Case study

Cafés in Prague 2: where homes outnumber cafés

Within 2 km of Náměstí Míru, OpenStreetMap lists 520 cafés. Compared with where people live, they are very unevenly spread: Vršovice has about 21 times more housing per café than the Old Town, and the three spots where a new café would draw the most customers all lie east and south-east of Náměstí Míru, towards Vršovice. This page shows the data, the method and its limits.

Updated 23 October 2026 · By AYKA TECHNOLOGIES

The map

Map data © OpenStreetMap contributors, 23 October 2026. Each square is about 140 m wide.

How much housing is there per café in each neighbourhood?

Each block of housing and each café is counted in its nearest neighbourhood. The housing index adds up residential buildings, with apartment blocks weighted more than houses; it compares areas with each other and is not a population count.

Vršovice has 17 cafés for a housing index of 13,103: about 771 per café. The Old Town has 125 cafés for 4,516, about 36 per café. That gap does not mean the Old Town has too many cafés: its customers are mostly visitors and office workers, whom housing does not measure. It does mean that residents of Vršovice and Vyšehrad have far fewer cafés close to home.

Where would a new café draw the most customers?

We tested candidate locations one by one with the Huff model, which estimates the chance that a resident chooses one place from its distance and from every competitor nearby. The three spots with the highest expected share:

1. Spot 1: 1.7 km south-east of Náměstí Míru (50.06219, 14.44817). Expected share of the area’s demand: about 0.7%.
2. Spot 2: 1.6 km south-east of Náměstí Míru (50.06752, 14.45646). Expected share of the area’s demand: about 0.6%.
3. Spot 3: 1.7 km east of Náměstí Míru (50.07151, 14.46061). Expected share of the area’s demand: about 0.6%.

For comparison, a new café on Náměstí Míru itself would get about 0.26%, because 20 cafés are already within an estimated 5-minute walk and 71 within 10 minutes. Spot 1 does about 2.7 times better. A real decision also needs rent, footfall, visibility and the concept, which this model does not see.

How many residents can one café reach on foot?

A café on Náměstí Míru has 16% of the area’s housing within an estimated 10-minute walk. Four well-placed cafés would cover 55%, which shows how much the area depends on many small, local places rather than a few large ones.

What does this mean for a café owner?

• Opening: compare candidate addresses by the customers they add, not by how busy the street looks.
• Google Business Profile: name as service areas the neighbourhoods where you are the closest choice for most homes.
• Local pages: in Czech and English, for the neighbourhoods you serve, with directions, opening hours and what makes you different.
• AI assistants: check whether they name you for your own area, for example “kavárna Vršovice” or “café near Náměstí Míru”, with the method in our AI visibility guide.

Method and limits

• Data: cafés (amenity=cafe), residential buildings and neighbourhood names from OpenStreetMap, read on 23 October 2026 within 2 km of Náměstí Míru.
• Model: Huff model with attractiveness exponent 1 and distance decay 2 (power), every café equally attractive; walking time estimated from distance at 4.8 km/h with a 1.25 detour factor.
• Demand: residential buildings, weighted by type. Visitors, students and office workers are not included, which is why central areas look over-supplied.
• Neighbourhoods: blocks and cafés are assigned to the nearest neighbourhood point, so borders are approximate, not official district borders.
• Completeness: OpenStreetMap is very complete in Prague, but a café that is not mapped is not counted.
• The shares are estimates for comparing places, not sales forecasts.

The same analysis works for clinics, pharmacies, gyms or shops, and with your own anonymised customer data it becomes much more precise. See location and catchment analysis.

Free catchment check: Free catchment check: how many competitors are within walking distance?