walmart-keyword-search

Open the Walmart search URL with your keyword and page number in a browser-act session, wait for the page to stabilize, then run the extraction script. It reads __NEXT_DATA__ from the page and returns structured JSON with itemId, price, rating, and availability for each product.

Each item includes itemId, canonical URL, title, brand, image, price, wasPrice, rating, reviewCount, availability status, seller name, fulfillment badge, classType, and shortDescription. Note that brand and shortDescription are often null in search listings.

Yes, the search URL accepts a sort parameter with values best_match, price_low, price_high, rating_high, and new. Append it to the search URL before running extraction to get results in the desired order.

Walmart caps search pagination at roughly 11 to 25 pages regardless of total result count. The extraction response includes a maxPage field, so stop iterating when the page number exceeds it or when itemCount returns zero.

Run keywords serially within one browser session with 1-2 second delays between navigations, and never parallelize within a single browser. For higher throughput, distribute work across multiple stealth browser sessions, each with an independent fingerprint.