Amazon data workflows: step-by-step guides with API calls
Amazon data workflows: step-by-step guides with API calls
Guides that take one seller task, such as product research, review analysis, price monitoring or building a keyword library, from the first request to a decision, with runnable examples.
Amazon Search Trends: How to Tell a Real Rise From a One-Week Spike
A search term that jumps one week is a question, not an answer. Read its 1-, 4- and 12-week rank growth together, know why the growth rate tops out near 100%, and separate brand terms from product terms before you stock anything.
Amazon Backend Search Terms: What Goes in the Field and How to Fill It From Reverse Lookup Data
The backend search terms field is short, counted in bytes, and ignored entirely when it runs over. This post covers the rules Amazon publishes for it, why nobody can read a competitor's backend field, and a six-step way to fill yours from the terms competitors already rank for.
Reading Negative Amazon Reviews by Variation: Which Complaints Belong to Which Product
A variation family pools its reviews, so the complaints on one page can belong to several different products. In a sample of 20 recent critical reviews we pulled, 3 were about a spare part sold as a variation, and the one-star and three-star reviews described two different problems. This post shows how to split them.
Choosing a Price Point on Amazon: Read the Category's Price Bands Before You Set a Price
In one category we tested, 36 of the top 100 products sat between $15 and $20 and took 42% of unit sales, while the four products at $20 to $25 each sold more than any other band. This post shows how to read a price distribution into three numbers per band and what they say about where to price.
How Much Does an Amazon Data API Cost? Calls, Pack Size and Where Calls Get Wasted
The price on a call pack is only the start. Turn a workflow into a call count, pick the pack by what you will use in 60 days, and see where calls are wasted before you buy. Prices are from the pricing page as of 31 October 2026.
Building an Amazon Niche Finder: What Each Filter Decides and What It Hides
A niche finder is a set of filters over category statistics. This post shows which filters answer which question, how to read one result row without being fooled by head-sample averages, and why the filters you choose matter more than the tool you use.
Choosing an Amazon Data Analysis Tool: Test It With Data You Already Know
Every tool looks accurate in a demo. What separates them is comparing their estimates against ASINs whose real sales you hold in Seller Central — reading direction, magnitude and ranking — and then asking a few questions about definitions.
Seasonal Product Research on Amazon: Measure Concentration Before You Pick a Month
The risk in a seasonal product is not the peak, it is the inventory still sitting there in the off-season. Use multi-year monthly search, per-ASIN monthly sales and launch-year cohorts to decide whether to enter and which month to decide in.
Choosing an Amazon Keyword Tool: First Work Out Which "Search Volume" It Shows
"Which keyword tool is most accurate" usually compares two different quantities. A single reverse lookup response holds four fields that look like search volume, each with its own period and scope. This post splits the keyword toolkit by task and lists the three questions to ask when evaluating one.
Product Research on Amazon US, Japan and Europe: The Same Analysis Does Not Transfer As-Is
Product research methods carry across marketplaces; the numbers do not. Every request needs a marketplace (four common endpoints silently default to the US), amounts come in each marketplace's currency, keywords are stored in its language and category paths are defined per marketplace. This post sets out what to redo and what to reuse.
Building an Amazon data report: if it cannot be compared to last week, it is not a report
Reports go unread not because they lack detail but because they cannot be set beside the previous one. The three things that must stay fixed, a minimum report structure, and three habits that void a report.
Five decisions to settle before you integrate an Amazon data API
The docs cover how to send a request. This covers what to settle before you send one, including the rule people trip over most: RPM, concurrency and credits aggregate per account, so extra API keys buy no extra quota.
Amazon category analysis: four distributions decide whether you can still get in
A big category total does not mean a share of it is available to you. Listing age, seller type and price distributions, plus one statistics call, read the competitive structure — above all whether new listings still have room.
