nike

name	nike
description	Proven scraping playbook for nike.com. Skip the browser entirely — replay the internal api.nike.com product_wall endpoint directly with fetch. Public API, no cookies, no tokens, no anti-bot. Requires the nike-api-caller-id header and count=24 per page. Activate for any nike.com gridwall (/w/ URLs). Covers endpoint template, pagination, response shape, and known API quirks.
metadata	{"author":"torch","version":"1.0.0"}

Nike (nike.com)

Custom React SPA with SSR bootstrap ("Web Shell"). The product grid hydrates from an internal api.nike.com/discover/product_wall/v1 JSON API. There is no __NEXT_DATA__ blob; only the first ~24 cards are server-rendered, the rest load via XHR on scroll.

Skip the browser entirely. Replay the API with fetch. It's public, fast, and anti-bot-free — just send the right headers.

Detection

Signal	Value
Framework	Custom React SPA with "Web Shell" SSR
API	api.nike.com/discover/product_wall/v1 — public JSON, cursor-paginated
Anti-bot	None on the API (for browsing). Akamai Bot Manager on www.nike.com, but the API bypasses it
Auth	Not required for catalog browse. SNKRS and checkout are gated.
Rate limit	None observed at ~7 req/s across 132 pages. 150ms sleep between pages is plenty.

Strategy

Skip browser entirely
  → GET api.nike.com/discover/product_wall/v1/... with nike-api-caller-id header
  → Follow pages.next cursor until empty
  → Flatten productGroupings[].products[] to one row per colorway

Never launch Puppeteer for Nike catalog scraping — the API is faster, cleaner, and has no rate limits.

Endpoint template

US English "new releases" gridwall (/w/new-3n82y):

GET https://api.nike.com/discover/product_wall/v1/marketplace/US/language/en/consumerChannelId/d9a5bc42-4b9c-4976-858a-f159cf99c647
    ?path=/w/new-3n82y
    &attributeIds=53e430ba-a5de-4881-8015-68eb1cff459f
    &queryType=PRODUCTS
    &anchor=0
    &count=24

Parameters:

Param	Value	Notes
path	The URL slug (e.g. /w/new-3n82y)	Identifies the gridwall
attributeIds	GUID for the category filter	Sniff once per category (see Gotchas)
queryType	PRODUCTS	Always
anchor	0, 24, 48, ...	Pagination offset, increments by count
count	24	Required — larger values return HTTP 400

For other gridwalls, swap path and attributeIds to the target category's values. Find them by sniffing the page once in devtools.

Required headers

const headers = {
  "nike-api-caller-id": "nike:dotcom:browse:wall.client:2.0",
  "anonymousid": "a".repeat(32),  // any 32-char hex works; a static one is fine
  "accept": "*/*",
  "origin": "https://www.nike.com",
  "referer": "https://www.nike.com/",
  "user-agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36",
};

Without nike-api-caller-id the API returns:

{ "error": "NIKE_API_CALLER_ID_HEADER_NOT_PRESENT" }

anonymousid can be any hex string; the API only checks presence, not validity.

Pagination

Follow pages.next (relative URL) until empty or null:

let url = `https://api.nike.com/discover/product_wall/v1/marketplace/US/language/en/consumerChannelId/d9a5bc42-4b9c-4976-858a-f159cf99c647?path=/w/new-3n82y&attributeIds=53e430ba-a5de-4881-8015-68eb1cff459f&queryType=PRODUCTS&anchor=0&count=24`;

const allProducts = [];
const seen = new Set();

while (url) {
  const res = await fetch(url, { headers });
  if (!res.ok) throw new Error(`HTTP ${res.status}`);
  const data = await res.json();

  for (const group of data.productGroupings ?? []) {
    for (const p of group.products ?? []) {
      if (seen.has(p.productCode)) continue;
      seen.add(p.productCode);
      allProducts.push(flatten(p));
    }
  }

  url = data.pages?.next
    ? `https://api.nike.com${data.pages.next}`
    : null;

  await new Promise((r) => setTimeout(r, 150));
}

pages.totalPages and pages.totalResources tell you how many cards remain — useful for progress logging.

