Saltar al contenido
Volver a guías

BlackPrint · Guides · Site selection

How to choose where to open a store in Mexico

Retail site selection in Mexico usually starts with a circle on a map and ends with a lease. This is the method that replaces the circle: seven measurements — catchment, resident demand, flow and rhythm, competition, leakage, cannibalization and cost — ranked on one yardstick, with numbers from real studies.

Updated November 10, 2026 · BlackPrint Team

95.6winning corridor index, 4 points above the second

The essentials, in 20 seconds

  • Seven things are measured on one yardstick: the real catchment with isochrones (not a radius), resident demand by census block with its income tier, and flow with its rhythm by day and hour
  • Then competing supply, consumption leakage, cannibalization of your own stores and occupancy cost are discounted
  • The deliverable is a ranking with each factor's weight declared: in the real case, the winner scored 95.6 against 91.6 for the second

01 · Radius vs. isochrone

A radius bills you for 8.7 km² of market the road network never delivers.

Almost every site study in Mexico starts the same way: draw a 1, 3 or 5 km circle and count what falls inside. The circle assumes people move in straight lines, equally in every direction. People move where there are streets.

In a 15-minute catchment, a simple radius covers 21.1 km² while the road-network isochrone covers 12.4 km², 41% less (BlackPrint site-selection method).

−41%of area: 21.1 → 12.4 km² for the same travel time

Radius against isochrone: same travel time, two different markets

Diagram: a 15-minute radius crossing a barrier covers 21.1 km²; the 15-minute isochrone on the real road network covers 12.4 km²

Diagram (labels in Spanish): radius on the left, road-network isochrone on the right.

15-minute radiussymmetric travel in every direction
21.1 km²
15-minute isochronereal road network, with barriers
12.4 km²
▲ THE READ
  • A case built on the circle carries 8.7 km² nobody crosses in fifteen minutes (−41%).
  • A real catchment is asymmetric: it stretches along corridors and stops at barriers — rivers, highways, rail lines.
  • It is modeled with isochrones — curves of equal travel time on the existing road network — and validated against where visitors actually come from.

Method diagram; areas are illustrative.

02 · Who lives around MEASURED CASE

The area has less high income than the city, but more people working.

The absolute number of households says nothing on its own. What decides is the contrast against the city average: above or below, and on what. This is how a 1,500-meter catchment read in a real study in Querétaro.

In a measured Querétaro site study, a 1,500 m catchment held 22,671 households with 25.8% in the top income tiers, 13.1 points below the city (BlackPrint, 2026).

22,671households in the area · 8.9% of the city

1,500 m catchment against the city average

22,671households in the area · 8.9% of the city
25.8%top socioeconomic tiers (A/B + C+) · −13.1 pp vs. the city
69.7%economic participation · +2.6 pp vs. the city
3.36people per dwelling · city: 3.42
▲ THE READ
  • Fewer high-income households (−13.1 pp) and more people working (+2.6 pp) than the city average.
  • That steers the site: it defines whom to serve and with what product, and rules out premium formats from the start.

Socioeconomic tiers follow the AMAI rule used in Mexico (AB to E).

03 · Flow and rhythm OWN STUDY

The two corridors peak five hours apart: 7 p.m. against 2 p.m.

The most common mistake when reading foot or vehicle traffic is mixing two questions: how many people pass and when they pass. Volume only means something against a known reference. "3,000 people a day" informs no one; "this point moves 8 of every 100 pedestrians a high-traffic reference point moves" places the candidate on a scale.

In a 2026 BlackPrint site study, the candidate corridor peaked at 7 p.m. and the reference at 2 p.m., and the candidate out-moved the reference in 20 of 24 hours.

20 of 24hours where the candidate moves more relative flow than the reference

Pedestrian rhythm by hour, normalized to each corridor's own peak

Line chart of hourly pedestrian rhythm: candidate peaks at 19 h, reference at 14 h

Hourly rhythm (labels in Spanish): candidate in pink, reference in blue; 100 = each corridor's own peak.

