Questions and plain answers.
Data, freshness, formats, compliance, licensing, refunds and the difference between the files and the actors.
All questions
What exactly is in a LatSouth file?
Company records from the official registry of each country, rebuilt into one English schema: tax identifier, legal and trade name, legal form, status, sector, native activity code with English label, ISIC Rev.4 section, division and class, size class, employee bracket where published, region, city and address, corporate e-mail domain, website, registration date, the registry extraction date and the source credit. Brazil-only files add the registered phone and e-mail with an accountant-contact flag.
Which countries are covered, and how fresh is each one?
Brazil (Receita Federal, monthly), Mexico (INEGI DENUE edition 05/2026), Chile (SII nómina 2026-08 and RES), Colombia (RUES, monthly pull), Peru (SUNAT Padrón RUC, monthly), Argentina (Registro Nacional de Sociedades, cut 2026-09-18) and Ecuador (SRI Catastro RUC, monthly). Every row carries data_as_of, and each country page shows its date.
How is 'active' defined?
By the registry's own status field wherever it exists: Brazil situação cadastral 02, Chile no término de giro, Colombia not cancelled or in liquidation and renewed in 2025-26, Peru estado ACTIVO, Ecuador estado ACTIVO. Mexico's DENUE lists operating establishments only. Argentina publishes no status, so the site and the actor use the activity-row signal (status = active_activity) as the proxy and say so.
Do you sell personal data or contacts of people?
No. The product is company data by design: natural persons and individual entrepreneurs are removed at build time, partner and director tables are never loaded, and sole-owner legal forms whose name is the owner's name are masked. Where a registry publishes a company's filed phone or e-mail (Brazil, Mexico, part of Colombia) it ships in that country's file as a company contact, never in the LATAM unified set.
What formats do I get, and how is the file delivered?
CSV (UTF-8, comma-separated, header row) and Parquet with identical columns, plus a data dictionary and the source credit, delivered as a zip download or a signed cloud URL within one business day of purchase. Files open in Excel and Google Sheets (large files need a split), pandas, DuckDB, BigQuery and Snowflake.
How are the row counts on this site measured?
With DuckDB queries on the current build at site-build time; the filter behind every count is published on the methodology page and in the data dictionary so you can reproduce it on the raw open data. Counts change slightly with each registry release.
Is the website column verified?
For Brazil, website is the HTTP-verified domain (2xx/3xx answer in the release month) and website_verified says whether a check ran; website_guess is the corporate e-mail domain before verification. For Mexico, website is the one declared to INEGI, unverified. Chile, Peru, Argentina and Ecuador publish no website or e-mail, so those columns are empty there.
Why are some names shown as ***?
Legal forms whose name must carry the owner's name (EIRL in Chile and Peru, EIRELI and single-member law firms in Brazil, empresa unipersonal in Colombia, E.U.R.L. in Ecuador, S.A.U. in Argentina) and partnerships named after their partners are masked. The row stays, with its activity, location and identifier, so counts are complete; the name is withheld because it is a person's name.
Can I resell or redistribute a file?
No, with one exception. The licence is internal use by the purchasing company (CRM, analysis, outbound, products that use the data without exposing it). Colombian rows are CC BY-SA 4.0 at the source, so they are delivered as a separate file that you may share onward under the same licence; we do not restrict that.
What is the difference between the files and the Apify actors?
Same data, different access. The actors (Brazil, Mexico, LATAM) are pay-per-row for filtered pulls from code or an agent: US$3 per 1,000 rows for Brazil and Mexico, US$4 for LATAM. The files are flat-priced segments or whole registries for teams that want the universe in their warehouse, with an optional monthly refresh.
What is the refund policy?
Seven-day integrity guarantee: if the file has fewer rows than published for its data_as_of, different columns from the data dictionary, or does not open, you get a full refund. There is no refund because the segment did not contain the accounts you hoped for; the free samples and the published counts exist so you can check before buying.
Can I get a custom cut?
Yes: Cut (US$450, one country, up to 50,000 rows), TAM (US$900, up to 7 countries, up to 250,000 rows, with a TAM memo) and TAM + verified (US$1,500, adds HTTP verification of up to 25,000 domains). You get a row count and a quote within one business day, before paying anything.
Invoices, VAT and procurement?
Checkout is being moved to a merchant-of-record provider that issues invoices and handles VAT automatically. Until then, every purchase is an e-mail to [email protected] with the package name; you receive a quote, an invoice and the file by signed URL. Purchase orders and vendor forms are fine.