Marketplace intelligence in e-commerce is the regular collection of public data from marketplaces such as Amazon, Walmart and eBay: the offers and sellers on each listing, the featured offer, prices, stock, ratings, bestseller ranks and positions in categories. Retailers and brands use it to decide prices and product range, both on the marketplace and in their own store.

A marketplace is not just one more competitor. A single product page can carry five, ten or thirty sellers. The price a shopper sees is chosen by the platform's algorithm, not by any one seller. And very often the marketplace sells the same product itself. Methods that work well for watching competitor websites give a distorted picture here, because the unit of analysis is no longer the product. It is the offer.

This guide is for retailers that sell on marketplaces and in their own store, for brands that sell through marketplaces, and for direct-to-consumer stores whose prices shoppers compare with Amazon. It covers which marketplace data you can actually collect, how to read it for pricing and for assortment, and how to set up collection that holds up over time.

What Is Marketplace Intelligence (and How It Differs from Monitoring Competitor Stores)

The broader term, market intelligence, covers everything a company learns about its markets, industries and customers. Marketplace intelligence is the e-commerce slice of it, tied to specific selling platforms. Vendors also call the same practice ecommerce intelligence, ecommerce market intelligence or marketplace analytics, though the latter is also used for sellers' analysis of their own sales.

The difference from watching competitor stores is structural:

The same fields, such as price, stock and rating, mean different things on a marketplace. A price is attached to a seller, not to a product. A product can be in stock with one seller and sold out with another. And third party sellers can enter or leave a listing in a day, changing the price shoppers see without any change in the product.

If your competitors are mostly independent web stores, the signals to watch are different; we cover them in our guide to monitoring competitor stores. The rest of this article is about what is specific to marketplaces.

Marketplace Data You Can Collect (and What Stays Hidden)

Marketplace pages show more than most retailers use. The table groups the public fields by level and shows what each one helps decide.

Some things never appear on public pages, and it helps to know where the line is:

• Visible at the moment of collection: everything in the table above.

• Only as an estimate: sales and revenue (third-party tools model them from bestseller ranks and other signals); the share of time a seller holds the Buy Box (you can only estimate it from your own series of snapshots).

• Not public at all: data from a seller's own account (your sales, conversion rate, exact Buy Box share), advertising bids, stock in units, costs.

The complete picture comes from combining two sources: public marketplace data for the market and your seller account data for your own performance. Neither replaces the other.

Two cautions. Prices and the featured offer can depend on the delivery location, so a data set collected for one region may not match what shoppers see in another. And field names and page layouts differ by marketplace, so the same metric needs separate handling on Amazon, Walmart and eBay.

Product fields such as name, price, stock, brand and ratings are standard; seller data and offer details are collected as custom fields such as seller data. If you want to collect data from marketplaces like Amazon and eBay, the main constraint is not whether the data can be collected but how often, because marketplaces run the strongest anti-bot protection of any type of site.

Using Marketplace Data for Pricing

The general methods for setting prices against competitors, such as deciding where to be cheapest and where you can charge more, apply to marketplaces too, and we cover them in our general competitor pricing plays. Below are the five decisions that only exist on marketplaces. Together they make up a practical marketplace pricing strategy.

Price Against the Featured Offer, Not the Lowest Offer

The lowest price on a listing is often misleading. It may belong to a used or renewed item, to a seller with a two-week delivery time, or to a seller with a poor rating. Shoppers rarely see it. What they see is the featured offer, the Buy Box.

So compare your price with two numbers: the featured offer price and the lowest price among offers of the same type as yours (new condition, similar delivery speed). If your offer is $2 above an obscure used offer but level with the Buy Box, you are competitive. If you are $2 above the Buy Box, you are not, whatever the lowest price says.

This is where an Amazon pricing strategy most often goes wrong: cutting price to match an offer that was never really competing for the sale.

Read Price Dispersion Across Sellers

Price dispersion is the spread between the lowest and highest prices for the same product on one listing. It tells you about the state of that market:

• A narrow spread (within a few percent) means the market has settled. Sellers are likely using repricers and reacting to each other, and a lower price will be matched quickly.

• A wide spread means something is moving: new sellers testing the market, someone clearing stock, or goods arriving from a channel outside normal distribution.

• A spread that grows over several weeks on a brand's products is a signal for the brand itself to look at where the low-priced stock comes from. If it involves unauthorized and below-MAP sellers, that is a separate enforcement process.

Dispersion only makes sense on the offer level, so it requires collecting the full offer list, not just the featured offer.

Know When the Marketplace Itself Is Your Competitor

On many listings the marketplace sells the product itself. When it does, it usually holds the featured offer at a price that third party sellers struggle to beat profitably.

Competing head-on there is often a losing race. The more useful questions are:

• On which of your products does the marketplace sell directly? This is a single field in your data, and the share alone tells you a lot.

• Can you sell a different version instead: a bundle, a larger pack, an exclusive color?