7 p.m.candidate peak · evening
2 p.m.reference peak · lunchtime
36.1candidate floor · never below 36
22.7reference floor · drops overnight
▲ THE READ
  • The candidate moves more relative flow than the reference in 20 of 24 hours; the largest gap is +14.9 points.
  • They share no useful hour: one opening schedule for both leaves out five hours of each one's peak.

Rhythm by day of week · candidate against reference

Monreference 88.1
93.8
Tuereference 94.8
96.7
Wedreference 100
97.8
Thureference 97.5
96.7
Frireference 95.9
100
Satreference 83.2
96
Sunreference 68.8
87.2
▲ THE READ
  • Sunday separates the corridors most: 87.2 against 68.8 (+18.4 points) — put staff on the floor there before weekdays.
  • The candidate is a steady-flow corridor: 12.8 points of range against 31.2 for the reference.
  • Presence is not spending: the "how much" is confirmed with the ticket observed in comparable stores, never assumed from the count.

Site-selection study for a motorcycle brand, Greater Mexico City, 2026. Index: 100 = busiest day.

04 · Traffic data in Mexico METHOD

A traffic number with no reference cannot be audited.

This is where the market has the most smoke. Mexico's public counts are vehicle counts on the highway network and some primary roads. There is no national public pedestrian count. Manual counts work for one intersection on one day — not for comparing fifty candidates or reading a weekly pattern.

3questions every traffic number has to answer

Three questions every traffic number must answer, whoever supplies it

Expanded against what reference?

A number with no reference cannot be audited or compared between points.

From what period, and how many days?

A long weekend, a season or road works distort a whole week.

Per pass-by point or per whole zone?

Two different metrics. Mixing them inflates or deflates a candidate as convenient.

The measurement that scales

Aggregated, anonymous mobile-phone samples: trajectories without individual identity, expanded to population and standardized against known references.

▲ THE READ
  • BlackPrint operates that data as its own provider and standardizes it — the part that actually costs: without standardization, two measurements of the same corridor in different months cannot be compared.
  • If a provider cannot say what it is expanded against, how many days it covers and whether it is per point or per zone, the number is fit for a slide, not for a multi-year lease.

05 · Consumption leakage MEASURED CASE

Almost 80% of the spending born in the area is spent outside it.

Leakage is the share of consumption that originates in an area and is spent outside it. It is uncomfortable and therefore useful: it is measured with where spending actually goes, not with what people say in a survey. This is how it read in a study in Peñaflor, Querétaro.

In a measured site in Peñaflor, Querétaro, 79.4% of consumption originating in the area was spent outside it; half of visitors came from within 0.48 km (BlackPrint, 2026).

79.4%of the area's consumption is spent elsewhere

Leakage and visitor origin, in four numbers

79.4%of the spending born in the area is spent outside
≤0.48 kmdistance from which half the visitors come · median
6.53 kmwhere the 90th percentile stretches · the long tail
48%of origin households are tier C · dominant profile
▲ THE READ
  • The capturable market is in the flow that already passes the site: hence the short median (0.48 km) and the long tail (6.53 km).
  • With almost 80% leakage, local demand exists and nobody is serving it from inside.

06 · The three discounts METHOD

Competition, cannibalization and cost don't add market: they subtract it.

The three remaining measurements are the ones that turn estimated demand into sales that actually come in.

3discounts between demand and the sales that come in

What each discount measures and what disqualifies a candidate

Competing supply

Measured by spacing, not by count: ten competitors spread across the catchment weigh less than three right at the candidate's entrance. Together with traffic-generating anchors, it defines how much of the passing flow can be captured.

Own cannibalization

What the new store takes from your own: catchment overlap weighted by households and spend. It is the difference between gross and incremental sales (how it is measured).

Occupancy cost

Rent, maintenance and fit-out against that site's projected sales, not the chain average. It is the factor that most often moves a candidate from first place to third.

▲ THE READ
  • Gross sales are not incremental sales: a store opening near one of your own can lift the area's sales and leave the chain's flat.
  • The number to defend in the investment committee is the one that has already discounted the overlap.

