A cafe is not a small restaurant. The economics are different enough that the location decision is different: a cafe lives or dies on a three-hour window, a small ticket size, and a rent bill that does not care about either.
A cafe makes most of its money before 2 pm. The deciding number is how many people walk past between 7 and 11 am on a weekday — not weekend brunch traffic, not evening footfall. A corner on the morning commute path — walking from transit toward offices, or from home toward transit — will consistently outperform an equally "busy" evening corner, because evening people are going somewhere with a bigger cheque.
When you look at foot-traffic data, ask which hours it covers. A daily total that is half lunch-and-evening does not help a cafe the way a morning count does. The free map's pedestrian layer, where cities publish it, is the place to start — see how foot traffic is measured.
The raw number of cafes already in a neighbourhood, set against its population, is the cleanest demand signal there is. The spread across Canadian cities is real and worth internalising: Toronto runs about 900 residents per cafe, Montréal 750, Vancouver 1,062 — but Calgary runs 3,403, Edmonton 3,381, and Winnipeg 6,815. A cafe opening in downtown Toronto is entering a very different room than one in Winnipeg.
But the citywide figure is almost useless on its own. The neighbourhood number varies by an order of magnitude across one city, and that spread is exactly where the opportunity (or the oversupply) is:
A Starbucks and an independent cafe are not the same competitor. Count them apart. An indie cafe opening next to a chain trades on being different; an indie opening next to three other indies is fighting for the same small ticket. The map lets you toggle chains and independents separately for exactly this reason.
Cafe tickets are small — a few dollars of margin per drink — so rent pressure shows up faster than it does in a restaurant. Target 6–10% total occupancy cost as a share of projected sales, and treat anything above 12% as a red flag. Most independents who fail, fail because the rent math didn't work on a realistic sales forecast. Run the numbers with the rent-to-sales guide.
A corner that has cycled through two cafes in three years is telling you the previous operators already did the market research you are about to do — and lost. Where licence records are public, that history is directly measurable; where they are not, the local business improvement association will tell you in five minutes. See the full six-factor framework for the rest of the checklist.