How WellValet Closes the Canadian Grocery Data Gap

Last month we published our research into how well existing product databases cover the Canadian grocery aisle. It ended with a promise: build a flow where a scan that comes up empty, or comes back with no ingredient list, becomes an invitation to fix that instead of a dead end. That flow has been live in the app for several weeks now. Here's how it actually works.

From a blank scan to a data entry point

When WellValet scans a barcode and either finds nothing or finds a product with no usable ingredient list, the app now prompts the shopper standing in front of it to point the camera at the printed ingredient panel instead. That's the same OCR ingredient scanner already built into the app for unbarcoded items, repurposed as the entry point for a missing product.

The photograph doesn't go straight into the live database. It's queued centrally, reviewed and curated by our team, and then merged into the same shared, Open Food Facts–based dataset that every WellValet install draws from. The next Canadian shopper who scans that exact barcode — whether that's a minute later or a month later — gets a real answer instead of another blank scan.

It's a small, unglamorous mechanic, and that's the point: it turns ordinary use of the app into ongoing maintenance of the dataset behind it, without asking anyone to do anything they weren't already doing.

Why most scans succeed before that flow ever kicks in

Canada's grocery retail sector is unusually concentrated by developed-market standards.

Loblaw, Sobeys (Empire), Metro, Costco and Walmart between them dominate the shelves most Canadians actually shop from week to week. Because those chains' private-label lines and major national brands are exactly the products most likely to already be well-documented in a shared, community-maintained database, the products people scan on an ordinary shopping trip disproportionately land in the part of the dataset that's already complete. In our own usage, that shows up as most scans simply working on the first try, before the enrichment flow above is ever needed.

That's not a claim that every product in every store resolves instantly — a corner store's imported specialty item is a different story than a Loblaws-brand pantry staple, and it's exactly that long tail of smaller and regional products where the scan-and-curate flow above earns its keep.

An ongoing pipeline, not a finished project

We're not going to claim this closes every gap for good. New products land on Canadian shelves constantly, and any crowdsourced or community-maintained dataset is chasing a moving target by design — Open Food Facts itself grows and changes daily. What's changed on our end is the direction of travel: the gap now narrows a little with every scan that comes up short, instead of sitting there until someone manually re-imports a dataset months later.

We build WellValet, a Canadian grocery barcode scanner. If you want the full method and numbers behind the database this flow feeds into, read our original database coverage research. And if you have thoughts on how we could do this better, we'd like to hear them: support@wellvalet.com