07 · One yardstick OWN STUDY

The top two finished 4 points apart: the real decision was between two, not four.

After measuring comes deciding — and comparing candidates without a common scale is comparing opinions. An attractiveness index normalizes each factor against a reference and combines them on a 0–100 scale. Its value is that you can open the box and see which factor pulled up and which one penalized.

In a 2026 BlackPrint site study, the winning corridor scored 95.6 of 100, 4.0 points above the runner-up, winning on 3.92× foot traffic and 0.56× rent.

95.6winning index · +4.0 points over the second

Four corridors scored with the same index · 0 to 100

Corridor Dwinner
95.6
Corridor C−4.0 pts
91.6
Corridor B−28.2 pts
67.4
Corridor A−41.6 pts
54.0
▲ THE READ
  • The winning corridor scored 95.6, just 4.0 points above the second (91.6): a practical tie.
  • The other two finished 28.2 and 41.6 points behind — visible only on the same scale.
  • The corridor won on flow and cost: 3.92× the reference's foot traffic and 0.56× its rent, with an income profile at 1.02×.

Site-selection study for a motorcycle brand, 2026. Corridors anonymized.

How you get to four finalists

Starting point · available inventory

Units and lots on the target corridors, with operable area and frontage for the format.

Cut 01 · Coverage

The candidate's isochrone must reach where the target customer lives or passes.

Cut 02 · Flow

Minimum volume against the format's reference, and a peak compatible with the selling window.

Cut 03 · Cost and operation

Rent against projected sales, access, parking and zoning permission.

▲ THE READ
  • Running all seven factors on the whole inventory is expensive and does not change the result: cut first what disqualifies fast, measure last what is costly.

08 · The method in seven steps METHOD

Seven measurements, in order, each with its source.

From catchment to final ranking. Each step leaves a number the next one uses.

7measurements and a single scale

The method, start to finish

01 · Draw the real catchment

Walking and driving isochrones on the road network, at the hour that matters for the format, validated against observed visitor origin.

02 · Size resident demand

Households, population, AMAI socioeconomic tier and economic participation by AGEB — the smallest official census unit — always against the city average.

03 · Measure flow and its rhythm

Pedestrian and vehicle, by format. Volume against a known reference, and shape normalized to its own peak by day and hour.

04 · Quantify competing supply

Establishments in the category, their spacing and the anchors that generate traffic.

05 · Measure consumption leakage

What share of the spending born in the area is spent elsewhere.

06 · Discount own cannibalization

Overlap with your stores weighted by households and spend, to go from gross to incremental sales.

07 · Rank with one yardstick

A comparable index across candidates, with the breakdown by factor always visible.

Frequently asked questions

How do you choose where to open a store in Mexico?
Seven things are measured and compared with the same yardstick: the real catchment with isochrones (not a radius), resident demand by census block with its socioeconomic tier, and foot and vehicle flow with its rhythm by day and hour. Then competing supply, consumption leakage, cannibalization of your own stores and occupancy cost are discounted. The output is a ranking with the weight of each factor declared.
Why doesn't a kilometer radius work?
Because people move along streets, not in straight lines. In the example in this guide, a 15-minute radius covers 21.1 km² and the real road-network isochrone covers 12.4 km²: the circle adds 8.7 km² of market nobody can reach in that time.
How is foot traffic measured in Mexico?
There is no national public pedestrian count. The method that scales uses aggregated, anonymous mobile-phone samples, expanded to population and standardized against known reference points so that one location can be compared with another.
What is consumption leakage?
The share of spending that originates in an area and is spent outside it. In a measured site in Querétaro it was 79.4%, meaning local demand existed but nobody was capturing it locally.
What is a site attractiveness index?
A 0–100 score that normalizes each factor (flow, income, competition, rent, cannibalization, leakage) against a reference and combines them, so candidates can be ranked on one scale. In the case shown, the winner scored 95.6 against 91.6 for the second.

Your territory

This analysis covers the city. Yours is about your corner.

If you have a decision in front of you — where to open, which site to lease, how much competition you walk into — we measure it with data for your exact area, not with city averages.