Chargement de la page…
Amazon price history describes past observations; it does not predict the next price, the volume that will sell, or your ability to win the offer. Use it to spot pricing regimes, apparent breaks, promotional episodes and data freshness. Always verify the ASIN, the product condition, the offer type, the store and the period shown. A flat line can hide the absence of any offer, while an isolated spike can come from an unusual seller. The right decision combines the history with current fees, landed cost, observed competition, account restrictions and a downside scenario. No chart turns an opportunity into guaranteed revenue.
What a chart actually lets you observe
Keepa officially presents Amazon price history charts and price-drop alerts. This definition matters: the tool stores and displays observations. A chart therefore lets you answer descriptive questions. Did the visible price stay in a band for several weeks? Are there short, repeated changes? Is the data continuous or interrupted? Is the current value close to recent observations or an outlier?
These questions help characterize the context of an analysis. On their own they do not reveal the number of units sold, the stock held by every seller, the cause of a variation, or the future price. A period with no data point can mean a series was not observed or that no matching offer was available; it should not automatically be read as a confirmed sales gap.
The first check is about identity. A neighboring ASIN, a size variant or a refurbished product can have a different history. The second is about the series chosen: Amazon price, new-offer price, used, or another indicator are not interchangeable. The third is about the national store. A history on Amazon.de does not directly describe Amazon.fr, even if the product shares the same GTIN.
Reading signals without inventing their cause
| Observed signal | Cautious reading | Conclusion not allowed without other evidence |
|---|---|---|
| long plateau | price band often observed over the period | guaranteed price tomorrow |
| brief drop | one-off episode to investigate | new market normal |
| rise during a data gap | rare observation or a different available offer | confirmed shortage |
| regular alternation | recurring change in context or offer | proven seasonality |
| isolated current price | value different from recent history | durable margin |
An explanation becomes credible only if independent evidence supports it: known calendar event, range-wide event, change in the number of offers, fee change, or stock evidence. Otherwise, write "variation observed" rather than "rising demand." This discipline avoids confusing visual correlation with causation.
Eight-step procedure
1. Confirm the identifier. Compare ASIN, brand, model, variant, lot quantity and condition. 2. Fix the store. Note the national domain and the currency of the series. 3. Name the chart. Record the offer type or indicator actually shown. 4. Choose several horizons. Read a recent window and a longer window without mixing their scales. 5. Mark the gaps. Distinguish observed points from missing segments. 6. Break into regimes. Identify plateaus, short episodes and extreme values without assigning a cause. 7. Recalculate profitability. Apply current fees and landed cost to several representative prices. 8. Check Seller Central. Before placing an order, check authorization, compliance, listing and logistics for your account.
The low price observed over a long period deserves as much attention as the current high price. Calculating the Amazon break-even price lets you test whether the model survives a return to that low band. The ROI or net margin comparison then clarifies capital use, without adding predictive power to the chart.
If the purchase or resale spans multiple marketplaces, the method in comparing eBay and Amazon with shipping costs lets you redo the total cost channel by channel. An Amazon chart represents neither eBay's fees nor its liquidity.
A methodological example, with no market data
Imagine an entirely fictional, teaching-only series: a price band around 30 monetary units for eight weeks, a dip around 24 for four days, then a current observation at 36. These numbers belong to no ASIN and represent no real rate. They only serve to build three scenarios: low at 24, central at 30, and current at 36.
The low scenario measures resilience. The central scenario describes a band frequently observed in the example. The current scenario tests the apparent opportunity, but must not be assigned an invented probability. For each scenario, use the same tax base and the fees matching the price and the logistics mode. If the fictional landed cost is 20 and the other hypothetical costs are 6, the result before out-of-scope costs is respectively −2, 4 and 10. The calculation is deliberately simplified and does not reproduce actual Amazon fees: it only shows why a conclusion depends on the price tested.
<video controls muted playsinline aria-label="Silent animation summarizing Amazon price history in four steps"> <source src="/blog/articles/historique-prix-amazon-limites/video.mp4" type="video/mp4"> </video>
The animation summarizes four steps — identify, qualify the series, build scenarios, verify — without constituting a demonstration of an authenticated feature. It contains no account data and no forecast.
The limits to record in the purchase file
An analysis note should mention the date consulted, the period covered, the gaps, the series and the store. It should also distinguish a displayed price from a completed sale. Even when a tool provides statistics, the definition of each aggregate and the calculation window must stay visible.
Amazon states that selling costs depend in particular on the plan, the category and the shipping strategy. An old price history therefore does not necessarily contain current fees. Recalculate them in Amazon's tools at decision time. The Seller Central checklist before ordering serves as the final barrier against an analysis that looks appealing but is not executable.
What ArbitragePro+ could automate / what the seller must verify
The analyse-historique (history analysis) feature is a proposed specification, not a feature presented as currently available.
| Automation proposed for ArbitragePro+ | Verification left to the seller |
|---|---|
| show several horizons and flag gaps | confirm the ASIN, variant, condition and store |
| detect price bands without naming their cause | interpret the context with independent evidence |
| generate low, central and high scenarios | validate the prices used and current fees |
| date every observation and its source | check changes in Seller Central |
| attach assumptions to the profitability calculation | decide whether the uncertainty is acceptable before buying |
The proposed automation would make the reading reproducible. It could not guarantee a sale, predict a price, infer a certain cause, or validate the seller's rights to an offer.
Reading checklist
- ASIN, variant, condition and packaging match the product purchased.
- The store, currency and series are named.
- Gaps are not turned into confirmed sales or stockouts.
- Several horizons and several scenarios are examined.
- Amazon fees are recalculated at decision date.
- Any assumed causality is labeled as a hypothesis.
- The decision includes compliance, stock, capital and risk.
Official sources
Sources accessed 17 August 2026:
- Keepa — official price-history tracking and charting service
- Keepa — official API documentation
- Amazon — fees, cost variables and revenue calculator