SEO for Marketplaces

By Brian Tr · Updated Oct 6, 2026

A single-brand store decides which pages exist. A marketplace doesn't. Sellers, hosts, drivers, or landlords create the listings, and the number of indexable pages moves with supply: up when a seller uploads a catalog, down when a listing expires, sideways when two sellers post the same item. That one difference changes almost everything about how SEO works.

This article is about that difference. It doesn't repeat the generic on-page checklist (titles, headings, internal anchors, image alt text). For that, read e-commerce SEO strategy and the page-level guide to product page SEO. What follows covers the problems those guides don't have to solve: pages you didn't write, pages that go stale on their own, and pages that are empty before the marketplace has enough supply.

If you need the background on how a marketplace differs from a store as a business, see marketplace vs e-commerce store. SEO is one of several demand channels, and it's the one most tightly coupled to supply.

Why search matters, and why it's fragile, for marketplaces

Public filings are blunt about this. Etsy's 2025 annual report says it obtains a significant number of visits via search engines such as Google, and that changes to search algorithms have and may continue to negatively affect the placement of links to its marketplaces. Airbnb's 2025 annual report lists search engine optimization among its paid and unpaid channels and warns that if listings are less visible, it may need to increase paid marketing, raising costs. Zillow's filing says changes to search engine models or algorithms may negatively affect its placement or require greater investment to optimize.

None of those filings gives a percentage of traffic from organic search, so don't borrow one. The useful reading is simpler: organic search substitutes for paid acquisition, and the three companies treat losing it as a cost risk. If you want the cost side of that trade, customer acquisition cost explains how to compare channels.

Supply creates the pages

In a store, SEO starts with a keyword and ends with a page. In a marketplace it often runs the other way. You have a large set of listings, and the question is which combinations of those listings deserve a URL a person might search for.

The common page types are:

Category-by-location pages are where marketplaces earn most of their organic reach, because buyers search for "thing + place" and no single seller can own that query. They're also where marketplaces most often get into trouble, because the number of possible combinations is far larger than the number that has real supply behind it.

Two rules help.

A page should exist because supply exists. If you can't show a visitor at least a few real, current listings on the page, it shouldn't be a crawlable, indexable URL yet. This is the practical reading of Google's own guidance. Its spam policies define scaled content abuse as many pages generated for the primary purpose of manipulating search rankings and not helping users, and the examples include stitching together content from other pages without adding value. Its doorway abuse definition includes pages targeted at specific regions or cities that funnel users to one page, and creating substantially similar pages that are closer to search results than a clearly defined, browseable hierarchy.

A page needs something the listings don't supply on their own. That might be local context, a count, a price range computed from live data, or answers to questions buyers ask in that category. This is a judgment call, not a Google formula. Without that, a page is just a list of listings repeated under thousands of city names, which is the pattern the scaled content examples describe.

User-generated content is the other half of the engine. Reviews, seller descriptions, and buyer questions give a marketplace unique text at a scale no editorial team could write. Google's site reputation policy is often read as a threat to this, but its own text says that sites designed to allow user-generated content, such as a forum website or comment sections, are not considered inconsistent with the policy. The policy targets third-party content hosted mainly to borrow the host's ranking signals. A listing that a real seller made to sell a real item isn't that. A section of third-party "best of" pages that sellers pay you to host, mainly to rank, might be.

The same document has a separate heading for user-generated spam: spammy content added to a site through a channel intended for user content. A marketplace is exactly that kind of channel. Moderation is an SEO function as well as a trust function, which is why marketplace supply quality matters here too.

The cold-start problem for search

Early on, a marketplace has the taxonomy but not the supply. Every category-by-location page you generate returns zero or one listing. The obvious temptation is to publish all of them anyway and hope Google fills them in.

Three things go wrong.

1. Thin pages get indexed or discovered and ignored. Google's crawl budget guide lists sites with a large portion of their URLs classified in Search Console as Discovered - currently not indexed among the sites that need to read it. If most of your URLs are empty shells, that's the report you'll fill.
2. Empty pages are a poor first impression for a buyer. Search is often the first contact. A visitor who lands on "No results" has learned something about your supply that no brand campaign will undo.
3. You spend your crawl on pages you have to rewrite later. Google says the factor you can control the most is perceived inventory: if many known URLs are duplicates or unimportant, a lot of crawling time is wasted.

The working answer is a supply threshold. Decide the minimum number of live listings a page needs before it's linked, included in your sitemap, and allowed to be indexed. Below it, the page can still exist for users who navigate there, but keep it out of the index (a noindex on pages below the threshold, or simply no link and no sitemap entry). Google's scaled content guidance says if you host such content on your site, exclude it from Search.

