Where visitors come from in Sao Paulo, and how to use it

In Sao Paulo you can see where the trips to a place come from: the districts they start in and each one's share, how people travel (car or ride, public transport, walking), why they travel, and how far and how long the trip takes on average. It is the closest thing to a map of a shop's catchment drawn from real movement rather than a circle. Today this trip data covers Sao Paulo only; elsewhere you get footfall, but not origins.

This guide is for retailers, restaurants, clinics, gyms, mall operators and agencies working in Sao Paulo who want to know which districts feed a location. It covers what the trip figures are, how to read them, a made-up example, the mistakes to avoid, and an AI employee that reports them for each of your sites.

What the trip figures are

Mobility Footfall has a choice under Return called Where visitors come from. For each place it reads the trips that end in the area of the map the place sits in (a block a little under 1 km2) and adds:

• visitors_from: the districts those trips start from, with each one's share, for example "Pinheiros 31%, Butanta 14%". A district appears only when at least five trips come from it.
• trip_modes: how the trips travel, as shares.
• trip_purposes: why, in plain words, as shares.
• avg_trip_km and avg_trip_min: how far and how long the trips are on average.
• ride_hailing_pct: the share made by ride-hailing.
• trips_counted: how many trips the figures rest on.

Everything is an aggregate: no person, device or single trip is ever shown. The trips are counted for the whole small area, so a shop and its neighbour share the same figures. When the trips are many, the block reads a limited number of pages of them and says so in trips_note.

How to read it

1. Check trips_counted first. A few dozen trips give rough shares; a few thousand give steadier ones.
2. Read visitors_from as the order of your feeder districts, not as exact percentages. The top three usually tell the story.
3. Put avg_trip_km next to the districts. A short average with a nearby top district means a local trade; a long average means people come across the city for you.
4. Read trip_modes with the street in mind. A high share of car or ride trips says parking and drop-off matter; walking says the pavement and the corner do.
5. Compare like places. Two of your branches side by side show which one draws from further away.

A made-up example: two gyms

Invented figures for illustration only. A gym brand has a branch in Vila Mariana and is looking at a second one. It runs the trip figures on its branch and on a candidate building in Moema.

Invented figures for illustration; not real data for these districts.

Reading it: the existing branch is a walking-distance gym for its own district. The candidate draws people by car from a wider area that overlaps only a little with the first branch: Vila Mariana is 12% of its trips. So a second gym in Moema would mostly reach new members rather than split the old ones, and the building's parking matters more than it did in Vila Mariana. These trips end in the area, not at the gym, so they include people going to every other business there too.

Mistakes to avoid

• Reading it as your customers. The trips end in the area around the place, at all of its shops, homes and offices. They describe the area's pull, not your own customer list.
• Trusting small numbers. With few trips, one district's share can swing a lot. Look at trips_counted.
• Expecting it elsewhere. Today trips are measured in Sao Paulo only. Do not build a plan for Rio de Janeiro or Lisbon on these fields.
• Forgetting the delay. Like footfall, the trips arrive some days late, so the figures describe a window that ended about 10 days ago.
• Drawing a circle instead. A 3 km circle and the real feeder districts can be very different, because rivers, main roads and transport lines shape where people come from.

Report it with an AI employee

1. Schedule: the first Monday of every month.
2. My data with your Sao Paulo branches and any candidate addresses, or Nearby places to bring in a rival's branches.
3. Mobility Footfall: 30 days, with Footfall and Where visitors come from.
4. xMap LLM, one place at a time, asked to write one line on the top districts and how people arrive, and to say when trips_counted is low.
5. Show on a map and Email to the marketing and expansion teams.

Each month the inbox gets a line per location: the busiest-day footfall, the three feeder districts, the average trip and how people travel, with a map of the sites. For neighbourhoods to compare in the first place, the template score every neighbourhood for a new branch ranks them, and where to open next scores candidate sites.