• Should these products go to other channels, while you focus marketplace effort on listings where the platform does not sell?

An Amazon competitor analysis that ignores the platform's own offers misjudges the market on exactly the products where the stakes are highest.

Watch Seller Count, Not Just Price

The number of sellers on a listing often moves before the price does. When three sellers become eight within a month, price erosion usually follows: new sellers enter by undercutting. When sellers leave, for example because a product becomes hard to source, there is room to hold or raise the price.

Keep seller count as a history for your key listings. A rising count is an early warning; a falling one is an opportunity that disappears once someone else notices it. Marketplace price monitoring that only stores prices misses this signal entirely.

Compare Marketplace Prices with Your Own Store

Shoppers compare channels. If a product is 15% cheaper on a marketplace than in your own store, many of them will buy it there, and your store becomes a showroom.

The reverse matters too. Large marketplaces compare offers with prices elsewhere and may stop featuring an offer that is priced well above other channels. Check each marketplace's current pricing policies before you set a gap on purpose.

The practical approach to omnichannel pricing is to set a target gap per channel, for example "marketplace price within 3% of our store price on comparable products", and check it on every collection cycle. That requires the same product to be matched across your store, the marketplace and other retailers, which brings us to setting up the data.

Using Marketplace Data for Assortment

The full method for comparing your range with competitors, with metrics, an overlap matrix and a scoring step, is in our full competitor assortment analysis guide. Marketplaces add signals that no competitor website shows. Here are four ways to use them.

Bestseller Ranks as a Demand Signal (with Caveats)

Best Sellers Rank is the closest thing to a public demand figure in e-commerce. On Amazon, every product page shows its rank within one or more categories; the category pages of the Amazon bestseller list and Walmart best sellers show which products lead right now.

Read the numbers carefully:

• Compare ranks only within one category. A rank of 500 in a small category and 500 in a huge one mean very different volumes.

• Use a series, not a snapshot. Ranks update frequently and jump with a single promotion. Weekly averages from daily snapshots show the real direction.

• Do not convert ranks into unit sales yourself. Third-party tools estimate sales from ranks with their own models; treat their figures as directional.

What ranks do show reliably is which subcategories and product types are climbing. A product type that holds several top-50 positions for a month, while it is missing from your range, is worth a closer look.

Spot Crowded and Thin Niches

Marketplace data shows how contested a niche is in a way competitor websites cannot:

• Many sellers on the same few listings means the niche is crowded and price is the main weapon. Entering with the same products leads straight into a price war.

• Few brands in a category's top 100 and mediocre ratings among the leaders (for example, an average below 4.0) suggest room for a better product. Shoppers buy what is available, not what they love.

• Review text on the leading products shows what buyers complain about, which tells you what a better product would need to fix.

This kind of Amazon market research works best as a comparison across several candidate subcategories, scored on the same criteria.

Decide What to Sell on the Marketplace and What to Keep in Your Store

Not every product belongs on every channel. Marketplace data helps decide where each one goes:

• Products where the marketplace sells directly and a dozen sellers compete are often better kept for your own store, or sold on the marketplace in a version nobody else offers: a bundle, a set, a different pack size.

• Products only you carry can use the marketplace for reach, because there is no one to undercut you on the listing.

• Products that rank well on the marketplace but are missing from your store may be worth adding there. If they come from a supplier's online catalog, you can turn it into import-ready catalog files rather than enter products by hand.

For Brands: Check Your Listing Coverage

Brands selling through marketplaces need a different view: how their own products appear.

• Which of your products are listed at all, and on which marketplaces?

• How many sellers offer each one, and who are they?

• Which listings were created without you, with wrong titles, missing images or incomplete attributes?

• Where do your products rank in their categories, and how do their ratings compare with competing brands?

Brands often use Amazon seller analytics tools for their own sales, but those tools see only their account. Listing coverage requires public data across all sellers, the same kind of data that digital shelf analytics platforms collect for large brands.

How to Set Up Marketplace Data Collection

A marketplace data project fails most often at scope: collecting too little to see the market, or too much to keep fresh. These six steps keep it manageable.

Step 1. Choose Marketplaces and Regions

Start where your customers buy and where your competitors sell: usually one or two of Amazon, Walmart and eBay, plus a regional marketplace if you sell in Europe or Latin America. For Amazon, choose the country sites that matter. Remember that prices and the featured offer can vary by delivery location, so fix one location per region and keep it stable.

Step 2. Define the Scope: Three Page Types

Each type of page answers a different question:

• Your product listings (by marketplace ID): who sells your products, at what price, who holds the featured offer.

• Category bestseller pages: what sells in your categories and how that changes.

• Search results for 10–30 key queries: which brands and products shoppers see first, and how many of those slots are sponsored.

Start with the first type for pricing and the second for assortment. Add search results once the first two are running.