This ties SEO to the harder problem in the chicken-and-egg problem. Search can't create supply density in a market you haven't seeded. It can only reward density you've already built, and the sensible order is to open pages in the cities and categories where supply acquisition has already worked. Launching SEO pages for a city you haven't reached produces traffic you can't convert.

Faceted navigation and crawl budget at scale

A marketplace with filters (price, size, color, distance, rating, availability) can generate a nearly unlimited number of URLs from a modest number of listings. Google's faceted navigation guide describes the cost directly: the most common URL-parameter implementation can generate infinite URL spaces, which causes overcrawling and slower discovery of new, useful URLs.

Google also says who this affects. Its crawl budget guide is aimed at large sites with 1 million or more unique pages that change about weekly, and medium or larger sites with 10,000 or more unique pages that change daily, noting that those figures are rough estimates and not exact thresholds. A marketplace with active listings gets to those sizes faster than a store does, because the catalog is never finished.

Google's options for faceted URLs, in the order it lists them:

• If you don't need them indexed, prevent crawling. Use robots.txt to disallow the parameters, and keep the individual item pages plus a dedicated unfiltered listing page crawlable. Or use URL fragments for filters, which Google generally doesn't crawl or index.
• If you do need some indexed, follow its rules. Use the standard & separator, keep the order of path filters consistent with no duplicates, and return a 404 when a filter combination has no results. Don't redirect empty combinations to a generic error page.
• Canonical and nofollow are weaker. Google says rel="canonical" and rel="nofollow" on links are generally less effective in the long term than robots.txt or fragments for controlling crawl.

For a marketplace the decision isn't "block all facets" or "allow all facets." It's choosing the few filter combinations that match real search demand and give them clean, stable, linked URLs, and walling off the rest. "Brand + category" or "category + location" might be worth indexing. "Category + location + price band + two colors + sort by newest" never is.

One more discovery point. Google's e-commerce structure guidance warns that if category pages don't link to every product, Googlebot might not find all of your products by crawling alone, because it generally doesn't submit searches into a search box. A marketplace whose listings are reachable only through on-site search has an indexing hole. Browsable category paths and a sitemap close it.

Expired, sold-out, and seasonal listings

Listings die constantly: the item sells, the booking window passes, the seller leaves, the season ends. Each one is a URL that Google already knows. What you do with it has three options, and each has a cost.

First, what Google does with the signals. Its crawling documentation says a page that returns 200 but shows an error or an empty page is treated as a soft 404 in Search Console. A real 404 or 410 removes the URL from the index if it was previously indexed, and the crawling frequency gradually decreases. The crawl budget guide is explicit about the cost of the soft version: soft 404 pages continue to be crawled and waste your budget.

The first two rows are mine, not Google's decision tree: Google documents the status code behavior and the markup values, and the choice between them depends on whether the page still helps someone. A good test is whether you'd be comfortable showing the page to a buyer who arrives from search today.

The failure to avoid is the in-between: a page that says "This item is no longer available" with a 200 status, no alternatives, and no end date. It's a soft 404, it keeps getting crawled, and it never leaves your sitemap because your sitemap generator doesn't know the listing expired. Make listing state a first-class field and have the page template, status code, structured data, and sitemap all read from it.

Duplicate listings across sellers

Marketplaces create duplicates by design. Ten sellers list the same phone, the same book, or the same standard service package, each with a different URL and often a copied manufacturer description. Google's canonicalization guide says when it finds duplicate or very similar pages, you can state a preference, and it ranks the methods: redirects are a strong signal, rel="canonical" is a strong signal, and sitemap inclusion is a weak signal. Those methods can stack. It also says that if you don't specify one, Google chooses the version it judges best.

The reasons Google gives for consolidating are directly relevant: showing the URL you want people to see, consolidating link signals into one preferred URL, simplifying metrics, and avoiding spending crawling time on duplicate pages. It also lists best practices that matter at marketplace scale: don't use robots.txt for canonicalization, and don't use the URL removal tool for it.

There are two design patterns, and the choice is a product decision, not just an SEO one.

• One product page, many offers. A single page per distinct product, with sellers listed as offers beneath it. Duplicates disappear by construction. Google's product snippet markup supports this with AggregateOffer, intended for a product sold by multiple merchants, with lowPrice and priceCurrency required. It also says not to use AggregateOffer to describe a set of product variants.
• One page per listing. Each seller's listing keeps its own URL, which suits unique goods (a used car, a rental, a handmade piece). Where several listings are near-identical, pick a canonical listing by a stable rule, and point the others at it only if they're truly the same thing for a buyer.

