datawebotCoupang product data: product IDs, options & crawler rules
How Coupang product URLs carry productId, itemId and vendorItemId, how options map to prices, and what Coupang's crawler rules say.
Updated · datawebot team
A Coupang product page shows one market and one page state
Coupang product research starts from a shopper-facing product page. Each page view is a dated observation of what one market showed for one product and selected option; it is not a seller inventory record or a market-wide feed. Keep the page evidence, the URL identifiers and the crawler rules as separate inputs.
Scroll to compare →
| Evidence | What it shows | What it does not establish |
|---|---|---|
| Buyer-facing product page | Shopper-visible product/catalog presentation, with market and page state | A complete set of all sellers, variants, stock or historic prices. |
| Product page URL | The productId path segment and the itemId and vendorItemId query parameters for the viewed page | A seller's own product code or a Coupang-wide SKU. |
| Robots and sitemap | Coupang's published crawl directives and sitemap location | Terms or permission to collect data. |
Coupang product URLs carry product, item and option IDs
Coupang product URLs take the form coupang.com/vp/products/{productId}?itemId=...&vendorItemId=.... The path and query parameters identify the product page and the option that was selected when the URL was captured. Store each value in its own field instead of collapsing them into one SKU.
Scroll to compare →
| URL part | Where it appears | Model it as |
|---|---|---|
| productId | Path segment after /vp/products/ | Product-page key for the listing group; store it with the capture time. |
| itemId | itemId query parameter | Item reference for the viewed page; keep it separate from productId. |
| vendorItemId | vendorItemId query parameter | Selected purchase option, distinct from the product group. |
| Full URL and capture time | The exact URL submitted or followed, plus when it was read | Provenance for every field taken from that page view. |
Price and availability are option-level page observations
A displayed price or availability state belongs to the option shown on the page at capture time. Keep it attached to the vendorItemId from the URL, the market and the timestamp, so a single page view is not presented as a price history or a checkout result.
Scroll to compare →
| Field | Attach it to | Interpretation |
|---|---|---|
| Displayed price | vendorItemId, market and capture time | The amount the page showed for that option at that time. |
| Availability shown on the page | vendorItemId and capture time | A page state, not proof that checkout will succeed. |
| Selected option | itemId and vendorItemId from the URL | The option the observation describes; other options need their own records. |
Coupang robots.txt blocks OpenAI crawlers and unspecified bots
In Coupang's robots.txt file checked October 4, 2026, the company explicitly disallows GPTBot, ChatGPT-User and OAI-SearchBot from /. Its final User-agent: * group also disallows /. Named Googlebot, Bingbot, Naver/Yeti and Daum groups publish different, selected path rules. These directives describe Coupang's crawler policy for those user-agent groups; they do not establish how any other AI system retrieves or answers about products.
Scroll to compare →
| Crawler group in the file | Published rules | Discovery implication |
|---|---|---|
| GPTBot, ChatGPT-User, OAI-SearchBot | Each has `Disallow: /` | These named OpenAI crawlers are not invited to fetch Coupang URLs directly. |
| Wildcard `*` | `Disallow: /` | Unspecified crawlers are disallowed by the catch-all group. |
| Googlebot, Bingbot, Naver/Yeti, Daum | Named groups allow selected product, category, campaign and search paths | Traditional search crawler policy is not the same as the OpenAI groups. |
| Operational paths | Cart, direct orders, vendor-items and other paths are disallowed for named search agents | Do not confuse private or transaction paths with public product discovery. |
| Sitemap reference | The file lists `https://www.coupang.com/sitemap/sitemap.xml` | A sitemap URL does not override the crawler directives above. |
Treat crawl rules and page observations as different boundaries
Coupang's robots file sets crawler directives; a product page shows what one market displayed at one time. Neither establishes the coverage of the other. For each record, retain the exact URL, market, field grain and observation time.
Scroll to compare →
| Evidence | What it can support | What it cannot establish by itself |
|---|---|---|
| Robots file | Crawl directives for named user agents and wildcard bots | Legal permission or an exception to account controls. |
| Buyer listing | Displayed product/catalog information at a particular time and market | A complete catalog or historical price/stock series. |
| Product URL identifiers | Which product page and option the observation describes | A seller's own product code or private inventory. |
Content reviewed 2026-10-24.