Coles Scraper

Scrape Coles Australian grocery products, prices, specials, and categories from public pages and export Coles data to CSV, JSON, or Excel. This Coles API alternative turns repeatable inputs into structured results for Australian grocery assortment and price research.

Each dataset row represents a public result or a documented alternate record type. The output includes source identifiers and links where available, so you can verify a row, compare later runs, and distinguish missing source data from an omitted column.

What data can you extract from Coles?

The dataset schema defines every field below. Examples come from one successful published run; null means that field was not exposed for that particular record. Availability may differ by input mode, page, and record type.

Field	Description	Example
recordType	Product or category row.	product
operation	Operation that produced the row.	category
query	Search keyword.	null
page	Native Coles page number.	1
productId	Coles product identifier.	8150288
productUrl	Canonical product page URL.	https://www.coles.com.au/product/coles-full-cream-milk-3l-8150288
name	Product name.	Full Cream Milk
brand	Brand.	Coles
description	Short product description.	COLES FULL CREAM MILK 3L
longDescription	Detailed product description.	null
size	Pack size.	3L
imageUrl	Product image URL.	https://cdn.productimages.coles.com.au/productimages/8/8150288.jpg
price	Regular price in Australian dollars; current price if no reduction.	4.95
currentPrice	Current price in Australian dollars.	4.95
discountPrice	Current reduced price, if a was price is shown.	null
wasPrice	Previous price when reduced.	null
currency	ISO currency code.	AUD
unitPrice	Comparable unit price.	1.65
unitMeasure	Unit for comparable pricing.	l
unitPriceText	Displayed comparable price label.	$1.65/ 1L
promotionType	Promotion classification.	null
specialType	Special classification.	null
offerDescription	Displayed offer text.	null
savePercent	Discount percentage when supplied.	null
isOnlineSpecial	Whether the offer is online special.	false
available	Availability in the anonymous shopping context.	true
availabilityType	Online or store availability type.	InStoreAndOnline
availableQuantity	Reported available quantity.	2440
retailLimit	Maximum retail quantity.	20
category	Product category.	Milk
subCategory	Top level department.	Dairy, Eggs & Fridge
aisle	Product aisle.	Full Cream Milk
categoryId	Coles category identifier.	8881800
gtin	Global trade item number from detail page.	null
ingredients	Ingredient text from detail page.	null
allergens	Allergen text from detail page.	null
servingSize	Nutrition serving size.	null
energyKjPer100	Energy per 100 g or ml.	null
proteinPer100	Protein per 100 g or ml.	null
fatPer100	Total fat per 100 g or ml.	null
sugarsPer100	Total sugar per 100 g or ml.	null
sodiumPer100	Sodium per 100 g or ml.	null
countryOfOrigin	Product origin statement.	null
lastUpdated	Source product update timestamp.	null
categoryUrl	Category browse URL.	null
categorySlug	Category path slug.	null
categoryName	Category name derived from path.	null

Use the identifier and URL fields as your join keys before comparing snapshots. Fields that describe a page, search, category, author, or tournament establish where the record came from; keep them when you export a subset. Numeric values and boolean flags reflect the public page at collection time, not a permanent claim about the underlying item.

How to scrape Coles

1. Open the actor input form and choose a narrow public source: a search phrase, page URL, item URL, or ID supported by the input fields below.
2. Set the relevant per-source page or result limit and the total item limit. Start with a small sample to check which optional fields the public source exposes.
3. Run the actor. Its dataset contains one structured row per saved result; inspect the first rows and their source links for the chosen input.
4. Download the dataset as JSON, CSV, or Excel. Preserve identifiers when combining runs so repeated results can be deduplicated.

For this actor, choose search, category, specials, product, or category-discovery operation according to the input schema. Product records and category records have different populated fields; inspect recordType before comparing prices.

Input

Use only the parameters relevant to your collection mode. An omitted optional filter uses the schema default shown here; an empty array, zero, and an omitted value can have different meanings, so keep intentional settings in your saved input.

| Parameter | Type | Default | Description | | --- | --- | --- | | operation | string | search | Choose product search, specials, a category shelf, individual items, or category discovery. | | query | string | No default; example milk | Product keyword for search. | | special | string | all | All specials or half-price offers when operation is specials. | | categoryUrl | string | No default; example https://www.coles.com.au/browse/dairy-eggs-fridge/milk | Full Coles browse URL for category operation. | | productUrls | array | No default; example ["https://www.coles.com.au/product/coles-full-cream-milk-3l-8150288"] | Full Coles product URLs for item operation. | | startPage | integer | 1 | First native Coles page for search, specials, or category. | | maxPages | integer | 1 | Maximum native pages to fetch from the start page. | | maxItems | integer | 100 | Stop after this many dataset rows; 0 means no cap. | | details | boolean | false | Fetch each listing product page for barcode, ingredients, allergens, and nutrition. | | proxyConfiguration | object | {"useApifyProxy":true} | Proxy settings for remote access; the actor escalates to an Australian residential proxy when blocked. |

Example input based on the published golden run (long URL lists are shortened):

{
  "operation": "search",
  "query": "milk",
  "startPage": 1,
  "maxPages": 1,
  "maxItems": 50,
  "details": false
}

The example is a starting shape, not a guarantee of a particular result count. Source inventory and page accessibility change. When you need repeatable comparisons, save the exact input JSON with the run date and inspect the returned source or record-type field.

Output example

The following is one real item from a successful published dataset. Long text and media arrays are shortened for readability; the actual dataset keeps the original values and all schema fields.