Amazon competitor analysis: the comparison table is useless, the difference is not
A side-by-side competitor table rarely produces an action. Four angles that do — keyword difference, dual-placement waste, unmet demand in negative reviews, variation structure — plus one number you can compute directly.
An Amazon Competitor Analysis Template That Ends in Actions
A downloadable competitor sheet that keeps what you observed, what you think it means, and what you will do in three separate columns — plus how to pick rivals and keep the rows comparable.
Amazon data analysis: separate facts, estimates and your own judgment first
The expensive mistake in Amazon data analysis is not miscalculating — it is using an estimate as if it were a fact. The boundary between three kinds of data, which to query for which question, and one recording rule that applies to all of it.
Choosing an Amazon Data API: SP-API, Third-Party Endpoints and Scraping
Three things go by the name "Amazon API" and they solve different problems. What each one can and cannot return, and the most common misjudgement: expecting SP-API to show you a competitor.
Where Amazon data comes from: why there is no single "Amazon database"
People searching for a database want a dataset to download. Data splits along authorization boundaries into three blocks, and it is a stream rather than a store. A map of what exists, how fast each field goes stale, and whether you can keep a copy.
How to optimize Amazon keywords: sort them into four conversion classes
Keyword optimization is not stuffing words into a title. Your listing already gets impressions for a set of terms, and optimizing means classifying that set — which to defend, which to investigate, which to stop paying for.
Amazon Market Analysis: Where the Market Size Number Comes From
Market size is the least reliable number in the dataset — it rests on a head sample and two layers of estimation. This post opens up how it is computed, the unit trap in ratio fields, and how to read share properly.
Choosing Amazon negative keywords: three sources and one line you should not cross too early
Negatives are not about deleting terms. They move budget away from positions that will not convert. Three classes safe to negate, what each looks like in the data, and when negating does damage.
Amazon PPC keyword bidding: why one term comes back with 36 numbers
Asking what to bid on a term expects one number. The endpoint returns 36 — three match types, each with a bid, CPA, budget and ACOS, and each of those a range rather than a point. How to read them.
Choosing Amazon PPC keywords: exclude three kinds first, then read four competition fields
Picking ad keywords is not picking SEO keywords. SEO needs relevance and volume; advertising also needs terms you can afford to win. The three kinds to exclude first, and the four fields that actually decide it.
How to Find Products to Sell on Amazon: Eliminate Before You Discover
How to find products to sell on Amazon starts with removing them, not finding them. Four elimination layers, what data each one uses, and why reversing the order throws away everything you did before it.
An Amazon Product Research Template That Makes Two Sheets Comparable
A downloadable product research sheet that fixes the marketplace, the collection date, and which numbers are estimates — plus the field mapping for when filling it in by hand stops scaling.
Choosing an Amazon product research tool: the difference is how much it decides for you
What separates product research tools is not whose data is more complete — it is who set the weights behind that "opportunity score". Three tool shapes, an acceptance checklist you can run, and how to write your own criteria instead.
Remote MCP: Query Amazon Data From ChatGPT Without Pasting a Key
The browser OAuth path, for clients that cannot run a local installer. Client support matrix, a first prompt that verifies the connection without spending credits, and what revoking a connection actually does.
Choosing a Reverse ASIN Tool: An Acceptance Checklist You Can Run Yourself
No rankings. A method instead: fix a set of ASINs, test any tool against the same conditions, and check the three things no reverse lookup can return. Downloadable checklist included.
How to analyse Amazon reviews: from a pile of them to three kinds of actionable conclusion
The output of review analysis should not be a summary. It should be three separated conclusions — fix the product, fix the listing, fix nothing — because blending them turns a description problem into an expensive product change.
Exporting Amazon reviews into a spreadsheet: 21 columns and a template to download
The problem with copy-pasting reviews into Excel is not that it is slow — it is that it drops the fields you need to judge them. A downloadable review analysis sheet, the endpoint field behind every column, and the three traps in exporting.