Step 3. Decide on Listing-Level or Offer-Level Data

Listing-level data (the featured offer, rating, rank) is lighter and enough for price positioning against the Buy Box. Offer-level data (every seller with price, shipping, condition and fulfillment) is needed for price dispersion, seller count and spotting the marketplace's own offers. It is heavier to collect and to store, so use it where those questions matter: usually your top products and the listings where you see the most price pressure.

Step 4. Map Marketplace IDs to Your Catalog

Marketplace IDs do not match your SKUs. Link each listing to your product by GTIN where the page shows it, then by brand, model and key attributes such as size and pack count. One product can have several listings, and one listing can carry several variants. Without careful mapping, a price for a 2-pack ends up compared with your single unit. Our guide explains how to match marketplace listings to your catalog and other stores.

Step 5. Set the Cadence per Question

Different questions need different refresh rates:

Scheduled price scraping handles the repetition. On the largest marketplaces, the practical limit on frequency is anti-bot protection; this is why marketplaces are the hardest sites to collect from, and it is worth agreeing realistic refresh rates before you plan decisions around them.

Step 6. Keep Snapshots and Join Them with Your Seller Data

Every collection is a snapshot. Only a series of them shows how often a seller held the featured offer during your checks, how a rank trends, and how seller count changes. Store every run with its timestamp. Then put the public data next to the data from your seller account: your sales and your own Buy Box share explain what the market data shows, and vice versa.

Example: A Beauty Retailer Selling on Amazon and Its Own Store

The scenario below is illustrative.

A beauty retailer sells 60 hair care products both on Amazon and in its own online store. Margins on Amazon have been falling for two quarters, and the team suspects it is losing on price. It collects daily offer-level data for its 60 listings, bestseller pages for its three main categories, and prices for the same products in its own store.

After a month, the data shows:

The first decision alone changed the picture. The team had been matching prices on listings where it could not win the featured offer, losing margin on every sale. Without the seller field in the data, the falling margins looked like a general price problem.

Marketplace Intelligence Tools: What Each Type Covers

Most tools that come up for marketplace analytics were built for one job. This is how the main types compare:

The types complement each other. A seller tool helps with product research and your own account, a repricer acts on prices, and a data service delivers the public market data in your own format, the only option that puts marketplaces and competitor websites in one table. When comparing vendors of any type, the questions in our guide on how to choose a price monitoring tool apply here as well, including coverage of your hardest sites and full data export.

Collecting Marketplace Data with ShopScraping

ShopScraping collects product data from public marketplaces and e-commerce websites and delivers it as clean, structured tables. For marketplace intelligence, that means:

• Amazon, eBay and other marketplaces alongside your competitors' own websites

• Prices, sale prices, stock status and variants

• Ratings and review counts

• Product identifiers, brand and category path

• Seller data and other marketplace-specific details as custom fields, where the marketplace shows them on public pages

You choose the listings, categories and fields, set a daily, weekly or monthly schedule, and export to CSV, Excel, JSON or Google Sheets.

What we don't do: estimate sales, access your seller account data, track ad spend, change your prices, compare runs or send alerts. ShopScraping is the data layer. Your team, your repricer or your BI tools draw the conclusions.

If prices across web stores and marketplaces are your starting point, see how we deliver competitor and marketplace price data. Check plans and pricing, or start collecting marketplace data.

FAQ

What is marketplace intelligence?

Marketplace intelligence is the regular collection and analysis of public data from online marketplaces: offers and sellers on each listing, the featured offer, prices, stock, ratings and bestseller ranks. Retailers and brands use it to make pricing and assortment decisions on marketplaces and in their own stores.

What is the difference between marketplace intelligence and market intelligence?

Market intelligence is the broad practice of learning about markets, industries, competitors and customers from many sources. Marketplace intelligence is narrower: it focuses on selling platforms such as Amazon, Walmart and eBay, and works on the level of individual listings, offers and sellers.

Can you see a competitor's sales on Amazon?

Not directly. Sales figures are not public. Best Sellers Rank shows a product's demand relative to others in its category, and third-party tools estimate sales from it, but those estimates are directional. Exact sales are visible only to the seller in their own account.

Can you track who wins the Buy Box?

You can see which seller holds the Buy Box, and at what price, at the moment of each collection. Over many snapshots, that gives an estimate of how often each seller held it during your checks. Your exact Buy Box share for your own offers is available only in your seller account.

How often should you collect marketplace data?

It depends on the question. Featured offers and prices on key listings need daily collection, more often during big sales events. Seller counts work daily or a few times a week, bestseller ranks are best collected daily and read as weekly averages, and listing coverage is fine monthly. On large marketplaces, anti-bot protection sets a practical limit on frequency.

Does Amazon allow web scraping, and is it legal?

Marketplaces set their own terms of use, and collecting data from them has legal and contractual considerations that depend on the method, the data and your jurisdiction. Collecting publicly displayed product information, without logging in, without overloading servers and without personal data, is a common business practice. Our guide Is Web Scraping Legal? covers the main cases. This is general information, not legal advice.