Size and presence
In 9 of 13 franchise categories, units opened from 2015 to 2020 remain active at 5 years in a larger share than other establishments in the same activities
Categories in which network units opened from 2015 to 2020 exceed the standardized comparison at five years, with a 95% confidence interval above zero
See the full table (13 rows)
| Category | Networks | Units followed | Active at 5 years | Standardized comparison | Difference (p.p.) | 95% CI of the difference | Reading |
|---|---|---|---|---|---|---|---|
| Compounding pharmacy | 7 | 108 | 97.2% | 77.8% | +19.3 | +9.2 to +28.1 | Above the comparison |
| Dentistry | 19 | 1,814 | 87.8% | 79.7% | +8.1 | +3.0 to +13.8 | Above the comparison |
| Intimate apparel | 6 | 152 | 87.5% | 57.2% | +30.3 | +21.3 to +39.3 | Above the comparison |
| Vocational courses | 7 | 414 | 83.3% | 68.0% | +15.3 | +6.0 to +25.4 | Above the comparison |
| Pharmacy | 11 | 171 | 83.0% | 70.4% | +12.6 | +7.4 to +18.0 | Above the comparison |
| Footwear | 27 | 898 | 82.3% | 62.8% | +19.4 | +13.9 to +24.4 | Above the comparison |
| Vehicle inspection | 13 | 423 | 79.7% | 72.6% | +7.1 | +0.8 to +13.7 | Above the comparison |
| Optical store | 17 | 1,316 | 79.1% | 71.8% | +7.3 | -4.5 to +13.8 | No statistical difference |
| Language schools | 15 | 646 | 72.9% | 65.8% | +7.1 | +0.6 to +14.5 | Above the comparison |
| Japanese food | 6 | 108 | 72.2% | 58.0% | +14.2 | +4.8 to +23.5 | Above the comparison |
| Açaí | 13 | 762 | 58.9% | 52.4% | +6.5 | -0.7 to +15.8 | No statistical difference |
| Credit and purchasing consortia | 8 | 205 | 55.1% | 63.7% | -8.6 | -16.4 to +2.8 | No statistical difference |
| Coffee shop | 12 | 454 | 52.2% | 52.5% | -0.3 | -8.4 to +8.9 | No statistical difference |
When only deregistration counts as terminationThe narrower criterion disregards inapt status and suspension. The shares rise in both groups, and the difference decreases in 12 of the 13 categories.
| Category | Without deregistration at 5 years | Standardized comparison | Difference (p.p.) |
|---|---|---|---|
| Compounding pharmacy | 97.2% | 82.8% | +14.3 |
| Dentistry | 90.3% | 82.9% | +7.4 |
| Intimate apparel | 88.8% | 65.7% | +23.1 |
| Vocational courses | 85.5% | 73.2% | +12.4 |
| Pharmacy | 87.1% | 79.1% | +8.0 |
| Footwear | 84.0% | 70.8% | +13.2 |
| Vehicle inspection | 82.5% | 78.2% | +4.3 |
| Optical store | 82.1% | 78.8% | +3.4 |
| Language schools | 80.0% | 72.9% | +7.2 |
| Japanese food | 77.8% | 69.3% | +8.4 |
| Açaí | 60.4% | 63.4% | -3.1 |
| Credit and purchasing consortia | 61.5% | 70.7% | -9.3 |
| Coffee shop | 57.5% | 63.3% | -5.8 |
The CNPJ (Brazilian company registry) makes it possible to follow the units of each business category from opening and compare them with other establishments that carry out the same activity and opened in the same year. Across the 478 measured networks with a unit opened in the period, 72.9% of the 18,527 units opened from 2015 to 2020 were still active five years later, compared with 63.4% in the standardized comparison.
The breakdown by category shows how much of this difference holds when the variation between networks in the same category is taken into account. Of the 13 categories with a robust confidence interval, which together account for 7,471 units, 9 are above the comparison, 4 are indistinguishable from it and none is below. The share active at five years reaches 97.2% in compounding pharmacies, and the largest distance from the comparison is in intimate apparel, at 30.3 percentage points.
The universe is that of networks operating today, which raises the networks' share, and the comparison is descriptive. The table follows the order of the values and does not constitute a ranking of categories.
How we calculate
We started from the head-office establishments opened from 2015 to 2020 that the census attributes to the measured networks, with at least 30 active CNPJs, in the CNPJ base of the reference month of October 2026, each network in its main category. Individual microentrepreneurs are excluded, based on the MEI (individual microentrepreneur regime) option in force at opening. A unit counts as active when it had no deregistration, inapt status or suspension dated before the 5-year anniversary of its opening. The comparison brings together the other head-office establishments with a trade name in the same CNAE (national classification of economic activities) subclasses, outside the CNPJ roots of the universe, and its share is calculated in each combination of subclass and year of opening and weighted by the number of units of the category in that combination.
The 95% interval treats the network as a cluster, because units of the same network are not independent observations. We publish the wider of the calculation with the t distribution and the resampling of networks, plus the sampling error of the comparison. A category is included with at least 5 networks in the cohort, standardization covering 80% of the units, a largest network with up to 50% of the units and a half-width of up to 10 percentage points. A reading above or below the comparison requires an interval excluding zero in the published value, to one decimal place, and a sign that holds when any single network is removed.
Of the 35 categories with 5 networks and 30 active CNPJs, 1 lack a sufficient cohort, 14 have an interval above the limit and 7 are left out because one network accounts for more than half of the CNPJs or of the cohort.
What this figure does not show
- There is survivorship bias in the universe. Only networks existing today, with at least 30 active CNPJs, are included, so networks that have closed or shrunk are not observed, which raises the networks' share in all categories.
- The trade name is the current one. A unit that left the network and changed its name drops out of the set, whether active or terminated.
- Registration status does not always coincide with the operation of the location. There are dormant companies without deregistration and units sold with a change of CNPJ, which appear as a termination and an opening.
- Only head offices are included, and a unit opened as a branch of a franchisee with more than one location is left out.
- The comparison brings together independent businesses, head offices of company-owned chains and units of networks not tracked. The difference is descriptive and does not measure a causal effect of the franchise model.
- The category is a classification by the observatory, based on the network's name, CNAE code and offering, and each network is placed only in its main category.
- With few networks, the cluster interval tends to be too narrow, which is why the wider of two calculations is adopted.
Frequently asked questions
Which type of franchise lasts longest?
Among the 13 categories with a robust confidence interval, the largest share of units active at five years is in compounding pharmacies, a category with 7 measured networks, in which 97.2% of the 108 units opened from 2015 to 2020 were still active, compared with 72.9% across all measured networks. The share does not measure the duration of the location, because the universe is that of networks in operation and a unit that changes its name drops out of the set. In this category, of the active CNPJs (Brazilian company registry) with a network brand in May 2023, 18 remained active under another trade name in October 2026 and 28 had been deregistered. The category's active CNPJs changed by +38.1% in that period.
Do franchises survive longer than independent businesses in every sector?
Not in all. Of the 13 comparable categories, 9 are above other establishments in the same activities, 4 are indistinguishable from them and none is below, with a 95% interval and the network as the cluster.
How many franchises are still open after five years?
In the 478 measured networks with a unit opened in the period, 72.9% of the units opened from 2015 to 2020 had an active registration five years later, according to the CNPJ (Brazilian company registry) records. An active CNPJ does not necessarily equal a location in operation.
Report an error on this page (in Portuguese)
Size and presenceOther indicators in the same group
Maintained by Marktech, which serves franchise networks and franchises. The brands cited belong to their owners and are mentioned here only to identify the network the data refers to.