Response shape

{
  productGroupings: [
    {
      cardType: "default",
      products: [
        {
          productCode,
          copy: { title, subTitle },
          prices: { currency, currentPrice, initialPrice, discounted, discountPercentage },
          displayColors: { simpleColor: { name, hex }, colorDescription },
          colorwayImages: { portraitURL, squarishURL },
          pdpUrl: { url, path },
          badgeLabel,
          featuredAttributes: [...],
          productType,      // "FOOTWEAR" | "APPAREL" | "EQUIPMENT" | "STORED_VALUE"
          productSubType,
          styleColor,
          groupKey,
          globalProductId,
          isNewUntil,
          // ...
        },
      ],
    },
    // ...
  ],
  pages: { next, prev, totalPages, totalResources },
}

Each productGroupings[*] is a style group; its .products[] are individual colorway variants. Flatten to one row per colorway and dedupe on productCode.

Anti-blocking summary

Technique	Needed?	Notes
Stealth plugin	❌ N/A	No browser
nike-api-caller-id header	✅ Yes	Required or API returns error
anonymousid header	✅ Yes	Any 32-char hex; presence-checked only
Realistic UA	⚠️ Recommended	Standard desktop Chrome UA
Residential proxies	❌ No	Datacenter IPs work fine
Rate limiting	❌ No	7 req/s across 132 pages, no throttle observed
Session cookies	❌ No	Stateless API

Note: curl -I (HEAD) returns HTTP 400 because CORS preflight trips. Use GET only.

Data shape (output rows, flattened)

{
  "title": "Air Max 90",
  "subtitle": "Men's Shoes",
  "productCode": "FB9658-101",
  "styleColor": "FB9658-101",
  "groupKey": "FB9658",
  "globalProductId": "a1b2c3...",
  "productType": "FOOTWEAR",
  "productSubType": "SNEAKERS",
  "currency": "USD",
  "currentPrice": 130.0,
  "initialPrice": 130.0,
  "discountPercentage": 0,
  "color": "White",
  "colorHex": "#FFFFFF",
  "colorDescription": "White/Black/Cool Grey",
  "badge": null,
  "badgeAttribute": null,
  "isNewUntil": "2026-05-15T00:00:00Z",
  "featuredAttributes": [],
  "portraitImage": "https://static.nike.com/a/images/.../portrait.png",
  "squarishImage": "https://static.nike.com/a/images/.../squarish.png",
  "url": "https://www.nike.com/t/air-max-90-mens-shoes-..."
}

Known gridwall sizes

Gridwall	path	attributeIds	Pages	Groupings	Colorways
New releases	/w/new-3n82y	53e430ba-a5de-4881-8015-68eb1cff459f	~132	~3,148	~4,800

Flattened colorway count is ~1.5× the totalResources because totalResources counts groupings, not variants.

Gotchas

1. count > 24 → HTTP 400. Stick to 24. This is hard-capped server-side.

2. attributeIds is category-scoped — every /w/... URL has a different GUID. Sniff it once via devtools → Network when you switch category, then hardcode.

3. totalResources is the grouping count, not colorway count. Expect ~1.5× more rows after flattening. Don't use it for progress % without adjusting.

4. DOM-scraping caps at ~72 cards. The scroll observer stops firing in headless mode after three batches. Always prefer the API — this is why we skip the browser entirely.

5. HEAD returns 400 — CORS preflight trips on HEAD requests. Use GET only, even for probing.

6. anonymousid is dummy-checked — the API only verifies presence, not validity. A static string works across all requests.

7. SNKRS is different — this playbook covers www.nike.com/w/* catalog gridwalls. SNKRS (www.nike.com/launch) has its own API with real auth, device fingerprinting, and Akamai _abck/bm_sz cookie gating. Do not use this playbook for SNKRS.

Output

./output/nike.json