XMap

26/08/2026

We grouped every New York hexagon by how many subway stations it contains, then looked at how its population changes across the day.

Zero stations: 1.00x at midday.
One or more: 1.16x.
Three or more: 1.71x.

Read the first line again.

Averaged across 642 hexagons, a New York neighbourhood with no subway station does not change size during the day. Not slightly. At all. Its 24 hour curve is a flat line.

I expected a gradient. I did not expect the zero-station group to be perfectly flat.

Here is what I think is actually being measured. Station count is not a proxy for transit quality. It is a count of how many independent flows of people a place can receive. One station connects you to a line. Three connects you to a network, and daytime population climbs 71 percent.

The practical version:

Before you look at rent, before demographics, before competitive density, count the stations in the cell. In New York that single variable separates locations with a daytime audience from locations that structurally cannot have one.

It takes about thirty seconds and it is free.

21/08/2026

Ranking São Paulo districts by daytime footfall gives you one list.

Ranking them by evening footfall gives you a completely different one.

Evening index, 8pm to 10pm against the overnight baseline:

República: 1.46x
Sé: 1.41x
Bom Retiro: 1.32x
Consolação: 1.30x
Pari: 1.22x

República is close to as busy at 9pm as it is at noon. For late-trading retail, food service, security staffing, or out-of-home media, that is your shortlist and it looks nothing like the daytime one.

Now the part I want to be straight about.

Our dataset ends at 11pm. The nightlife-dense hexagons were still climbing when it stopped. So the real evening peak in these districts sits outside the window we measured, and our numbers understate it.

Same with our density figures. The public map bands its values and the top band is open-ended, which means the busiest blocks are capped at the top of the scale. Every gain we reported for Itaim Bibi and Bom Retiro is conservative.

I would rather publish the limits than publish a number that looks cleaner than the evidence supports.

If you work with footfall or mobility data, you already know this is where most published analysis quietly falls apart.

20/08/2026

One variable in São Paulo separates locations that gain people during the day from locations that lose them.

Not income. Not resident density. Not district.

A metro station.

Hexagons with at least one station: 1.26x at midday, rising to 1.28x at 5pm.
Hexagons with none: 0.93x.

Those two groups sit on opposite sides of the break-even line for the entire working day. One imports people. The other exports them.

The gap is 33 percentage points.

For context, of the 2,281 hexagons we analysed, 1,307 have no station and 82 do.

There is a version of this post that ends with "so build near transit," which is not insight, it is a bumper sticker. The more useful read is this:

If you are evaluating two sites with similar rent, similar resident demographics, and similar competitive density, transit adjacency is doing more work than any of those variables to determine whether your daytime audience exists at all.

And it is the one you can check in about thirty seconds.

19/08/2026

A Jardim Paulista block gains 2,100 people by lunchtime.

It has six restaurants and one café.

We found this by isolating the 177 São Paulo hexagons that gain more than 2,000 people between overnight and midday, then dividing midday population by the number of food and beverage venues inside each one.

The spread is not subtle.

República: 25 people per venue. Saturated.
Consolação, Sé, Santo Amaro: 38 to 40. Mature.
Individual blocks in Pinheiros, Moema, Morumbi: 300 to 1,100.

One Moema hexagon takes in 3,270 extra people against twelve restaurants.

These are not underserved because they are poor or empty. They are high-value catchments where the daytime population showed up and the supply never followed.

That is a site selection shortlist. Two data layers, no fieldwork, one afternoon.

The uncomfortable part for anyone doing expansion planning: this analysis is not hard. It is just not what most teams look at, because most teams are still working from resident population.

18/08/2026

What is built in a place tells you when that place is busy.

We tested this across 2,281 São Paulo hexagons by pairing each one's hourly footfall curve with its venue mix.

The separation was clean.

Office-dense areas: 1.28x at midday. Sharp ramp from 6am.
Retail-dense areas: 1.20x at midday, but flat for twelve straight hours.
Logistics areas: already at 1.15x by 6am, before anyone else has moved.
Residential areas: 0.86x. They empty.

The one that surprised us:

Office hexagons do not have one peak. They have two.

Midday hits 1.32x. But 5pm hits 1.29x, which is essentially the same number.

That is the return commute and the after-work window stacking on top of each other. If you are staffing a business in an office cluster, scheduling deliveries, or weighting ad dayparts, the 5pm hour is worth roughly what noon is worth.

Most planning treats lunch as the event and 5pm as the wind-down.

The data says they are the same size.

17/08/2026

São Paulo has 11.4 million people at 4am.

At 1pm it has 10.9 million.

Almost no change. Which sounds like a boring finding, until you look at where those people are standing.

Roughly 840,000 of them are somewhere completely different.

We analysed 2,281 hexagons of hourly footfall across all 96 districts of the municipality. One Tuesday, every hour, 0.7 km² per cell.

Barra Funda nearly doubles between 4am and 1pm. A pull index of 1.96x.
Bom Retiro: 1.70x.
Sé: 1.58x.

Meanwhile São Rafael and Jardim Ângela lose close to a quarter of their people.

The city does not get bigger during the day. It gets rearranged.

This matters because almost every dataset used for commercial decisions describes where people sleep. Census counts, resident demographics, district population.

But you do not open a store for people who sleep nearby. You open it for people who are standing there at 1pm on a Tuesday.

In São Paulo those two numbers differ by a factor of two in exactly the districts where the money is.

Full analysis in the comments.

07/08/2026

Find your next high-potential location using real visitor activity.

Before opening a store, restaurant, franchise, or commercial property — understand the people around it.

Click any location on the map and discover:
✓ Visitor counts by time of day
✓ Area activity patterns
✓ Location popularity
✓ Customer movement insights

Built for:
Retail expansion teams, real estate developers, franchise operators, and location strategy professionals.

Explore locations smarter with AI-powered geospatial intelligence.

02/08/2026