How to Compare Grocery Prices Across 16 Canadian Chains in One Call

Start here

A grocery list and a city. Paste this into the Input tab of the Canadian Grocery Price Comparison API, switch to JSON view, and hit Start.

{
  "queries": ["eggs", "chicken breast", "celery"],
  "location": "Vancouver, BC",
  "retailers": ["loblaws", "saveonfoods", "tnt"]
}

Same run over HTTP:

curl -X POST \
  "https://api.apify.com/v2/acts/sunny_eternity~canada-grocery-price-comparison/run-sync-get-dataset-items?token=YOUR_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"queries": ["eggs", "chicken breast", "celery"], "location": "Vancouver, BC", "retailers": ["loblaws", "saveonfoods", "tnt"]}'

Rows stream in as each retailer finishes. One query against three retailers gives you three rows:

The matching is the whole product

Grocery products don’t line up cleanly across chains. Brands differ, sizes differ, and every retailer names things its own way. So the Actor doesn’t claim a match is correct. It shows its work on every row:

• match_score runs 0 to 1.
• match_confidence buckets that into high, medium or low.
• match_reasons and mismatch_reasons name the signals behind the score, things like brand_match and size_match.

My advice: start at minMatchConfidence: "medium" and look at source_url on a handful of rows before you trust a whole basket. Low confidence rows are useful for coverage and bad for a price claim.

Basket totals per store

Set includeSummaries and each run emits three extra row types on top of the comparison rows:

{
  "queries": ["eggs", "milk", "bread", "bananas", "chicken breast", "rice"],
  "location": "Vancouver, BC",
  "retailers": ["loblaws", "saveonfoods", "tnt"],
  "includeSummaries": true,
  "minMatchConfidence": "medium"
}

• basket_summary totals the basket per retailer.
• item_best_options names the cheapest source for each item.
• run_meta carries per-retailer status and any warnings.

Schedule that same input weekly and you have basket totals, per-item winners and sale frequency over time, with no post-processing.

From Python

pip install apify-client, then:

from apify_client import ApifyClient

client = ApifyClient("YOUR_API_TOKEN")

run = client.actor("sunny_eternity/canada-grocery-price-comparison").call(run_input={
    "queries": ["eggs", "chicken breast", "celery"],
    "location": "Vancouver, BC",
    "retailers": ["loblaws", "saveonfoods", "tnt"],
    "minMatchConfidence": "medium",
})

for row in client.dataset(run["defaultDatasetId"]).iterate_items():
    if row.get("_type") != "comparison":
        continue
    print(row["query"], row["retailer"], row["price"], row["match_confidence"])

From JavaScript

npm i apify-client, then:

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: 'YOUR_API_TOKEN' });

const run = await client.actor('sunny_eternity/canada-grocery-price-comparison').call({
  queries: ['eggs', 'chicken breast', 'celery'],
  location: 'Vancouver, BC',
  retailers: ['loblaws', 'saveonfoods', 'tnt'],
  minMatchConfidence: 'medium',
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
const cheapest = items
  .filter((r) => r._type === 'comparison')
  .sort((a, b) => a.comparable_unit_price - b.comparable_unit_price);

Three things to know before you trust a number

• Unit prices need a shared basis. comparable_unit_price normalizes to whatever comparable_unit_basis names (100g, 100ml, each or lb). That’s what puts a 1.89 L Superstore carton and a 2 L Save-On-Foods one on one scale. Check is_comparable_unit_price before you sort on it.
• A partial run still returns. If one retailer fails, run_meta.retailer_status gives it a verdict of success, no_results, failed with a reason, or skipped. Read that before you conclude a chain was expensive when it actually returned nothing.
• Only six retailers run by default. tnt, loblaws, superstore, nofrills, saveonfoods and pricesmart. The other ten are opt-in, so name them in retailers if you want Quebec (maxi, provigo) or Atlantic Canada.

Output modes

outputMode changes the shape of what lands in the dataset. Start with comparison unless you know you want another.

If you want one chain in depth instead of a shallow slice of sixteen, the retailer guides go further: Loblaws and its eleven sibling banners, Save-On-Foods and the Pattison chains, or T&T.