{
  "recordType": "product",
  "operation": "category",
  "page": 1,
  "productId": "8150288",
  "productUrl": "https://www.coles.com.au/product/coles-full-cream-milk-3l-8150288",
  "name": "Full Cream Milk",
  "brand": "Coles",
  "description": "COLES FULL CREAM MILK 3L",
  "size": "3L",
  "imageUrl": "https://cdn.productimages.coles.com.au/productimages/8/8150288.jpg",
  "price": 4.95,
  "currentPrice": 4.95,
  "currency": "AUD",
  "unitPrice": 1.65,
  "unitMeasure": "l",
  "unitPriceText": "$1.65/ 1L",
  "isOnlineSpecial": false,
  "available": true,
  "availabilityType": "InStoreAndOnline",
  "availableQuantity": 2440,
  "retailLimit": 20,
  "category": "Milk",
  "subCategory": "Dairy, Eggs & Fridge",
  "aisle": "Full Cream Milk",
  "categoryId": "8881800"
}

This row illustrates the observed output structure, including its identifiers and public links. Empty values elsewhere in the dataset should be interpreted field by field; a field shown in this example is not promised for every result.

Use cases

• Grocery analysts can compare product and unit prices across recurring category snapshots.
• Brands can monitor public assortment, promotion labels, and shelf visibility.
• Nutrition researchers can compile available ingredients, allergens, and per-100 values for selected products.
• Retail operations can reconcile product IDs, barcodes, and availability signals.

The strongest analyses keep source context. A field such as price, rating, engagement, or rank has meaning only with its associated item, query, date, and public URL. Keep the raw export and create a separate cleaned view for charts or alerts.

How much does it cost to scrape Coles?

The price is $1.00 per 1,000 saved results. Platform usage is included. The charge scales with output rows, so a restrictive filter or inaccessible page can produce fewer billable results than the requested maximum.

• 100 results: $0.10. This is useful for checking a small cohort and confirming which optional fields are present.
• 1,000 results: $1.00. This is the reference price for a larger export.
• 5,000 results: $5.00. Reaching this size may require multiple focused sources or scheduled runs, depending on public inventory and source caps.

Compute any other estimate as saved result count × $1.00 / 1,000. A maximum input is a ceiling, not a purchase of that many rows. For planning, use the actual saved-item count from an initial representative run.

Tips for best results

For comparable prices, keep a consistent operation and query across runs. Use productUrls for known items, and details when ingredients or nutrition matter. Compare unitPrice with unitMeasure rather than raw package price alone.

Collect a small baseline first, record the exact input and date, and inspect both a typical row and a sparse row. Expand by adding focused sources instead of assuming one broad input can reveal the full public inventory. When comparing two runs, match stable IDs or canonical URLs and use the same filters so changes reflect the source rather than a changed query.

Export the full JSON when nested arrays or objects matter. CSV and Excel are convenient for sorting and joins, but nested structures may need flattening before spreadsheet analysis. Keep numeric fields numeric and preserve source URLs as text; do not infer a zero from a null.

Limits

Retail prices and availability can depend on storefront context and change after collection. Nutrition, allergens, offers, and origin are only populated where the public product data exposes them.

Public pages can change, disappear, or expose different fields for different records. The schema is a list of possible output columns, not a promise that each column is filled in each row. The maxItems-style input limits cap saved results; they do not bypass source pagination, public visibility, or a site-specific result ceiling.

Treat a saved result as a snapshot. If a later run returns fewer rows, first compare the input, source access, and public inventory before concluding that the underlying market changed. If you need an audit trail, retain the source URL, stable identifier, and run timestamp with the export.

Use with AI agents (MCP)

An agent can supply the documented input JSON, run this actor, and work from its dataset. Ask it to keep source links and identify null values explicitly when summarizing results. Example prompts:

Run Coles Scraper for the sample input above. Return the first 20 results with their source URLs and the fields needed for Australian grocery assortment and price research. Mark unavailable fields as null.

Schedule a repeat Coles collection with the same filters. Compare records by stable ID or canonical URL and report only new, removed, or changed public values.

FAQ

Is it legal to scrape Coles?

This actor collects public data. Check the site’s terms, applicable law, and the rights of people whose data appears in your export. Follow GDPR and other privacy rules where relevant; do not use personal data for misuse, intrusive profiling, or unauthorized contact.

Do I need to configure proxies?

The input includes proxyConfiguration and defaults to Apify Proxy. Usually the default is enough to start. Availability can vary by page and region; changing network settings cannot make private or login-only information public.

How fast will a run finish?

Duration depends on the number of inputs, pages, detail requests, and source responses. A small sample is the best way to measure your workload. Increase the limits gradually, then use the observed run duration for scheduling and monitoring.

Can I schedule and monitor recurring runs?

Yes. Save the input and schedule recurring runs in Apify. Monitor run status, item count, and any missing-field changes; store the run date alongside exports when comparing snapshots.

Can I export to Google Sheets or Excel?

Yes. Download CSV or Excel from the dataset, or pass JSON to a spreadsheet integration. For nested fields, flatten the specific child values you need rather than losing the original JSON.

What if a field is empty?

An empty or null field means it was not available from that public source record in that run. Check the source URL, input mode, and record type. Do not replace missing prices, counts, dates, or flags with zero unless your own analysis has a documented rule for doing so.

Are search and category rows identical?

No. recordType and operation identify the source. Category discovery rows describe taxonomy, while product rows carry prices and product attributes.

Integrations

Use the Apify API to start runs and retrieve the dataset, webhooks to react when a run finishes, or Zapier, Make, and n8n to route records into other systems. Google Sheets supports lightweight review, while scheduled runs provide repeat snapshots. Keep the raw JSON when downstream workflows need nested fields or exact null values.

Support

Open an issue on the Issues tab; we reply within 24h and add fields on request.