Mexico City foot traffic: comparing neighbourhoods for a new shop
Roma Norte, Polanco, Coyoacan or Centro? How to compare presence around candidate corners in Mexico City on equal terms before you sign for a shop.
Part of the guide Location intelligence automation
To compare Mexico City foot traffic across neighbourhoods for a new shop, measure the people-hours of presence around one candidate corner in each colonia, at the same radius and over the same weeks, then read the daily total next to the shape of the day, the busiest days and what already trades within a short walk. The city's districts work on very different clocks: an office strip in Santa Fe, a dining street in Condesa and a plaza in Coyoacan can show similar totals for entirely different reasons.
This is for retailers, cafe and restaurant founders, and franchise teams opening their first or next unit in the capital, and for the brokers who help them. It suggests how to choose comparison points, what to read, a made-up worked example, the mistakes newcomers make, and how an AI employee in xMap Flows does the measuring across a shortlist.
Why neighbourhoods in Mexico City need a like-for-like read
Mexico City is a patchwork of colonias, each with its own rhythm. A few patterns are worth keeping in mind before looking at any figures.
- Office districts. Areas such as Santa Fe and parts of Reforma fill on weekday working hours and empty in the evening and at weekends.
- Leisure and dining streets. Roma Norte, Condesa and Juarez tend to build through the afternoon and peak in the evening, with busy weekends.
- Historic and plaza districts. Centro Historico and Coyoacan draw tourists and families, often strongest at weekends and around midday.
- Upmarket shopping. Polanco mixes offices, hotels and prime retail, so presence is spread across the day.
- Residential colonias. Del Valle and similar districts carry a large share of presence early and late, much of it residents at home.
These are tendencies to test, not facts about any single corner. Two blocks apart, the picture can change completely, which is why a measured comparison beats reputation.
What to measure at each candidate
- Pick a real corner. Use the address of the unit you might rent, not the neighbourhood's centre. Presence on a side street is very different from the main avenue.
- Fix the radius. 150 to 200 metres for a shop that depends on passers-by. Keep it the same for every candidate.
- Use the same window. The last 30 days with data, read once for weekdays and once for weekends.
- Read four figures. The average day's people-hours, the split between morning, midday, evening and late night, the busiest hour, and the busiest and quietest days.
- Add the neighbours. Count rivals and anchors within a short walk: cafes, bakeries, metro and Metrobus stations, offices, markets, gyms.
A worked example (made up)
The brand and all figures here are invented. A Monterrey bakery-cafe wants its first unit in the capital and has three offers, measured at 150 metres over the last 30 days.
| Candidate | People-hours a day (weekdays) | Weekend vs weekday | Busiest hour | Bakeries and cafes within 300 m |
|---|---|---|---|---|
| Roma Norte side street | 19,000 | +35% | 18:00 | 17 |
| Polanco avenue | 27,000 | -10% | 13:00 | 11 |
| Coyoacan near the plaza | 15,500 | +60% | 12:00 | 9 |
Invented figures for illustration only.
The Polanco unit has the most presence but it falls at weekends and peaks at lunch, which suits office breakfasts and takeaway. Coyoacan is quieter on weekdays but its weekends are strong and it has fewer rivals, a good fit for a bakery that sells whole cakes to families. Roma Norte is crowded with competitors. The founder picks Coyoacan for the first unit and keeps Polanco as the second, for a grab-and-go format.
Mistakes newcomers make
- Trusting the neighbourhood's name. A famous colonia does not make every corner in it busy. Measure the address.
- Comparing an avenue with a side street at different radii. Same radius, every time.
- Ignoring the day of the week. A district that looks average on a weekly total can be the best in the city on Sundays.
- Counting rivals on a map by eye. Use a list with distances; units on upper floors and inside markets are easy to miss.
- Treating people-hours as customers. Presence is the ceiling, not the forecast. Your own conversion from a test or a pop-up is what turns it into sales.
Measuring the shortlist with an AI employee
If you are weighing more than two or three corners, or want to keep comparing as new offers arrive, an AI employee in xMap Flows can do the measuring every week. It is the kind of repeat task covered in the guide to an AI employee for site selection.
- Schedule: every Monday.
- My data: the candidate addresses your broker sends, in a table you add rows to.
- Mobility Footfall: footfall, busy hours and busiest days, 150 metres, weekdays; a second block for weekends.
- Nearby places: bakeries and cafes within 300 metres of each candidate, with a count and the nearest one.
- Sort and limit: order the candidates by weekday footfall.
- xMap LLM: one paragraph per candidate on how well it fits your format, then Email to the founders.
Each Monday the founders get a ranked table of every corner on the shortlist, its weekday and weekend presence, its busiest hour, its rival count and a short verdict. When the broker adds a new address to the table, it appears in the next week's email without anyone touching the flow.
Questions
Is there footfall data for Mexico City?
Yes. Footfall in xMap Flows covers most of the world, Mexico City included, as people-hours of presence around any address, with each day arriving about 10 days later.
Can I see which colonias a shop's visitors come from?
Not in Mexico City yet. Where visitors come from is measured in Sao Paulo only so far; elsewhere you can see how busy an area is and when.
What radius should I use for a shop on a busy avenue?
Around 150 metres keeps the figure close to the frontage. A wider circle on a major avenue or near a metro station can mix in presence that never passes your door.
How many weeks of data should I compare?
At least four full weeks per candidate, read for weekdays and weekends apart. The footfall history began in mid-October 2026, so longer comparisons will become possible as it grows.