For unique inventory, don't canonicalize different items to one another just to reduce page count. Two different apartments in the same building are not duplicates, even if the descriptions look alike. Fix thin descriptions at the source instead, through listing forms that require specifics, which belongs in seller onboarding.

Google's spam policy also names copying: thin affiliation describes product pages where descriptions and reviews are copied directly from the original merchant without any original content or added value. A marketplace full of seller-pasted manufacturer text is exposed to the same pattern.

Structured data for listings, reviews, and local providers

Structured data tells Google what a page is. For a marketplace, the rule is to mark up only what the page is, and to pick the type that matches the page's role.

• Product pages. Google's merchant listing guidelines say only pages where a shopper can purchase a product are eligible, not pages with links to other sites that sell the product. Product rich results support pages that focus on a single product or its variants, and Google recommends marking up product pages instead of pages that list products or a category. Don't put product markup on category or search result pages.
• Local providers. Google's local business markup asks you to define each local business location as a LocalBusiness type and to use the most specific subtype. For a services marketplace, a provider profile maps to that type. Keep the details on the page itself consistent with the markup.
• Reviews. Google's review snippet documentation says LocalBusiness and Organization review markup is supported only for sites that capture reviews about other local businesses or organizations, and it points to its guidelines on self-serving reviews. A marketplace collecting buyer reviews of sellers is in that situation. A marketplace marking up reviews of itself is not.

Structured data doesn't fix thin pages. It describes what's there. A listing page with a complete, accurate offer, clear availability, and real reviews gets accurate markup nearly for free if the data lives in structured fields.

Review quality is part of SEO

Reviews are the cheapest unique content a marketplace produces, which is why they attract abuse. Google's guidance on writing high quality reviews asks for evidence of firsthand experience, quantitative measurements, benefits and drawbacks, and a focus on the most important decision factors. Those are written for editorial reviewers, but they point at what separates useful reviews from filler: specifics.

Practical implications for a marketplace:

• Ask for specifics at submission (what was bought, when, what the experience was) rather than a star rating alone.
• Moderate for spam and fake reviews. Google's user-generated spam category makes this a search risk as well as a trust one.
• Don't import reviews from elsewhere without adding value, since copying falls under the scraping and thin-affiliation definitions in the spam policies.

A sequence that fits a young marketplace

1. Open the plumbing first. Fix status codes, canonical rules, and sitemaps before generating pages at scale, because it's far harder to unwind 100,000 bad URLs than to launch 1,000 good ones.
2. Set the supply threshold. Decide how many live listings a page needs before it's indexable, and enforce it in code.
3. Choose which facets are real landing pages. Index a handful with clean URLs, and block the rest from crawling.
4. Model listing state. One status field that drives page content, HTTP status, markup, and sitemap membership.
5. Pick the duplicate pattern. Product-with-offers or one-page-per-listing, per category.
6. Mark up honestly. Product, LocalBusiness, and review markup on the right pages only.
7. Expand by density, not by taxonomy. Add new category and city pages only after supply exists there, in step with your supply-first or demand-first choice.

Key Facts: Marketplace SEO

• Google's crawl budget guide is aimed at sites with 1 million or more unique pages that change about weekly, or 10,000 or more pages that change daily, and says those figures are rough estimates, not thresholds.
• Google says soft 404 pages continue to be crawled and waste crawl budget, while a 404 or 410 removes an indexed URL and gradually reduces crawling.
• Google's faceted navigation guide says URL-parameter facets can generate infinite URL spaces and recommends returning a 404 for filter combinations with no results.
• Google's spam policies define scaled content abuse as many pages generated mainly to manipulate rankings, and its site reputation policy does not treat sites designed to allow user-generated content as inconsistent with it.
• On canonicalization, Google ranks redirects and rel="canonical" as strong signals and sitemap inclusion as a weak one.
• Etsy's 2025 10-K says it obtains a significant number of visits via search engines such as Google, and Airbnb's says weaker SEO visibility may force it to increase paid marketing.

Related Reading

• E-commerce SEO Strategy
• Product Page SEO
• Marketplace vs E-commerce Store
• The Chicken-and-Egg Problem in Marketplaces
• Acquiring Supply for a Marketplace
• Marketplace Supply Quality
• Acquiring Buyers for a Marketplace
• Why Density Beats Reach in Local Marketplaces