Amazon's presence as a seller is a competitive signal, not an automatic ban nor proof of future losses. Confirm the exact ASIN and offer, record who is selling, who is shipping, the price, the condition, the featured offer, and the timestamp. Repeat the observation at an authorized cadence, then test your costs at several prices without assuming Amazon will stay in stock or change its price. Cross-reference this signal with the number of offers, your restrictions, your minimum order quantity, and your breakeven point. If a small price drop is enough to wipe out the margin, reduce the batch size or reject the purchase, regardless of who the current seller is.
What Amazon's presence actually proves
A product page can group several sellers under a single listing. Amazon explains on its official pages that several sellers can offer the same product and that a featured offer appears in the main buy box area. Observing Amazon as a seller only proves that an offer attributed to that entity was visible in the context and at the moment observed.
This observation does not prove stock depth, how long the seller has been present, the pricing strategy, or the number of sales. Nor does it mean another seller can never win the featured offer. The conditions and criteria belong to Amazon and must be checked in current official resources.
Before interpreting the signal, verify the packaging, the variant, the condition, and the marketplace. An Amazon offer on a two-pack is not direct competition for a single unit. A matching error turns the entire analysis into a false positive.
Building a traceable observation
| Field | Example status | Limit of the observation |
|---|---|---|
| ASIN, marketplace, condition | confirmed | does not yet prove the physical variant |
| displayed seller | observed | may change after collection |
| displayed shipper | observed | distinct from the seller |
| price and delivery | observed and dated | not a completed sale |
| featured offer | observed | may change |
| other visible offers | counted per context | visibility may be partial |
| displayed availability | raw value | does not prove an exact quantity |
| evidence | authorized redacted screenshot or response | must contain no account data |

This public screenshot only attests to the redacted editorial asset for the indicated route. It proves neither authenticated data nor the existence of the filter described below.
Four-step evaluation procedure
1. Confirm the product and the rights
Validate the ASIN, GTIN, brand, model, size, color, unit count, and condition. Also check that your account is authorized to sell the brand or category before interpreting the competition. The guide on checking brand restrictions before buying covers this step.
2. Observe without extrapolating
Record the offer at a precise time. Follow access rules and use the account's official tools. If you repeat the observation, keep each snapshot instead of overwriting the previous one. A one-time absence is "not observed," not "Amazon has left the ASIN."
3. Test the economic scenarios
Calculate your full cost and breakeven point, then test several lower prices. The Amazon price-drop simulation must recalculate commissions and other price-sensitive costs. Add a scenario where Amazon stays a seller for the whole period, and a scenario where the offer changes, without assigning an invented probability.
4. Decide on the size of the risk
Cross-reference margin, minimum order quantity, storage, tied-up capital, restrictions, and ability to exit. The decision can be: refuse, order a canary batch, reduce the quantity, or accept with monitoring. Document the reasoning; don't use an unexplained color code.
Distinguishing four situations
Amazon's presence can take several operational forms:
- Amazon is both the seller and shipper of the observed offer;
- a third party sells and Amazon ships via FBA;
- Amazon is a seller, but another offer is featured at the moment observed;
- Amazon is not observed as a seller, which does not prove a lasting absence.
These situations must not be merged under amazon_present = true. Seller, shipper, and featured offer are three distinct fields. The source, date, and marketplace complete the signal.
An example decision matrix without an invented score
A team can use a qualitative decision matrix instead of an opaque score:
| Observation | Margin under tested drop | Minimum order quantity | Treatment |
|---|---|---|---|
| Amazon observed, fragile margin | negative | high | block or negotiate |
| Amazon observed, resilient margin | positive | low | canary batch and monitoring |
| Amazon not observed, restrictions unknown | irrelevant | irrelevant | verify authorization |
| uncertain seller or ambiguous variant | not calculable | irrelevant | matching review |
The terms "fragile," "resilient," and "low" must be defined in the seller's own policy. The table provides no universal rule and does not promise a canary batch will be profitable.
Cross-referencing other signals without conflating them
The BSR is a ranking, not a sales figure. The upcoming article on interpreting Amazon's BSR complements the observation, but does not turn Amazon's presence into a reliable demand estimate. Likewise, a large number of offers does not prove their stock or quality.
Favor a listing where each signal keeps its own definition. The price answers "how much is shown?" The seller answers "who is offering this listing?" The featured offer answers "which offer occupies that spot right now?" Sales reports answer questions about your own account's activity. None of these fields should be used in place of another.
What ArbitragePro+ can automate / what the seller must verify
The filtre-amazon-vendeur function is a proposed specification, not a confirmed feature. It could record seller, shipper, featured offer, and date, keep a history, and trigger a review when the identity changes. It should show unknowns and never predict Amazon's strategy.
The seller must confirm the ASIN, their authorization to sell, the offers in their account, the costs, and the scenarios. They choose the quantity, verify the authorized collection cadence, and own the decision. Amazon's presence remains a signal, not an automatic verdict.
Checklist before ordering
- ASIN, variant, batch, and condition are confirmed.
- Seller and shipper are recorded separately.
- The featured offer carries a date and a marketplace.
- One-time absences remain "not observed."
- The account's restrictions are checked.
- Several prices are tested with recalculated fees.
- Minimum order quantity and exit capacity are assessed.
- The BSR is not converted into invented sales figures.
- The decision explains its evidence and its unknowns.
Official sources
- Amazon France, Amazon Seller app and definition of the featured offer: https://sell.amazon.fr/programmes/amazon-seller-app — accessed 2026-08-17.
- Amazon France, advertising and competition between multiple sellers for the featured offer: https://sell.amazon.fr/en/publicite?mons_sel_locale=en_GB — accessed 2026-08-17.
- Amazon France, Seller University: https://sell.amazon.fr/vendre-sur-internet/seller-university — accessed 2026-08-17.
