2,550 municipalities have received their first unit of a measured franchise network since 1990, and the median population of these municipalities fell from 39,300 inhabitants in the 1990s to 11,500 since 2020

The CNPJ (Brazilian company registry) records register the municipality and the start-of-activity date of each establishment, which makes it possible to date the arrival of the first unit of a franchise network in each municipality. Of the 5,570 municipalities in the 2022 Census, 2,971 have at least one CNPJ attributed to the 503 networks measured by the observatory, in any registration status. In 363 of them the first unit predates 1990, and in 2,550 it arrived between 1990 and the reference month of October 2026. A unit that opened and later had its CNPJ deregistered or left the network counts as an arrival, by the CNPJ start date, and 433 of these 2,550 municipalities currently have no active establishment of the measured networks.

First arrivals ranged from 42 to 97 per year in the 1990s, reached 140 in 2021, when 4.52% of the municipalities still without a record received their first unit, and fell back to 46 in 2025, with 27 in 2026 up to the reference month. The decline holds in the 338 networks that already had 30 CNPJs before 2021, with 132 first arrivals in 2021 and 30 in 2025. The size of the municipalities reached decreased over the period. The median population in the 2022 Census went from 39,300 inhabitants between 1990 and 1999 to 24,800 between 2000 and 2009 and 11,500 between 2020 and 2026, and the share of municipalities with up to 20,000 inhabitants among first arrivals rose from 20.0% to 80.9%.

Between 1990 and 1999, health, beauty and wellness accounted for 55.6% of first arrivals, 81.0% of them from O Boticário, and kept the lead between 2000 and 2009, with 54.1%. O Boticário exceeded 50% of the country's first arrivals in 1993, 1999, 2000 and 2001. From 2020 to 2026, the categories with the most first arrivals are dentistry (30.2%) and chocolate (16.1%), both with a single network above half of the category's arrivals (OdontoCompany, with 61.7%, and Cacau Show, with 100.0%). Among first arrivals, pharmacy has a share 7.1 to 10.0 times the share it holds in the other openings of the measured networks in each of the three periods (4.7% versus 0.7% from 2020 to 2026), with no network above half of the arrivals. The ratio does not support the conclusion that the category reaches municipalities earlier. In the measured pharmacy networks, 2.6% of the CNPJs predate the first identified unit of the network itself, compared with 0.5% across all networks, an indication that the category brings together existing drugstores that joined the network's banner, and this share is a floor, because joining after the network's first unit cannot be distinguished in the records. Since the arrival date is that of the CNPJ opening, not that of joining, the older company brings forward the network's arrival in the municipality, and the share partly reflects the age of these companies. Language schools accounted for 19.0% of first arrivals from 1990 to 2009 and 5.4% from 2020 to 2026. Açaí, with 5.1% of first arrivals from 2020 to 2026, has a share close to the one it holds in the other openings, 4.0%, and does not stand out as an entry category.

The second measured network arrives a median of 8.4 years after the first, by the Kaplan-Meier estimate, which takes into account the municipalities still without a second network up to the reference month. The interval follows size, with a median of 3.6 years in municipalities with more than 100,000 inhabitants, 9.8 years in those with 10,000 to 20,000 and 19.3 years in those with 5,000 to 10,000, and, in those with up to 5,000 inhabitants, the median is not reached, with 5.8% receiving the second network within 10 years. In the most recent arrivals the interval has shortened, and 21.3% of the municipalities reached from 2020 to 2026 received the second network within two years, compared with 13.1% of those reached from 1990 to 1999. Of the 2,550 municipalities reached since 1990, 782 have no CNPJ of another measured network up to the reference month.

How we calculate

We start from the CNPJs that the observatory's census attributes to the 503 networks with at least 30 establishments, in the open CNPJ data of Receita Federal (the federal tax authority), reference month October 2026, by the trade name expression and the CNAE (national classification of economic activities) subclasses of each network profile, in any registration status (active, deregistered, classified as inapt and suspended). Added to these are the 6,049 CNPJs that had the name of one of these networks in the snapshots of the records of May 2023, October 2024, October 2025 and September 2026 and no longer have it, as in the indicator of municipalities without a network. The Receita Federal municipality code is converted into the IBGE (the national statistics institute) code by name and state, and size is the resident population in the 2022 Census (SIDRA, table 4714) in the 5,570 municipalities enumerated.

The first unit of a municipality is the CNPJ with the earliest start-of-activity date. Two types of record do not date the arrival, although the municipality is still counted as reached. The first is a CNPJ opened before the first identified unit of the network itself (529 CNPJs), usually a company that predates the brand and adopted it later, the network's first unit being the earliest date followed by at least two other CNPJs within five years, as in the indicator of network births. The second is a CNPJ opened in a batch, with 5 or more CNPJs of the same company on the same day (3,293 CNPJs), the criterion of the opening indicators, because the date reflects a registration act. In these cases the next dated CNPJ applies, unless an undated record may have arrived before it (a batch opened earlier, or a CNPJ predating the brand of a network that already existed), in which case the date is left undetermined. The date becomes that of a later CNPJ in 49 municipalities and is undetermined in 58, which are counted as reached and left out of the series. Arrivals before 1990 are included only in the total of municipalities already reached, because the oldest dates in the records concentrate companies that predate the brand, with 139 of the 529 CNPJs predating the network itself opened before 1990, and records of deregistrations from that period are scarce, with 37 deregistrations before 1990 in the measured networks, compared with 896 openings.

The entry rate divides the year's first arrivals by the municipalities still without a dated record at the start of the year. The second network is the first dated CNPJ of another measured network in the municipality, on the date of the first arrival or later, and the time until it is estimated by the Kaplan-Meier method, in which a municipality still without a second network is followed up to October 8, 2026, the effective date of the reference month, without counting as an event. The segment and category are those in the first network's profile, compared with the other openings, that is, the dated CNPJs of the same networks opened in the same period that are not first arrivals. This comparison excludes the 20 networks with a namesake risk identified when reading the trade names or with a website that lists fewer than half of the attributed CNPJs, and another 5 in which reading a sample of the trade names of first arrivals showed a similar indication. Cells with fewer than 5 municipalities are not published, with secondary suppression when the difference would reveal the cell, and the tables indicate the network that exceeds 50% of the first arrivals in an aggregate.

The scope differs from that of other indicators in the observatory. The indicator of presence in municipalities counts municipalities with an active establishment today (2,526 in this same base), and this indicator also counts those that only had a unit that has since closed. The indicator of municipalities without a network uses the same records with the original date of all CNPJs and counts 231 entries since 2023, compared with 230 on this page, a difference that comes from the treatment of dates. The indicator of openings outside state capitals counts all openings by location, not only the first in each municipality, and the indicator of networks in small towns counts each network's active CNPJs in municipalities with up to 50,000 inhabitants.

Calculated on November 8, 2026. Method version 1.0. Sample: 5,570 municipalities, 503 measured networks, 107,541 CNPJs in any status.

Report an error on this page (in Portuguese)