WatchBase Scraper — Watch Reference Specs (Caliber, Case, Dial)

Turn a watch brand (or a list of specific WatchBase reference URLs) into clean, structured JSON: every model reference's caliber/movement, case material, bezel, glass, shape, dimensions, water resistance, and dial color, finish, indexes, hands — sourced from WatchBase, the reference-grade watch specification database.

Built for watch e-commerce, valuation/insurance, price comparison apps, and collection/inventory tools that already have a reference number (from a listing, an invoice, a trade-in) and need the full authoritative spec sheet behind it.

Why this actor exists: watch resale marketplaces (Chrono24) have several dedicated Apify scrapers — the top one has 16,000+ runs — proving buyers pay for watch data. But those scrapers only get listing prices. The reference-number → full spec layer those same buyers need to normalize and enrich that data had zero dedicated actors. This actor fills that gap.

---

What it does

• Input: one or more brand slugs (rolex, omega, patek-philippe, …) and/or specific reference-page URLs.
• Discovery: for each brand, the actor walks WatchBase's own hierarchy — brand → model family → individual reference — so you don't need to know reference numbers in advance.
• Output: one clean record per watch reference — brand, family, reference number, name, caliber/movement, case specs, dial specs, and a full flexible specs object with every field WatchBase publishes for that watch.
• Reliable by design: reads only public WatchBase pages (allowed by its robots.txt for standard browser traffic), plain server-rendered HTML — no login, no JS rendering, no CAPTCHA farms.

Input parameters

Field	Type	Default	Description
brands	array of strings	–	WatchBase brand slugs, e.g. ["rolex", "omega"]. Find the slug in the brand's URL: watchbase.com/<slug>. The actor scrapes every family and every reference for each brand.
referenceUrls	array of strings	–	Optional. Specific https://watchbase.com/<brand>/<family>/<reference> URLs, instead of or alongside whole brands.
maxItems	integer	0	Cap on total reference records returned across all brands. 0 = no limit.
includeMovement	boolean	true	Include caliber name, caliber page URL and movement/complications text.
includeCaseSpecs	boolean	true	Include case material, bezel, glass, back, shape, diameter, height, lug width, water resistance.
includeDialSpecs	boolean	true	Include dial color, finish, indexes, hands.
maxConcurrency	integer	5	Parallel requests. Keep low to stay polite to WatchBase.
proxyConfiguration	object	Apify Proxy	Proxy settings. Default datacenter proxy is enough.

Example input

{
  "brands": ["rolex"],
  "maxItems": 200,
  "includeMovement": true,
  "includeCaseSpecs": true,
  "includeDialSpecs": true
}

Output

One dataset record per watch reference. Real output for Rolex Daytona 126500LN-0001, captured live:

{
  "brand": "Rolex",
  "family": "Daytona",
  "reference": "126500LN-0001",
  "akaReference": null,
  "name": "Cosmograph Daytona Stainless Steel - Cerachrom / White / Oyster",
  "produced": "2023",
  "limited": false,
  "caliber": "Rolex caliber 4131",
  "caliberUrl": "https://watchbase.com/rolex/caliber/4131",
  "movement": "Rolex caliber 4131 Hours, Minutes, Small Seconds | Chronograph, Column wheel | Chronometer",
  "material": "Ceramic, Stainless Steel",
  "bezel": "Tachymeter",
  "glass": "Sapphire",
  "back": "Closed",
  "shape": "Round",
  "diameterMm": 40,
  "heightMm": 11.9,
  "lugWidthMm": 20,
  "waterResistanceM": 100,
  "dialColor": "White",
  "dialFinish": null,
  "indexes": "Stick / Dot",
  "hands": "Stick",
  "specs": { "...": "every raw label/value pair WatchBase shows for this reference" },
  "url": "https://watchbase.com/rolex/daytona/126500ln-0001",
  "scrapedAt": "2026-07-24T00:00:00.000Z"
}

Output field reference

Field	Description
brand / family / reference	Brand name, model family (e.g. Daytona), and the manufacturer reference number.
akaReference	Alternate reference number WatchBase lists (e.g. Rolex's M-prefixed full reference), if any.
name	Full model name / dial description.
produced	Production year or year range, as published (not always a clean single year).
limited	Whether WatchBase flags this as a limited edition.
caliber / caliberUrl / movement	Movement caliber name, WatchBase's caliber detail page, and the raw movement/complications text.
material, bezel, glass, back, shape, diameterMm, heightMm, lugWidthMm, waterResistanceM	Case specifications. Not every field is published for every watch (older/rare references have thinner data); missing fields are null.
dialColor, dialFinish, indexes, hands	Dial specifications.
specs	Every raw label → value pair from WatchBase's spec tables for this reference, unfiltered. Use this if you need a field not promoted to the top level, or the exact original text (e.g. "41.00 mm" instead of the parsed 41).
url	Canonical WatchBase reference URL.
scrapedAt	ISO timestamp of capture.

Pages that don't resolve are returned as { "url": "...", "error": "not_found" } so a bad reference never silently disappears from your run.

Typical use cases

• Marketplace / dealer listing enrichment: you have a reference number from a trade-in or a scraped listing (e.g. from a Chrono24 dataset) and need the full spec sheet — movement, case, dial — to build a complete product page.
• Valuation & insurance: authoritative case material, diameter and movement data to normalize condition/value assessments across references.
• Watch comparison / recommendation apps: structured, comparable specs across brands instead of scraping marketing copy.
• Collection & inventory tools: bulk-hydrate a collection by brand instead of looking up each reference by hand.

How to choose brands

Browse watchbase.com/watches for the full brand list, or use the brand slug from any WatchBase URL you already have (watchbase.com/<slug>). Popular brands: rolex, omega, patek-philippe, audemars-piguet, cartier, breitling, tudor, seiko, tag-heuer.

Limitations (read before buying)

• Field coverage varies by watch. WatchBase's data depth differs by reference (a current Rolex has more fields than a 1960s vintage piece); missing fields are null, not fabricated.
• specs keys vary slightly across watches (e.g. Material vs Materials). The flattened top-level fields normalize the common cases; specs has the raw truth.
• Discovery is brand → family → reference, WatchBase's own navigation. There is no keyword search input — this keeps the actor on stable, allowed content pages.
• A full brand (e.g. Rolex, ~30 families × 100+ references each) can be several thousand records. Use maxItems to cap a run, or pass specific referenceUrls if you only need a handful.

FAQ

Do I need a WatchBase account or API key? No. The actor reads public reference pages only.

Is this allowed? The actor requests only public content pages, which WatchBase's robots.txt permits for standard browser traffic, at low volume. It collects factual product specifications — no personal data, no logged-in content.

What if a reference doesn't exist / the page is empty? You get a record with "error": "not_found" for that URL; the rest of the run continues.

Can I scrape just one specific watch instead of a whole brand? Yes — use referenceUrls with the exact WatchBase reference URL(s).

How many watches can I scrape? As many as exist for the brands you request. Use maxItems to cap large brands, and keep maxConcurrency modest (default 5).