Amazon competitor analysis in seven practical steps

Start with a decision and an evidence boundary

Amazon competitor analysis is the structured comparison of competing products, seller offers, and listing content to decide what to investigate or change. Its useful output is a short list of actions with evidence, limitations, and a way to evaluate the result. A category board alone cannot reveal a rival's profit, advertising budget, search conversion, or reasons customers purchased.

Write the decision before collecting data: revise a listing, investigate a loss of the Featured Offer, assess a price position, or explore a product range. Fix the marketplace, product type, collection window, delivery destination, and condition. Keep a record of what you could observe and what required an account report or an estimate. Scoutvo's methodology explains the difference between observations and interpretations.

1. Build a comparable competitor set

Start with products that solve the same problem for the same buyer. Use relevant category boards and a documented set of search queries; these produce different samples. A bestseller board is not the same as an organic search result, and a sponsored placement is not an organic position. Record how each ASIN entered the sample.

• Save the child ASIN, parent where known, brand, category or browse node, and collection time.
• Separate close substitutes from premium references, bundles, accessories, and products in a different capacity or use case.
• Deduplicate ASINs. If several children belong to one family, retain the child observations but report family counts separately.
• Record unavailable pages, missing fields, and exclusions. An inaccessible listing is missing data, not evidence that a feature is absent.

Output: a dated comparison set and an exclusion log. Stop expanding the sample when added products no longer match the decision; a larger mixed sample can be less informative than a smaller comparable one. There is no universal minimum number of competitors.

2. Separate the product, brand, and selling offer

One product detail page may carry offers from several sellers. Capture the observed seller ID, displayed business name, fulfillment method, and Featured Offer separately from the brand and manufacturer. Repeat seller IDs across ASINs are not additional independent businesses.

If seller location matters to distribution research, inspect the seller's business information and record the evidence and confidence. Registered business location does not establish manufacturing country, ownership, cost base, or deceptive intent. Leave unresolved cases unresolved. The seller-location guide covers those distinctions.

Output: a product-to-offer-to-seller map. Investigate a reseller problem at offer level and a product positioning question at product level. Do not attribute every seller's behavior to the brand.

3. Compare prices within meaningful product groups

Collect item price, shipping, condition, pack size, capacity, promotion conditions, seller, availability, and timestamp. Keep Prime, subscription, business, and coupon-dependent prices separate from a generally available offer. Compare like with like before calculating a median or price position.

Amazon's Featured Offer guidance highlights total price including shipping, delivery, customer experience, and stock. The lowest displayed item price does not guarantee the Featured Offer. For ongoing observation, use the price-monitoring workflow.

Output: a price range and position for a defined group, with missing values excluded and counted. Treat “low price, lower rank” as a reason to investigate other factors. It does not prove that price has no effect or that visibility is the only bottleneck.

4. Audit attributes and candidate keyword gaps

Compare titles, bullets, images, and category attributes against both the product's verified specifications and current marketplace requirements. Separate a field that is genuinely empty from a field your collection failed to retrieve. Fill only attributes that describe the product accurately.

Competitor vocabulary can suggest useful phrases. It cannot establish query demand, indexing, search rank, or conversion. Use a documented corpus, preserve language, and inspect phrases in context. A term common in dual-basket products may be irrelevant to a single-basket product. The keyword research guide shows coverage, TF-IDF, and validation.

Output: a list of valid field corrections and candidate terms with supporting product evidence. Stop a recommendation if its claim cannot be substantiated. An empty attribute is a potential completeness issue; any effect on filters or visibility needs separate checking.

5. Read reviews as a sampled set of experiences

Record the source, review count, date range, language, and products represented before grouping review themes. Define whether you count reviews, sentences, or mentions; one review can discuss several subjects. If you count mentions, report mentions rather than people.

Use themes to form questions: is cleaning difficult, does capacity match expectations, or is an instruction unclear? Check the original passages and negative examples before summarizing. A frequently mentioned feature is not necessarily the reason most customers buy, and reviewers do not represent all purchasers.

Output: a traceable theme list and product or communication hypotheses. Never turn a competitor's benefit claim into your own claim without product-specific support.

6. Map the range without treating variants as sales

Compare family and child counts, capacity, color, pack size, and functional differences. Mark the fields used to identify a family and any uncertain parent-child relationship. More children can indicate broader choice; they do not demonstrate higher sales or demand for each option.

Before proposing a new child, check the category's permitted variation theme and product relationship. A functional change may require a separate listing. Use the variation guide before turning market research into a catalog change.

Output: an assortment map and a shortlist of opportunities that still require demand, margin, supply, and policy validation. Do not infer a missing sales volume from a missing size.

What the July amazon.de example can actually show

The published air-fryer sample report describes a July 2026 amazon.de snapshot. It is historical context, not a current market census or an experiment. The article's original figures use different units:

The original report described the analyzed product as second cheapest in the priced subset and at board position 12. That juxtaposition supports a wider diagnosis, not a claim that a price cut would fail. Category identifiers, exact collection times, full exports, and exclusion records are not provided in this article, so the snapshot cannot be independently reconstructed here. Do not combine these denominators to create new market shares.

7. Turn findings into a decision worksheet

Separate corrections supported by product facts from hypotheses requiring performance evidence. Fix an inaccurate specification before testing a new phrase. For each task, name the evidence, expected mechanism, owner, measurement, and reason to stop.

For measurement, eligible brands can use Amazon Brand Analytics to examine search-funnel data. Record report definitions and periods rather than mixing sessions, clicks, purchases, and units in one rate. A before-and-after improvement can coincide with promotions, inventory changes, or seasonality; it does not automatically prove the edit caused it.

Repeat the parts that match the decision. Review pricing and offer events on an operational schedule; refresh a full competitor set before a major range decision or when the market materially changes. A quarterly review is a possible planning rhythm, not a requirement or a freshness guarantee.

Frequently asked questions