has higher net margin on retail arbitrage?

eBay, once FBA fees and long term storage are included. My measured average net margin was 17 percent on Amazon FBA against 24 percent on eBay, on the same pool of items priced within 5 percent of each other. Amazon wins on sell through speed, eBay wins on per unit profit.

Is retail arbitrage allowed on Amazon?

Yes in most categories, with brand gating catches. Ungated categories let you list sourced retail units freely, but a growing list of brands requires prior approval, and some brands file complaints even when the sale is legal under first sale doctrine. I keep a personal block list of brands that have triggered warnings.

How much capital do you need to start?

My first thirty flips were funded with 400 euros rotated across six weekends. The key constraint is not capital, it is cash conversion cycle; on Amazon FBA the payout lag is two weeks after sale, on eBay it is three to five days. If you have 400 euros you can start on eBay; on Amazon you want closer to a thousand to absorb the lag.