Google Shopping Scraper: search by keyword, get prices and merchants

This google shopping scraper takes a search term the way a shopper would type it and returns the product grid as structured JSON: what each product is, what it costs as text and as a number, which shop is selling it at that price, and how it is rated. One row per product.

It also returns Google's own product identifiers, which is the part most shopping scrapers leave out. With those you can look a product up again later without matching on its title, which is the difference between a price history that tracks a product and one that tracks a string.

What you get

• The product grid for any search term, roughly 40 products per search, in the order Google ranked them (position).
• The price twice: price exactly as displayed (€ 469,00) and price_value as a number (469.0), so sorting and comparison need no parsing on your side, and the currency as its own field.
• The merchant selling at that price (merchant), not just the product.
• Ratings and review counts, with the count as a real number (6200) next to the label Google showed (6,2K).
• Google's product identifiers (catalog_id, cluster_id, merchant_id, offer_id), the handles for re-fetching a specific product later.
• Market targeting by country and language, so a Dutch query returns Dutch shops in euros and a US query returns US shops in dollars.
• No image blobs. Google ships each thumbnail inline as tens of kilobytes of base64; those are dropped so a row stays a few hundred bytes instead of bloating every export.

Why scrape Google Shopping

Google Shopping is the one place where a product's offers from many shops sit side by side, already matched to each other. A retailer's own site tells you what that retailer charges; the grid tells you where you sit against everyone else selling the same thing, which is the number that decides whether you get the sale.

The awkward part is the entry point. Most price data is keyed on a barcode or a product ID you have to know in advance, which works for your own catalogue and fails for everything else. Searching by keyword is how you find products you do not already have on file: a category you are considering entering, a competitor's new range, a term customers use that you have no product for.

The identifiers matter more than they look. A grid row is a snapshot, and a title is a bad key because shops write it differently and Google rewrites it over time. catalog_id names the product itself, so tomorrow's run can be compared against today's with confidence rather than by fuzzy-matching strings.

Input

Field	Type	Required	Default	Description
queries	array of strings	yes	—	One search term per line.
country	string	no	nl	Two-letter country code. Decides the market's shops and the currency.
language	string	no	nl	Two-letter language code for titles and labels.
maxResultsPerQuery	integer	no	40	Cap on products per term, 1 to 100.

Set country and language together. Asking for Dutch shops with English labels returns a mix that matches neither, and the currency follows the country.

Output

One row per product.

{
  "query": "sonos era 300",
  "country": "nl",
  "language": "nl",
  "position": 1,

  "title": "Sonos Era 300",
  "product_url": "https://www.google.nl/search?ibp=oshop&q=sonos%20era%20300&...",

  "price": "€ 469,00",
  "price_value": 469.0,
  "currency": "EUR",
  "merchant": "Amazon.nl - Retail",

  "rating": 4.7,
  "reviews": 6200,
  "reviews_label": "6,2K",

  "catalog_id": "3719662623030852347",
  "cluster_id": "7079904976121266844",
  "merchant_id": "576462703233852971",
  "offer_id": "3320132363282883836",
  "image_id": "17103979546685020772"
}

Use cases

Competitive price monitoring. Run your product names on a schedule and read price_value and merchant per run. What you get is not just the lowest price but who is holding it, which is the difference between "we are 8% over market" and "one specific reseller is undercutting us and nobody else is".

Finding products you do not have on file. Barcode-keyed price tools need you to know the product first. Searching by category term surfaces what is actually being sold and ranked for that term, which is how you scope a range before committing to it rather than after.

Building a price history that survives. Store catalog_id with every row and tomorrow's run lines up with today's on the product rather than on its title. Titles drift, get rewritten by shops, and differ per merchant; the identifier does not.

Market entry research. Run the same term with different country values and compare the merchant sets and price levels. A category crowded in one market and thin in another is a straightforward read from two runs.

How it compares

	This actor	Typical alternative
Entry point	Keyword search	Often a barcode or product ID you must already have
Price	As displayed and as a number, plus currency	Usually a string you parse yourself
Merchant	Returned per product	Sometimes missing
Review count	Number and original label	Label only, e.g. "6,2K"
Product identity	Google's catalog, cluster, merchant and offer ids	Title only
Image data	Dropped, product URL kept	Base64 blobs inflating every row

Among independent listings, automation-lab/google-shopping-scraper is the closest comparison at roughly 9,000 runs; its per-result price is tiered by Apify plan, so compare against your own tier. It returns title, price, merchant, rating and delivery, and does not expose Google's product identifiers.

If you already have barcodes rather than search terms, the sibling actor linked below is the cheaper route: it goes straight from EAN or SKU to offers without searching at all.

Pricing

Priced per product returned. Search terms that match nothing are reported in the run summary and cost nothing. All pricing is pay-per-event, so you only pay for results you receive. There are no per-compute-unit charges.

Limits and gotchas

• Google returns roughly 40 products per search term and there is no deeper pagination here. For exhaustive coverage of one product's offers, use the EAN/SKU actor below, which is built for depth rather than discovery.
• The grid is ranked by Google, not sorted by price. Sort on price_value yourself if you want cheapest-first; position reflects Google's ordering.
• price is one offer's price, the one Google surfaced on the card. It is not the lowest across all merchants for that product.
• Some cards carry no rating, and rating, reviews and reviews_label are then empty. That is normal for newer or low-volume products.
• reviews is parsed from a shortened, locale-formatted label. Where it cannot be parsed cleanly it is left empty rather than guessed, and the original label is always kept in reviews_label.
• Results are regionalised. The same term in nl and us returns different shops, different currencies and often different products.
• Free Apify plans are capped at 10 rows per run. Split larger lists across runs or upgrade to remove the cap.

FAQ

Can I scrape Google Shopping without an API key? Yes. Run it from the Apify Store or call it through the Apify API. No Google account or key is involved.

Does it return the lowest price for a product? No. It returns the offer Google put on the card. For every offer on a product, follow up with the EAN/SKU actor below using the catalog_id from the row.

Can I search by barcode instead of a keyword? You can pass one as a search term, but the sibling EAN/SKU actor is built for that and is the better route.

Why are some rows missing a rating? Google only shows ratings on products that have them. Those fields are left empty rather than filled with a placeholder.

How do I track the same product over time? Store catalog_id alongside each row and join on it between runs. Matching on the title breaks as soon as a shop rewrites it.

Related Actors

• Free Google Shopping Scraper for every offer on a product, straight from an EAN or SKU.
• Google Shopping Variants for the colour and size variants of a product.
• Google AI Product Answers for a written, sourced answer about a product rather than a price grid.