Launching in a new country or category? Get the full competitive intelligence landscape — a complete SKU census, price bands, brand coverage, assortment gaps and marketplace penetration — extracted from the platforms that actually sell in that market. Built for strategic decisions, not daily ops.
The same drill is sold on three marketplaces by nine sellers. Counted as listings it is twenty-seven data points; counted as products it is one. Market sizing built on listing counts overstates every band it touches, and the error is not uniform across price tiers.
Off-the-shelf market data covers Amazon and eBay and stops. In most European markets the local platform — Bol, Allegro, eMAG, OTTO — carries the assortment that decides whether a category is crowded or open.
Entry pricing decided from a twenty-SKU sample produces positioning errors that cost market share for quarters. The gap you are looking for is usually a narrow band with few brands in it, and a sample is exactly the wrong instrument for finding a thin band.
A category captured during a promotional week reads as structurally cheaper than it is. Without separating the list price from the current selling price, a snapshot taken in November describes a market that does not exist in February.
Market data should feed models and dashboards, not just slides. Structured, SKU-level output means your team keeps analysing after the initial study ships — and can re-run the same scope a quarter later without re-scoping it.
Per 1,000 listing pages extracted — a full category census costs less than an analyst day
Ready-made marketplace and retailer endpoints, plus any regional platform you add yourself
Free requests on every new account — sample one category on one marketplace before you commit
Tell us the country, category and platforms you are assessing. We agree the taxonomy, the inclusion rules and what counts as in-category before a single page is collected.
Every listing in the category is enumerated through facets, extracted, normalised and matched across marketplaces. Products are deduplicated before any total is computed, and list prices are kept separate from promotional ones.
Delivered as JSON, CSV or straight to your data warehouse. Feed it into pricing ladders, market sizing models or assortment plans, and re-run the same scope later to measure what changed.
This is the actual sequence behind the landscape above — cordless drills in the Netherlands, across four marketplaces, from raw listings to a defensible price band.
Marketplace category pages cap pagination long before they run out of products — typically around 400 results, regardless of how many exist. The category is enumerated by traversing brand, price and attribute facets rather than by paging a single listing, so a category containing 1,284 listings does not silently become the 400 that one sort order was willing to show.
1,284 listings resolve to 612 distinct products. The same model appears on three marketplaces, and on each of them under several sellers with slightly different titles. Matching is done on identifier where published and on brand, model and normalised attributes where it is not. Every count downstream — band size, brand share, assortment depth — is computed on products, not listings.
Both prices are captured per listing, so a band can be measured at the price the market normally sells at and at the price it is selling at today. In this category the EUR 180–320 band holds 63 products from 9 brands — thin against 247 products from 24 brands in the band below it. That is the finding an expansion case is built on, and it only holds if the prices behind it were not a promotion week.
Market research extractions rarely fail on the price field. They fail on coverage and on double-counting, and both errors are invisible in the output — a wrong number looks exactly like a right one.
Every marketplace caps how deep a single category listing will page. Enumerating through facets — brand, price step, attribute — is what gets past that cap. A census that misses the deep tail systematically under-counts cheap, high-variant products, which is precisely where entry-level competition lives.
The same product carries different titles, different marketplace IDs and sometimes different brand spellings on every platform. Without cross-marketplace matching, a market with 600 products looks like a market with 1,300, and the brands present on the most platforms look artificially dominant.
A category's price structure is the list prices; its current competitive reality is the selling prices. Both are needed, and so is who is selling — a band dominated by third-party sellers behaves very differently on entry than one the marketplace stocks itself.
Cross-border listings inflate an assortment with products nobody in that market can actually buy within a reasonable lead time. Market, currency, tax treatment and stock status separate the real addressable assortment from the catalogue-level one.
Every listing collected lands in one of three states. A research dataset that does not distinguish them is a dataset whose totals cannot be defended in a board room.
First time this product has been seen in the census. It contributes one unit to its price band, one to its brand's share, and carries the listings it was seen on as evidence.
Matched to a product already counted, on another marketplace or from another seller. It adds an offer and a price point but does not add to the product count. Kept visible, because the number of platforms a product appears on is itself a finding.
Accessories in a tool category, bundles, non-stocked cross-border listings, or products outside the defined taxonomy. Excluded rather than dropped, so a reviewer can see what the census chose not to count and pull categories back in if the scope was wrong.
You define the market, the category and the platforms. You receive a deduplicated, SKU-level census with prices, brands and seller mix — as data, not as a deck.
The country and the category you are sizing. Defined as a taxonomy path, a set of category URLs, or a keyword scope you approve before extraction starts.
NL — Power tools > Cordless drillsGlobal marketplaces plus the regional platforms that actually carry the assortment in that market, and any DTC or brand sites you want mapped alongside them.
bol.comamazon.nlcoolblue.nlgamma.nlOptional. Supply your own category and attribute structure and the census is mapped into it, so the output drops straight into your existing models.
category_l1, category_l2attribute set: voltage, chuck, batteryOn by default. The same product on four platforms counts once toward product totals, and its four listings are retained as offers.
Once a scope is defined it can be re-run against it without re-scoping, which is what makes assortment change measurable rather than anecdotal.
One row per listing, keyed to one product, 13 columns each
Delivered by REST API, CSV export or straight into your data warehouse. Product-level totals and listing-level detail arrive together, so a band size can be traced back to the listings that produced it.
One row per listing, each carrying the product it resolved to — which is what lets you count products and offers from the same dataset. Three listings, two products.
| matched_product_id | marketplace | product_title | brand | list_price | selling_price | seller_type | count_status |
|---|---|---|---|---|---|---|---|
| P-00418 | bol.com | Accuboormachine 18V 2x5.0Ah | BrandCo | EUR 219.00 | EUR 199.00 | 1P | counted |
| P-00418 | amazon.nl | BrandCo 18V Cordless Drill Kit | BrandCo | EUR 219.00 | EUR 214.00 | 3P | merged |
| P-00418 | coolblue.nl | BrandCo accuboormachine 18 volt | BrandCo | EUR 219.00 | EUR 219.00 | 1P | merged |
| P-00907 | gamma.nl | Accuschroevendraaier 12V compact | RivalTools | EUR 74.95 | EUR 74.95 | 1P | counted |
| — | bol.com | Boorset 40-delig | BrandCo | EUR 24.95 | EUR 19.95 | 3P | out_of_scope — accessory |
Categories are enumerated through facets rather than paged, so the deep tail that pagination caps normally hide is included. Price bands, brand share and assortment depth are computed on the whole category — which is the only way a thin band shows up as an opportunity rather than as noise.
Global datasets miss Bol, Allegro, MediaMarkt, eMAG and the regional DTC leaders that decide whether a market is crowded. ScrapeWise extracts from whatever platforms actually sell in your target country, not from a supported-site list.
Delivered as JSON or CSV, deduplicated across marketplaces and mapped to your taxonomy, ready for market sizing models, pricing ladders and assortment plans. Product totals and listing detail arrive together, so every number can be traced to its source rows.
Stop waiting a quarter for consultants. ScrapeWise turns any e-commerce market into a structured, strategy-ready dataset — a deduplicated SKU census, price bands, brand share and assortment gaps — that you can re-run whenever the market moves.
Everything you need to know about e-commerce market research and competitive intelligence with ScrapeWise.
It is a point-in-time census of a category in a market: every product sold on the platforms that matter there, with its brand, price, seller and availability, deduplicated so the same product on four marketplaces counts once. From that dataset you derive price bands, brand share, assortment depth and the gaps — rather than inferring them from a sample or a panel.